25
The End of Affirmative Action in Washington State and Its Impact on the Transition from High School to College Susan K. Brown University of California, Irvine Charles Hirschman University of Washington Changes in affirmative action policies in some states create possibilities for “natural experi- ments” to observe the effect of public policy on racial and ethnic inequality in American soci- ety. This study measured the impact of Initiative 200, a ballot measure that eliminated affir- mative action in Washington State, on the transition from high school to college. As of 1999, the year after I-200 passed, the proportion of minority high school seniors who went to col- lege in Washington State decreased temporarily. The impact of I-200 was registered almost entirely at the University of Washington, the flagship public institution in the state. This decrease, however, stemmed less from changes in minority admission rates than from declines in application rates. Affirmative action programs may provide a signal of an institutional “wel- coming environment” that serves as a counterweight to the normal reluctance of prospective students to apply to institutions that may be perceived as intimidating. Although the impact of I-200 was short-lived, significant racial and ethnic differences remain in the transition from high school to college. Sociology of Education 2006, Vol. 79 (April): 106–130 106 F ollowing the civil rights movement of the 1950s and 1960s, the United States began to dismantle the apartheid struc- ture of racial segregation and inequality that had haunted American society since the end of slavery. During the second half of the 19th century and the first half of the 20th century, African Americans and other minority groups experienced separate and unequal opportuni- ties in employment, voting, housing, and education. Discrimination was woven into the fabric of American society and openly prac- ticed by individuals; firms; social groups; schools and colleges; and governmental agencies at the federal, state, and local levels. The end of official discrimination in the 1960s did not create a fully equal opportunity soci- ety, however. Overt forms of discrimination have decreased, but have sometimes been replaced by more subtle actions, which have discriminatory consequences, if not inten- tions. Moreover, the historical legacy of discrimi- nation can continue to affect racial and ethnic inequality, even in the absence of contempo- rary discrimination, because minorities gener- ally begin life in poorer circumstances and have fewer resources for social mobility. The Delivered by Ingenta to : University of Washington Fri, 16 Jun 2006 17:53:42

The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

The End of Affirmative Action inWashington State and

Its Impact on the Transition from HighSchool to College

Susan K BrownUniversity of California Irvine

Charles HirschmanUniversity of Washington

Changes in affirmative action policies in some states create possibilities for ldquonatural experi-

mentsrdquo to observe the effect of public policy on racial and ethnic inequality in American soci-

ety This study measured the impact of Initiative 200 a ballot measure that eliminated affir-

mative action in Washington State on the transition from high school to college As of 1999

the year after I-200 passed the proportion of minority high school seniors who went to col-

lege in Washington State decreased temporarily The impact of I-200 was registered almost

entirely at the University of Washington the flagship public institution in the state This

decrease however stemmed less from changes in minority admission rates than from declines

in application rates Affirmative action programs may provide a signal of an institutional ldquowel-

coming environmentrdquo that serves as a counterweight to the normal reluctance of prospective

students to apply to institutions that may be perceived as intimidating Although the impact

of I-200 was short-lived significant racial and ethnic differences remain in the transition from

high school to college

Sociology of Education 2006 Vol 79 (April) 106ndash130 106

Following the civil rights movement of the1950s and 1960s the United Statesbegan to dismantle the apartheid struc-

ture of racial segregation and inequality thathad haunted American society since the endof slavery During the second half of the 19thcentury and the first half of the 20th centuryAfrican Americans and other minority groupsexperienced separate and unequal opportuni-ties in employment voting housing andeducation Discrimination was woven into thefabric of American society and openly prac-ticed by individuals firms social groupsschools and colleges and governmental

agencies at the federal state and local levelsThe end of official discrimination in the 1960sdid not create a fully equal opportunity soci-ety however Overt forms of discriminationhave decreased but have sometimes beenreplaced by more subtle actions which havediscriminatory consequences if not inten-tions

Moreover the historical legacy of discrimi-nation can continue to affect racial and ethnicinequality even in the absence of contempo-rary discrimination because minorities gener-ally begin life in poorer circumstances andhave fewer resources for social mobility The

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 107

dilemma was expressed in 1965 by PresidentLyndon Johnson ldquoYou do not take a manwho for years has been hobbled by chainsand liberate him bring him up to the startingline in a race and then say lsquoYou are free tocompete with all the othersrsquo and still believethat you have been completely fairrdquo (quotedin Bowen and Bok 19986)

A bold response to this dilemma was theintroduction of affirmation action policieswhich sought to increase educational andemployment opportunities for minoritygroups who had historically experienced dis-crimination The origins effectiveness andunintended consequences of affirmativeaction policies have been among the mostcontested issues in American politics andsocial science research (Skrentny 2001)Without evidence from the counterfactual ofminority socioeconomic progress in theabsence of affirmative action policies thedebates over the presumed effects and meritsof affirmative action have often generatedmore heat than light

In this article we take the opportunity pro-vided by a change in public policy namelythe end of affirmative action in WashingtonState as a ldquonatural experimentrdquo to assess theimpact of affirmative action on minority stu-dentsrsquo transition from high school to collegeOn November 3 1998 voters in WashingtonState ended affirmative action by passing astatewide measure Initiative 200 (I-200) by amargin of 58 percent to 42 percent The ini-tiative was worded ldquoShall government beprohibited from discriminating or grantingpreferential treatment based on race sexcolor ethnicity or national origin in publicemployment education and contractingrdquo

The presumption of the sponsors and sup-porters of I-200 was that affirmative actionpolicies allowed less qualified minority stu-dents to enter universities at the expense ofmajority students The opponents of I-200feared that without affirmative action theeducational progress of minority studentswould be slowed (for more background onthe politics and aftermath of I-200 see Fryer2002) The assumption that I-200 would havean automatic impact was reinforced withwidespread press reports of a decrease inAfrican American Latino and American

Indian freshmen at the University ofWashington (UW) in 1999 following the pas-sage of I-200 (Davila and Mayo 2002)

The decrease in minority enrollments inWashington State is not an isolated case InCalifornia minority applications and enroll-ments dropped dramatically at many cam-puses of the University of California systemfollowing rulings by the universityrsquos Board ofRegents and a proposition passed by statevoters to eliminate race-conscious admissionsdecisions (Lomibao Barreto and Pachon2004 University of California 2003) Thedeclines were the steepest at the flagshipcampuses of Berkeley and Los Angeles whichhave also been among the slowest to recoverFollowing the 1996 Hopwood v Texas caseand the decision of the state of Texas to banaffirmative action in college admissionsenrollments of African American and Hispanicfreshman declined at the University of Texasat Austin (Tienda and Niu 2004) There hasbeen a partial recovery in minority enroll-ments in Texas with the implementation of aplan guaranteeing admission to the top 10percent of graduating high school classes(Horn and Flores 2003) Similar declines inminority enrollments have been reported inFlorida (Marin and Lee 2003) Georgia (Goseand Schmidt 2001) and Michigan (Dobbs2004 Winter 2004) as affirmative action pro-grams have come under challenge and havebeen modified or eliminated

The number and composition of collegefreshmen from year to year are shaped how-ever by demographic and economic forces aswell as by public policy The number of newstudents who enter colleges and universitiesdepends on the supply of students (the num-ber and composition of potential applicants)as well as institutional decisions about whomto admit and studentsrsquo acceptances of offersof admission The appropriate index to evalu-ate the impact of I-200 (or other policies) oncollege attendance is the ldquotransition raterdquo orthe percentage of in-state high school seniorswho enroll in Washington State colleges anduniversities The first objective of our analysisis to estimate transition rates by race and eth-nicity and to compare the 1999 rates after thepassage of I-200 with rates for earlier andlater years We also examine transition rates

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

108 Brown and Hirschman

for different types of four-year colleges in thestate

The transition from high school to collegeis a sequential process of decision making inwhich admission to a university is a criticalstep but only one stage in the processStudents (with advice and encouragementfrom others) take the first step by applying tocolleges and then make the final decision bydeciding whether and where to enrollassuming that they have been admitted toone or more colleges Through a detailedanalysis of administrative records from UWwe discovered that the decline in minoritystudents in 1999 was partially affected bychanges in ethnic-specific admission ratesbut primarily the result of declines in collegeapplications from prospective minority stu-dents

The decline in applications from minoritystudents included many students who werealmost certain to have been admitted on uni-versalistic standards of grades and test scoresWe interpret these findings as suggestive ofan unexpected impact of affirmative actionpolicies namely as a symbolic beacon of awelcoming environment Minority studentsfrom modest backgrounds may be reluctantto apply to the most competitive anddemanding colleges and universities as maymany majority students Universitiesrsquo affirma-tive action policies are often accompanied byoutreach efforts that provide informationpersonal contacts and encouragement forminority students to apply Even without aformal policy of affirmative action these out-reach effects may have spurred the reboundin minority application rates to UW in recentyears

AFFIRMATIVE ACTION AS A POLICY RESPONSE

For the past three decades affirmative actionprograms have been a bulwark of govern-mental efforts to redress racial ethnic andgender inequality in the United States(Bowen and Bok 1998 Holzer and Neumark2000 Reskin 1998) Nonetheless the effectsof affirmative action efforts remain uncertain

and controversial The problem of assessingthe effects of affirmative action arises not onlybecause programs are so varied but becauseit is difficult to measure discriminationmdashthephenomenon that affirmative action isdesigned to prevent or remedy A fewdecades ago racial and gender discrimina-tion was openly practiced (even in newspa-persrsquo want ads) and often reinforced by statelaws In such circumstances it was reasonableto assume that almost all intergroup inequal-ity was caused by discrimination eitherdirectly or indirectly Thus the impact ofefforts to combat discrimination would beevident in any observable changes in segre-gation and inequality

At present overt or formal discriminationhas diminished although informal practicesand everyday expressions of prejudice haveproved remarkably resistant to change(Lelyveld 2001) Discrimination that is subtleand hidden (and perhaps even unconsciousby the actor) is much more difficult to detectand its consequences are much more uncer-tain Over the past 40 years social scienceresearchers have expended considerableeffort and ingenuity trying to measure thechanges in the relative role of discriminationas a cause of racial and ethnic gaps in educa-tion occupations income and access to sub-urban residence The standard method ofestimation is to assume that discrimination isreflected in the net racial or ethnic differencethat remains after multivariate statistical mod-els are used to adjust for all measurable socialbackground characteristics that are unrelatedto discrimination (Duncan 1969 Feathermanand Hauser 1978)

This approach is limited however becauseracial and ethnic inequality is affected notonly by contemporary discrimination but bythe legacy of historical patterns of discrimina-tion that have produced unequal resources(eg family background characteristics)between groups The complex and often elu-sive relationships between historical patternsof discrimination and contemporary inter-group inequality have influenced the originsof affirmative-action programs and have ledto the political vulnerability of these pro-grams

In addition to the removal of discriminato-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 109

ry barriers that are sanctioned by law or cus-tom affirmative action programs can bearrayed along a continuum from ldquoopportuni-ty enhancementrdquo to mandating outcomessuch as proportional representation (Boboand Kluegel 1993) The most basic type ofaffirmative action consists simply of efforts toensure that minorities and women are includ-ed in the pool of eligible candidates throughoutreach programs and a reliance on formalmeans of recruitment rather than on infor-mal networks The second type of affirmativeaction program is one that gives qualifiedminority candidates extra (but not exclusive)consideration For example if several quali-fied candidates were being considered for aposition (or admission) affirmative actionwould give priority to a minority or womancandidate much as geographic diversity oran alumnus parent may give an advantage toothers in a college admission decision Thethird type of affirmative action program andthe one that creates the most political con-troversy is a program that admits or recruitsminorities or women for targeted slots orpositions without consideration of other can-didates

In the famous Bakke case (Regents of theUniversity of California v Bakke) the SupremeCourt justices split over the question of whetherracial quotas could be used to make up for pastracism The deciding vote was cast by JusticeLewis Powell who argued against racial quotasbut suggested that admissions officers in uni-versities could ldquotake race into accountrdquo as oneof several factors in evaluating minority candi-dates (Bowen and Bok 19988) This significantdistinction has allowed colleges to maintainaffirmative action programs as long as they dis-avowed the use of quotas This rather inchoatepolicy was continued with the mixed SupremeCourt decisions in 2003 involving the Universityof Michigan In one case the court rejected6ndash3 the use of a system that provided extrapoints to minorities in undergraduate admis-sions decisions But in Grutter v Bollinger thecourt voted 5ndash4 to allow raceethnicity to con-tinue to be considered as one of several criteriafor admissions to law schools Justice SandraDay OrsquoConnor writing for the majority saidthat affirmative action was still needed(Ogletree 2004243)

At the national level it is probably impos-sible to disentangle the impact of affirmativeaction programs from broader patterns thathave expanded opportunities for minoritystudents to enroll in colleges and universitiesFrom the 1950s through the 1970s the pro-portion of African Americans native-bornHispanics and American Indians with at leastsome college education doubled but theracial and ethnic differences remained widebecause white students were also makingimpressive educational gains (Mare1995173) Although most minority studentswere still handicapped by their disadvan-taged social backgrounds there was a briefwindow in the 1970s when African Americanhigh school graduates were slightly more like-ly to enter college than were whites with thesame socioeconomic and geographic charac-teristics (Jaynes and Williams 1989339)Although affirmative action programs mayhave played an important role in attractingminority students to highly selective institu-tions many other factors including improve-ments in minority household incomesreduced discrimination and rising ambitionswere also important The tenor of the timesduring the 1960s and 1970s was one ofgreater opportunities for able high school stu-dents including those from disadvantagedbackgrounds to aspire to a college educa-tion

For the next two decades college-enroll-ment rates resembled a roller coaster By the1980s the rates of college attendance andgraduation were declining among all groupsespecially disadvantaged minorities (Mare1995) The most plausible reason seems tohave been the decrease in financial aid grantsand the reluctance of students from poorhouseholds to borrow money for college(Jaynes and Williams 1989340ndash345) In the1990s college enrollments began to rise forall groups including minority groupsalthough wide differences by race and ethnic-ity persisted By the fall of 1998 45 percentand 52 percent of 20ndash21-year-old white non-Hispanic men and women respectively wereenrolled in college (Martinez and Curry1999) The comparable figures were 33 per-cent and 37 percent for African American(non-Hispanic) men and women and 22 per-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

110 Brown and Hirschman

cent and 25 percent for Hispanic men andwomen Not all minorities are disadvantagedin terms of educational enrollment howeverThe comparable percentages of Asian andPacific Islander 20 and 21 year olds who wereenrolled in college were 65 percent and 73percent for men and women respectively Infact native-born Asian Americans have hadhigher educational attainments than whitesfor many decades even before World War II(Folger and Nam 1967 Hirschman and Wong1986 Mare 1995)

Affirmative action programs have thepotential to affect both the number of minor-ity students who go on to college and thesestudentsrsquo choice of college These two issueshave typically been conflated in research onaffirmative action because the focus has gen-erally been limited to one college or at mosta few similar colleges Changes in enrollmentat other colleges have rarely been comparedIn the study presented here which investigat-ed the impact of I-200 we examined changesin enrollment for the universe of four-year col-leges and universities in Washington State

If the impact of affirmative action is solelya result of collegesrsquo admission decisions onlya small proportion of (potential) students areaffected by these decisions About 70 percentto 80 percent of colleges and universities inthe United States are nonselective that isthey admit almost all applicants who meetbasic entrance requirements (Bowen and Bok199815) Most of the attention and researchon affirmative action has focused on the rela-tively few colleges and universities that havecompetitive admissions and are assumed tooffer a superior education

Graduates of selective colleges with high-er tuition and average SAT scores havereported above-average earnings (Kane1998 Solmon and Wachtel 1973) and highlevels of job satisfaction and civic involve-ment (Bowen and Bok 1998) and are morelikely to obtain advanced degrees than aregraduates of other types of schools (Wegnerand Sewell 1970) These associations how-ever do not prove that attendance at selec-tive colleges actually causes these outcomesRecent research has suggested that the char-acteristics of students prior to their entryinto selective colleges such as socioeco-

nomic background and motivations mayaccount for many of the positive effects thatare usually attributed to attendance at selec-tive colleges (Dale and Krueger 1999Pascarella and Terrenzini 1991) Since thelabor market probably values many of thesame traits as do college admissions officersthe correlation between entry into competi-tive colleges and labor market success maybe inflated

Affirmative action programs may influencethe college choice of minority students forseveral reasons beyond giving an edge inadmissions decisions Perhaps most impor-tant is financial aid because students from dis-advantaged backgrounds may be reluctant orunable to assume debt Also these studentsmay have different perceptions of academicprograms and campus social life (Chapman1984 Hanson and Litten 1982 HosslerSchmit and Vesper 1999 Manski and Wise1983 McDonough 1997)

Assuming that students especially minori-ty students are looking at potential collegeswith the question of whether ldquostudents likethemselvesrdquo are present public policies thateliminate or modify affirmative action pro-grams may be presumed to affect perceptionsof the relative attractiveness of different col-leges and universities The prospect of fewerldquoin-grouprdquo classmates or a general sense thatminorities are no longer welcome maychange the perceived climate of campusesthat have ended affirmative action policies(Gorov 1997) More broadly changes in affir-mative action policy could influence howhigh school counselors advise students andhow universities organize recruitment efforts(Traub 1999) In the immediate aftermath ofthe ban on affirmative action some minoritystudents especially those with multiple choic-es may enroll in out-of-state colleges espe-cially historically black ones In fact such aspike in out-of-state enrollments occurredamong Washington residents according toour analysis of data from the US Departmentof Educationrsquos National Center for EducationStatistics (httpncesedgovipedsdataasp)In the fall of 1998 just before WashingtonState banned affirmative action 76 studentsfrom Washington enrolled as freshmen in his-torically black colleges In 2000 that enroll-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 2: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 107

dilemma was expressed in 1965 by PresidentLyndon Johnson ldquoYou do not take a manwho for years has been hobbled by chainsand liberate him bring him up to the startingline in a race and then say lsquoYou are free tocompete with all the othersrsquo and still believethat you have been completely fairrdquo (quotedin Bowen and Bok 19986)

A bold response to this dilemma was theintroduction of affirmation action policieswhich sought to increase educational andemployment opportunities for minoritygroups who had historically experienced dis-crimination The origins effectiveness andunintended consequences of affirmativeaction policies have been among the mostcontested issues in American politics andsocial science research (Skrentny 2001)Without evidence from the counterfactual ofminority socioeconomic progress in theabsence of affirmative action policies thedebates over the presumed effects and meritsof affirmative action have often generatedmore heat than light

In this article we take the opportunity pro-vided by a change in public policy namelythe end of affirmative action in WashingtonState as a ldquonatural experimentrdquo to assess theimpact of affirmative action on minority stu-dentsrsquo transition from high school to collegeOn November 3 1998 voters in WashingtonState ended affirmative action by passing astatewide measure Initiative 200 (I-200) by amargin of 58 percent to 42 percent The ini-tiative was worded ldquoShall government beprohibited from discriminating or grantingpreferential treatment based on race sexcolor ethnicity or national origin in publicemployment education and contractingrdquo

The presumption of the sponsors and sup-porters of I-200 was that affirmative actionpolicies allowed less qualified minority stu-dents to enter universities at the expense ofmajority students The opponents of I-200feared that without affirmative action theeducational progress of minority studentswould be slowed (for more background onthe politics and aftermath of I-200 see Fryer2002) The assumption that I-200 would havean automatic impact was reinforced withwidespread press reports of a decrease inAfrican American Latino and American

Indian freshmen at the University ofWashington (UW) in 1999 following the pas-sage of I-200 (Davila and Mayo 2002)

The decrease in minority enrollments inWashington State is not an isolated case InCalifornia minority applications and enroll-ments dropped dramatically at many cam-puses of the University of California systemfollowing rulings by the universityrsquos Board ofRegents and a proposition passed by statevoters to eliminate race-conscious admissionsdecisions (Lomibao Barreto and Pachon2004 University of California 2003) Thedeclines were the steepest at the flagshipcampuses of Berkeley and Los Angeles whichhave also been among the slowest to recoverFollowing the 1996 Hopwood v Texas caseand the decision of the state of Texas to banaffirmative action in college admissionsenrollments of African American and Hispanicfreshman declined at the University of Texasat Austin (Tienda and Niu 2004) There hasbeen a partial recovery in minority enroll-ments in Texas with the implementation of aplan guaranteeing admission to the top 10percent of graduating high school classes(Horn and Flores 2003) Similar declines inminority enrollments have been reported inFlorida (Marin and Lee 2003) Georgia (Goseand Schmidt 2001) and Michigan (Dobbs2004 Winter 2004) as affirmative action pro-grams have come under challenge and havebeen modified or eliminated

The number and composition of collegefreshmen from year to year are shaped how-ever by demographic and economic forces aswell as by public policy The number of newstudents who enter colleges and universitiesdepends on the supply of students (the num-ber and composition of potential applicants)as well as institutional decisions about whomto admit and studentsrsquo acceptances of offersof admission The appropriate index to evalu-ate the impact of I-200 (or other policies) oncollege attendance is the ldquotransition raterdquo orthe percentage of in-state high school seniorswho enroll in Washington State colleges anduniversities The first objective of our analysisis to estimate transition rates by race and eth-nicity and to compare the 1999 rates after thepassage of I-200 with rates for earlier andlater years We also examine transition rates

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

108 Brown and Hirschman

for different types of four-year colleges in thestate

The transition from high school to collegeis a sequential process of decision making inwhich admission to a university is a criticalstep but only one stage in the processStudents (with advice and encouragementfrom others) take the first step by applying tocolleges and then make the final decision bydeciding whether and where to enrollassuming that they have been admitted toone or more colleges Through a detailedanalysis of administrative records from UWwe discovered that the decline in minoritystudents in 1999 was partially affected bychanges in ethnic-specific admission ratesbut primarily the result of declines in collegeapplications from prospective minority stu-dents

The decline in applications from minoritystudents included many students who werealmost certain to have been admitted on uni-versalistic standards of grades and test scoresWe interpret these findings as suggestive ofan unexpected impact of affirmative actionpolicies namely as a symbolic beacon of awelcoming environment Minority studentsfrom modest backgrounds may be reluctantto apply to the most competitive anddemanding colleges and universities as maymany majority students Universitiesrsquo affirma-tive action policies are often accompanied byoutreach efforts that provide informationpersonal contacts and encouragement forminority students to apply Even without aformal policy of affirmative action these out-reach effects may have spurred the reboundin minority application rates to UW in recentyears

AFFIRMATIVE ACTION AS A POLICY RESPONSE

For the past three decades affirmative actionprograms have been a bulwark of govern-mental efforts to redress racial ethnic andgender inequality in the United States(Bowen and Bok 1998 Holzer and Neumark2000 Reskin 1998) Nonetheless the effectsof affirmative action efforts remain uncertain

and controversial The problem of assessingthe effects of affirmative action arises not onlybecause programs are so varied but becauseit is difficult to measure discriminationmdashthephenomenon that affirmative action isdesigned to prevent or remedy A fewdecades ago racial and gender discrimina-tion was openly practiced (even in newspa-persrsquo want ads) and often reinforced by statelaws In such circumstances it was reasonableto assume that almost all intergroup inequal-ity was caused by discrimination eitherdirectly or indirectly Thus the impact ofefforts to combat discrimination would beevident in any observable changes in segre-gation and inequality

At present overt or formal discriminationhas diminished although informal practicesand everyday expressions of prejudice haveproved remarkably resistant to change(Lelyveld 2001) Discrimination that is subtleand hidden (and perhaps even unconsciousby the actor) is much more difficult to detectand its consequences are much more uncer-tain Over the past 40 years social scienceresearchers have expended considerableeffort and ingenuity trying to measure thechanges in the relative role of discriminationas a cause of racial and ethnic gaps in educa-tion occupations income and access to sub-urban residence The standard method ofestimation is to assume that discrimination isreflected in the net racial or ethnic differencethat remains after multivariate statistical mod-els are used to adjust for all measurable socialbackground characteristics that are unrelatedto discrimination (Duncan 1969 Feathermanand Hauser 1978)

This approach is limited however becauseracial and ethnic inequality is affected notonly by contemporary discrimination but bythe legacy of historical patterns of discrimina-tion that have produced unequal resources(eg family background characteristics)between groups The complex and often elu-sive relationships between historical patternsof discrimination and contemporary inter-group inequality have influenced the originsof affirmative-action programs and have ledto the political vulnerability of these pro-grams

In addition to the removal of discriminato-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 109

ry barriers that are sanctioned by law or cus-tom affirmative action programs can bearrayed along a continuum from ldquoopportuni-ty enhancementrdquo to mandating outcomessuch as proportional representation (Boboand Kluegel 1993) The most basic type ofaffirmative action consists simply of efforts toensure that minorities and women are includ-ed in the pool of eligible candidates throughoutreach programs and a reliance on formalmeans of recruitment rather than on infor-mal networks The second type of affirmativeaction program is one that gives qualifiedminority candidates extra (but not exclusive)consideration For example if several quali-fied candidates were being considered for aposition (or admission) affirmative actionwould give priority to a minority or womancandidate much as geographic diversity oran alumnus parent may give an advantage toothers in a college admission decision Thethird type of affirmative action program andthe one that creates the most political con-troversy is a program that admits or recruitsminorities or women for targeted slots orpositions without consideration of other can-didates

In the famous Bakke case (Regents of theUniversity of California v Bakke) the SupremeCourt justices split over the question of whetherracial quotas could be used to make up for pastracism The deciding vote was cast by JusticeLewis Powell who argued against racial quotasbut suggested that admissions officers in uni-versities could ldquotake race into accountrdquo as oneof several factors in evaluating minority candi-dates (Bowen and Bok 19988) This significantdistinction has allowed colleges to maintainaffirmative action programs as long as they dis-avowed the use of quotas This rather inchoatepolicy was continued with the mixed SupremeCourt decisions in 2003 involving the Universityof Michigan In one case the court rejected6ndash3 the use of a system that provided extrapoints to minorities in undergraduate admis-sions decisions But in Grutter v Bollinger thecourt voted 5ndash4 to allow raceethnicity to con-tinue to be considered as one of several criteriafor admissions to law schools Justice SandraDay OrsquoConnor writing for the majority saidthat affirmative action was still needed(Ogletree 2004243)

At the national level it is probably impos-sible to disentangle the impact of affirmativeaction programs from broader patterns thathave expanded opportunities for minoritystudents to enroll in colleges and universitiesFrom the 1950s through the 1970s the pro-portion of African Americans native-bornHispanics and American Indians with at leastsome college education doubled but theracial and ethnic differences remained widebecause white students were also makingimpressive educational gains (Mare1995173) Although most minority studentswere still handicapped by their disadvan-taged social backgrounds there was a briefwindow in the 1970s when African Americanhigh school graduates were slightly more like-ly to enter college than were whites with thesame socioeconomic and geographic charac-teristics (Jaynes and Williams 1989339)Although affirmative action programs mayhave played an important role in attractingminority students to highly selective institu-tions many other factors including improve-ments in minority household incomesreduced discrimination and rising ambitionswere also important The tenor of the timesduring the 1960s and 1970s was one ofgreater opportunities for able high school stu-dents including those from disadvantagedbackgrounds to aspire to a college educa-tion

For the next two decades college-enroll-ment rates resembled a roller coaster By the1980s the rates of college attendance andgraduation were declining among all groupsespecially disadvantaged minorities (Mare1995) The most plausible reason seems tohave been the decrease in financial aid grantsand the reluctance of students from poorhouseholds to borrow money for college(Jaynes and Williams 1989340ndash345) In the1990s college enrollments began to rise forall groups including minority groupsalthough wide differences by race and ethnic-ity persisted By the fall of 1998 45 percentand 52 percent of 20ndash21-year-old white non-Hispanic men and women respectively wereenrolled in college (Martinez and Curry1999) The comparable figures were 33 per-cent and 37 percent for African American(non-Hispanic) men and women and 22 per-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

110 Brown and Hirschman

cent and 25 percent for Hispanic men andwomen Not all minorities are disadvantagedin terms of educational enrollment howeverThe comparable percentages of Asian andPacific Islander 20 and 21 year olds who wereenrolled in college were 65 percent and 73percent for men and women respectively Infact native-born Asian Americans have hadhigher educational attainments than whitesfor many decades even before World War II(Folger and Nam 1967 Hirschman and Wong1986 Mare 1995)

Affirmative action programs have thepotential to affect both the number of minor-ity students who go on to college and thesestudentsrsquo choice of college These two issueshave typically been conflated in research onaffirmative action because the focus has gen-erally been limited to one college or at mosta few similar colleges Changes in enrollmentat other colleges have rarely been comparedIn the study presented here which investigat-ed the impact of I-200 we examined changesin enrollment for the universe of four-year col-leges and universities in Washington State

If the impact of affirmative action is solelya result of collegesrsquo admission decisions onlya small proportion of (potential) students areaffected by these decisions About 70 percentto 80 percent of colleges and universities inthe United States are nonselective that isthey admit almost all applicants who meetbasic entrance requirements (Bowen and Bok199815) Most of the attention and researchon affirmative action has focused on the rela-tively few colleges and universities that havecompetitive admissions and are assumed tooffer a superior education

Graduates of selective colleges with high-er tuition and average SAT scores havereported above-average earnings (Kane1998 Solmon and Wachtel 1973) and highlevels of job satisfaction and civic involve-ment (Bowen and Bok 1998) and are morelikely to obtain advanced degrees than aregraduates of other types of schools (Wegnerand Sewell 1970) These associations how-ever do not prove that attendance at selec-tive colleges actually causes these outcomesRecent research has suggested that the char-acteristics of students prior to their entryinto selective colleges such as socioeco-

nomic background and motivations mayaccount for many of the positive effects thatare usually attributed to attendance at selec-tive colleges (Dale and Krueger 1999Pascarella and Terrenzini 1991) Since thelabor market probably values many of thesame traits as do college admissions officersthe correlation between entry into competi-tive colleges and labor market success maybe inflated

Affirmative action programs may influencethe college choice of minority students forseveral reasons beyond giving an edge inadmissions decisions Perhaps most impor-tant is financial aid because students from dis-advantaged backgrounds may be reluctant orunable to assume debt Also these studentsmay have different perceptions of academicprograms and campus social life (Chapman1984 Hanson and Litten 1982 HosslerSchmit and Vesper 1999 Manski and Wise1983 McDonough 1997)

Assuming that students especially minori-ty students are looking at potential collegeswith the question of whether ldquostudents likethemselvesrdquo are present public policies thateliminate or modify affirmative action pro-grams may be presumed to affect perceptionsof the relative attractiveness of different col-leges and universities The prospect of fewerldquoin-grouprdquo classmates or a general sense thatminorities are no longer welcome maychange the perceived climate of campusesthat have ended affirmative action policies(Gorov 1997) More broadly changes in affir-mative action policy could influence howhigh school counselors advise students andhow universities organize recruitment efforts(Traub 1999) In the immediate aftermath ofthe ban on affirmative action some minoritystudents especially those with multiple choic-es may enroll in out-of-state colleges espe-cially historically black ones In fact such aspike in out-of-state enrollments occurredamong Washington residents according toour analysis of data from the US Departmentof Educationrsquos National Center for EducationStatistics (httpncesedgovipedsdataasp)In the fall of 1998 just before WashingtonState banned affirmative action 76 studentsfrom Washington enrolled as freshmen in his-torically black colleges In 2000 that enroll-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 3: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

108 Brown and Hirschman

for different types of four-year colleges in thestate

The transition from high school to collegeis a sequential process of decision making inwhich admission to a university is a criticalstep but only one stage in the processStudents (with advice and encouragementfrom others) take the first step by applying tocolleges and then make the final decision bydeciding whether and where to enrollassuming that they have been admitted toone or more colleges Through a detailedanalysis of administrative records from UWwe discovered that the decline in minoritystudents in 1999 was partially affected bychanges in ethnic-specific admission ratesbut primarily the result of declines in collegeapplications from prospective minority stu-dents

The decline in applications from minoritystudents included many students who werealmost certain to have been admitted on uni-versalistic standards of grades and test scoresWe interpret these findings as suggestive ofan unexpected impact of affirmative actionpolicies namely as a symbolic beacon of awelcoming environment Minority studentsfrom modest backgrounds may be reluctantto apply to the most competitive anddemanding colleges and universities as maymany majority students Universitiesrsquo affirma-tive action policies are often accompanied byoutreach efforts that provide informationpersonal contacts and encouragement forminority students to apply Even without aformal policy of affirmative action these out-reach effects may have spurred the reboundin minority application rates to UW in recentyears

AFFIRMATIVE ACTION AS A POLICY RESPONSE

For the past three decades affirmative actionprograms have been a bulwark of govern-mental efforts to redress racial ethnic andgender inequality in the United States(Bowen and Bok 1998 Holzer and Neumark2000 Reskin 1998) Nonetheless the effectsof affirmative action efforts remain uncertain

and controversial The problem of assessingthe effects of affirmative action arises not onlybecause programs are so varied but becauseit is difficult to measure discriminationmdashthephenomenon that affirmative action isdesigned to prevent or remedy A fewdecades ago racial and gender discrimina-tion was openly practiced (even in newspa-persrsquo want ads) and often reinforced by statelaws In such circumstances it was reasonableto assume that almost all intergroup inequal-ity was caused by discrimination eitherdirectly or indirectly Thus the impact ofefforts to combat discrimination would beevident in any observable changes in segre-gation and inequality

At present overt or formal discriminationhas diminished although informal practicesand everyday expressions of prejudice haveproved remarkably resistant to change(Lelyveld 2001) Discrimination that is subtleand hidden (and perhaps even unconsciousby the actor) is much more difficult to detectand its consequences are much more uncer-tain Over the past 40 years social scienceresearchers have expended considerableeffort and ingenuity trying to measure thechanges in the relative role of discriminationas a cause of racial and ethnic gaps in educa-tion occupations income and access to sub-urban residence The standard method ofestimation is to assume that discrimination isreflected in the net racial or ethnic differencethat remains after multivariate statistical mod-els are used to adjust for all measurable socialbackground characteristics that are unrelatedto discrimination (Duncan 1969 Feathermanand Hauser 1978)

This approach is limited however becauseracial and ethnic inequality is affected notonly by contemporary discrimination but bythe legacy of historical patterns of discrimina-tion that have produced unequal resources(eg family background characteristics)between groups The complex and often elu-sive relationships between historical patternsof discrimination and contemporary inter-group inequality have influenced the originsof affirmative-action programs and have ledto the political vulnerability of these pro-grams

In addition to the removal of discriminato-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 109

ry barriers that are sanctioned by law or cus-tom affirmative action programs can bearrayed along a continuum from ldquoopportuni-ty enhancementrdquo to mandating outcomessuch as proportional representation (Boboand Kluegel 1993) The most basic type ofaffirmative action consists simply of efforts toensure that minorities and women are includ-ed in the pool of eligible candidates throughoutreach programs and a reliance on formalmeans of recruitment rather than on infor-mal networks The second type of affirmativeaction program is one that gives qualifiedminority candidates extra (but not exclusive)consideration For example if several quali-fied candidates were being considered for aposition (or admission) affirmative actionwould give priority to a minority or womancandidate much as geographic diversity oran alumnus parent may give an advantage toothers in a college admission decision Thethird type of affirmative action program andthe one that creates the most political con-troversy is a program that admits or recruitsminorities or women for targeted slots orpositions without consideration of other can-didates

In the famous Bakke case (Regents of theUniversity of California v Bakke) the SupremeCourt justices split over the question of whetherracial quotas could be used to make up for pastracism The deciding vote was cast by JusticeLewis Powell who argued against racial quotasbut suggested that admissions officers in uni-versities could ldquotake race into accountrdquo as oneof several factors in evaluating minority candi-dates (Bowen and Bok 19988) This significantdistinction has allowed colleges to maintainaffirmative action programs as long as they dis-avowed the use of quotas This rather inchoatepolicy was continued with the mixed SupremeCourt decisions in 2003 involving the Universityof Michigan In one case the court rejected6ndash3 the use of a system that provided extrapoints to minorities in undergraduate admis-sions decisions But in Grutter v Bollinger thecourt voted 5ndash4 to allow raceethnicity to con-tinue to be considered as one of several criteriafor admissions to law schools Justice SandraDay OrsquoConnor writing for the majority saidthat affirmative action was still needed(Ogletree 2004243)

At the national level it is probably impos-sible to disentangle the impact of affirmativeaction programs from broader patterns thathave expanded opportunities for minoritystudents to enroll in colleges and universitiesFrom the 1950s through the 1970s the pro-portion of African Americans native-bornHispanics and American Indians with at leastsome college education doubled but theracial and ethnic differences remained widebecause white students were also makingimpressive educational gains (Mare1995173) Although most minority studentswere still handicapped by their disadvan-taged social backgrounds there was a briefwindow in the 1970s when African Americanhigh school graduates were slightly more like-ly to enter college than were whites with thesame socioeconomic and geographic charac-teristics (Jaynes and Williams 1989339)Although affirmative action programs mayhave played an important role in attractingminority students to highly selective institu-tions many other factors including improve-ments in minority household incomesreduced discrimination and rising ambitionswere also important The tenor of the timesduring the 1960s and 1970s was one ofgreater opportunities for able high school stu-dents including those from disadvantagedbackgrounds to aspire to a college educa-tion

For the next two decades college-enroll-ment rates resembled a roller coaster By the1980s the rates of college attendance andgraduation were declining among all groupsespecially disadvantaged minorities (Mare1995) The most plausible reason seems tohave been the decrease in financial aid grantsand the reluctance of students from poorhouseholds to borrow money for college(Jaynes and Williams 1989340ndash345) In the1990s college enrollments began to rise forall groups including minority groupsalthough wide differences by race and ethnic-ity persisted By the fall of 1998 45 percentand 52 percent of 20ndash21-year-old white non-Hispanic men and women respectively wereenrolled in college (Martinez and Curry1999) The comparable figures were 33 per-cent and 37 percent for African American(non-Hispanic) men and women and 22 per-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

110 Brown and Hirschman

cent and 25 percent for Hispanic men andwomen Not all minorities are disadvantagedin terms of educational enrollment howeverThe comparable percentages of Asian andPacific Islander 20 and 21 year olds who wereenrolled in college were 65 percent and 73percent for men and women respectively Infact native-born Asian Americans have hadhigher educational attainments than whitesfor many decades even before World War II(Folger and Nam 1967 Hirschman and Wong1986 Mare 1995)

Affirmative action programs have thepotential to affect both the number of minor-ity students who go on to college and thesestudentsrsquo choice of college These two issueshave typically been conflated in research onaffirmative action because the focus has gen-erally been limited to one college or at mosta few similar colleges Changes in enrollmentat other colleges have rarely been comparedIn the study presented here which investigat-ed the impact of I-200 we examined changesin enrollment for the universe of four-year col-leges and universities in Washington State

If the impact of affirmative action is solelya result of collegesrsquo admission decisions onlya small proportion of (potential) students areaffected by these decisions About 70 percentto 80 percent of colleges and universities inthe United States are nonselective that isthey admit almost all applicants who meetbasic entrance requirements (Bowen and Bok199815) Most of the attention and researchon affirmative action has focused on the rela-tively few colleges and universities that havecompetitive admissions and are assumed tooffer a superior education

Graduates of selective colleges with high-er tuition and average SAT scores havereported above-average earnings (Kane1998 Solmon and Wachtel 1973) and highlevels of job satisfaction and civic involve-ment (Bowen and Bok 1998) and are morelikely to obtain advanced degrees than aregraduates of other types of schools (Wegnerand Sewell 1970) These associations how-ever do not prove that attendance at selec-tive colleges actually causes these outcomesRecent research has suggested that the char-acteristics of students prior to their entryinto selective colleges such as socioeco-

nomic background and motivations mayaccount for many of the positive effects thatare usually attributed to attendance at selec-tive colleges (Dale and Krueger 1999Pascarella and Terrenzini 1991) Since thelabor market probably values many of thesame traits as do college admissions officersthe correlation between entry into competi-tive colleges and labor market success maybe inflated

Affirmative action programs may influencethe college choice of minority students forseveral reasons beyond giving an edge inadmissions decisions Perhaps most impor-tant is financial aid because students from dis-advantaged backgrounds may be reluctant orunable to assume debt Also these studentsmay have different perceptions of academicprograms and campus social life (Chapman1984 Hanson and Litten 1982 HosslerSchmit and Vesper 1999 Manski and Wise1983 McDonough 1997)

Assuming that students especially minori-ty students are looking at potential collegeswith the question of whether ldquostudents likethemselvesrdquo are present public policies thateliminate or modify affirmative action pro-grams may be presumed to affect perceptionsof the relative attractiveness of different col-leges and universities The prospect of fewerldquoin-grouprdquo classmates or a general sense thatminorities are no longer welcome maychange the perceived climate of campusesthat have ended affirmative action policies(Gorov 1997) More broadly changes in affir-mative action policy could influence howhigh school counselors advise students andhow universities organize recruitment efforts(Traub 1999) In the immediate aftermath ofthe ban on affirmative action some minoritystudents especially those with multiple choic-es may enroll in out-of-state colleges espe-cially historically black ones In fact such aspike in out-of-state enrollments occurredamong Washington residents according toour analysis of data from the US Departmentof Educationrsquos National Center for EducationStatistics (httpncesedgovipedsdataasp)In the fall of 1998 just before WashingtonState banned affirmative action 76 studentsfrom Washington enrolled as freshmen in his-torically black colleges In 2000 that enroll-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 4: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 109

ry barriers that are sanctioned by law or cus-tom affirmative action programs can bearrayed along a continuum from ldquoopportuni-ty enhancementrdquo to mandating outcomessuch as proportional representation (Boboand Kluegel 1993) The most basic type ofaffirmative action consists simply of efforts toensure that minorities and women are includ-ed in the pool of eligible candidates throughoutreach programs and a reliance on formalmeans of recruitment rather than on infor-mal networks The second type of affirmativeaction program is one that gives qualifiedminority candidates extra (but not exclusive)consideration For example if several quali-fied candidates were being considered for aposition (or admission) affirmative actionwould give priority to a minority or womancandidate much as geographic diversity oran alumnus parent may give an advantage toothers in a college admission decision Thethird type of affirmative action program andthe one that creates the most political con-troversy is a program that admits or recruitsminorities or women for targeted slots orpositions without consideration of other can-didates

In the famous Bakke case (Regents of theUniversity of California v Bakke) the SupremeCourt justices split over the question of whetherracial quotas could be used to make up for pastracism The deciding vote was cast by JusticeLewis Powell who argued against racial quotasbut suggested that admissions officers in uni-versities could ldquotake race into accountrdquo as oneof several factors in evaluating minority candi-dates (Bowen and Bok 19988) This significantdistinction has allowed colleges to maintainaffirmative action programs as long as they dis-avowed the use of quotas This rather inchoatepolicy was continued with the mixed SupremeCourt decisions in 2003 involving the Universityof Michigan In one case the court rejected6ndash3 the use of a system that provided extrapoints to minorities in undergraduate admis-sions decisions But in Grutter v Bollinger thecourt voted 5ndash4 to allow raceethnicity to con-tinue to be considered as one of several criteriafor admissions to law schools Justice SandraDay OrsquoConnor writing for the majority saidthat affirmative action was still needed(Ogletree 2004243)

At the national level it is probably impos-sible to disentangle the impact of affirmativeaction programs from broader patterns thathave expanded opportunities for minoritystudents to enroll in colleges and universitiesFrom the 1950s through the 1970s the pro-portion of African Americans native-bornHispanics and American Indians with at leastsome college education doubled but theracial and ethnic differences remained widebecause white students were also makingimpressive educational gains (Mare1995173) Although most minority studentswere still handicapped by their disadvan-taged social backgrounds there was a briefwindow in the 1970s when African Americanhigh school graduates were slightly more like-ly to enter college than were whites with thesame socioeconomic and geographic charac-teristics (Jaynes and Williams 1989339)Although affirmative action programs mayhave played an important role in attractingminority students to highly selective institu-tions many other factors including improve-ments in minority household incomesreduced discrimination and rising ambitionswere also important The tenor of the timesduring the 1960s and 1970s was one ofgreater opportunities for able high school stu-dents including those from disadvantagedbackgrounds to aspire to a college educa-tion

For the next two decades college-enroll-ment rates resembled a roller coaster By the1980s the rates of college attendance andgraduation were declining among all groupsespecially disadvantaged minorities (Mare1995) The most plausible reason seems tohave been the decrease in financial aid grantsand the reluctance of students from poorhouseholds to borrow money for college(Jaynes and Williams 1989340ndash345) In the1990s college enrollments began to rise forall groups including minority groupsalthough wide differences by race and ethnic-ity persisted By the fall of 1998 45 percentand 52 percent of 20ndash21-year-old white non-Hispanic men and women respectively wereenrolled in college (Martinez and Curry1999) The comparable figures were 33 per-cent and 37 percent for African American(non-Hispanic) men and women and 22 per-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

110 Brown and Hirschman

cent and 25 percent for Hispanic men andwomen Not all minorities are disadvantagedin terms of educational enrollment howeverThe comparable percentages of Asian andPacific Islander 20 and 21 year olds who wereenrolled in college were 65 percent and 73percent for men and women respectively Infact native-born Asian Americans have hadhigher educational attainments than whitesfor many decades even before World War II(Folger and Nam 1967 Hirschman and Wong1986 Mare 1995)

Affirmative action programs have thepotential to affect both the number of minor-ity students who go on to college and thesestudentsrsquo choice of college These two issueshave typically been conflated in research onaffirmative action because the focus has gen-erally been limited to one college or at mosta few similar colleges Changes in enrollmentat other colleges have rarely been comparedIn the study presented here which investigat-ed the impact of I-200 we examined changesin enrollment for the universe of four-year col-leges and universities in Washington State

If the impact of affirmative action is solelya result of collegesrsquo admission decisions onlya small proportion of (potential) students areaffected by these decisions About 70 percentto 80 percent of colleges and universities inthe United States are nonselective that isthey admit almost all applicants who meetbasic entrance requirements (Bowen and Bok199815) Most of the attention and researchon affirmative action has focused on the rela-tively few colleges and universities that havecompetitive admissions and are assumed tooffer a superior education

Graduates of selective colleges with high-er tuition and average SAT scores havereported above-average earnings (Kane1998 Solmon and Wachtel 1973) and highlevels of job satisfaction and civic involve-ment (Bowen and Bok 1998) and are morelikely to obtain advanced degrees than aregraduates of other types of schools (Wegnerand Sewell 1970) These associations how-ever do not prove that attendance at selec-tive colleges actually causes these outcomesRecent research has suggested that the char-acteristics of students prior to their entryinto selective colleges such as socioeco-

nomic background and motivations mayaccount for many of the positive effects thatare usually attributed to attendance at selec-tive colleges (Dale and Krueger 1999Pascarella and Terrenzini 1991) Since thelabor market probably values many of thesame traits as do college admissions officersthe correlation between entry into competi-tive colleges and labor market success maybe inflated

Affirmative action programs may influencethe college choice of minority students forseveral reasons beyond giving an edge inadmissions decisions Perhaps most impor-tant is financial aid because students from dis-advantaged backgrounds may be reluctant orunable to assume debt Also these studentsmay have different perceptions of academicprograms and campus social life (Chapman1984 Hanson and Litten 1982 HosslerSchmit and Vesper 1999 Manski and Wise1983 McDonough 1997)

Assuming that students especially minori-ty students are looking at potential collegeswith the question of whether ldquostudents likethemselvesrdquo are present public policies thateliminate or modify affirmative action pro-grams may be presumed to affect perceptionsof the relative attractiveness of different col-leges and universities The prospect of fewerldquoin-grouprdquo classmates or a general sense thatminorities are no longer welcome maychange the perceived climate of campusesthat have ended affirmative action policies(Gorov 1997) More broadly changes in affir-mative action policy could influence howhigh school counselors advise students andhow universities organize recruitment efforts(Traub 1999) In the immediate aftermath ofthe ban on affirmative action some minoritystudents especially those with multiple choic-es may enroll in out-of-state colleges espe-cially historically black ones In fact such aspike in out-of-state enrollments occurredamong Washington residents according toour analysis of data from the US Departmentof Educationrsquos National Center for EducationStatistics (httpncesedgovipedsdataasp)In the fall of 1998 just before WashingtonState banned affirmative action 76 studentsfrom Washington enrolled as freshmen in his-torically black colleges In 2000 that enroll-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 5: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

110 Brown and Hirschman

cent and 25 percent for Hispanic men andwomen Not all minorities are disadvantagedin terms of educational enrollment howeverThe comparable percentages of Asian andPacific Islander 20 and 21 year olds who wereenrolled in college were 65 percent and 73percent for men and women respectively Infact native-born Asian Americans have hadhigher educational attainments than whitesfor many decades even before World War II(Folger and Nam 1967 Hirschman and Wong1986 Mare 1995)

Affirmative action programs have thepotential to affect both the number of minor-ity students who go on to college and thesestudentsrsquo choice of college These two issueshave typically been conflated in research onaffirmative action because the focus has gen-erally been limited to one college or at mosta few similar colleges Changes in enrollmentat other colleges have rarely been comparedIn the study presented here which investigat-ed the impact of I-200 we examined changesin enrollment for the universe of four-year col-leges and universities in Washington State

If the impact of affirmative action is solelya result of collegesrsquo admission decisions onlya small proportion of (potential) students areaffected by these decisions About 70 percentto 80 percent of colleges and universities inthe United States are nonselective that isthey admit almost all applicants who meetbasic entrance requirements (Bowen and Bok199815) Most of the attention and researchon affirmative action has focused on the rela-tively few colleges and universities that havecompetitive admissions and are assumed tooffer a superior education

Graduates of selective colleges with high-er tuition and average SAT scores havereported above-average earnings (Kane1998 Solmon and Wachtel 1973) and highlevels of job satisfaction and civic involve-ment (Bowen and Bok 1998) and are morelikely to obtain advanced degrees than aregraduates of other types of schools (Wegnerand Sewell 1970) These associations how-ever do not prove that attendance at selec-tive colleges actually causes these outcomesRecent research has suggested that the char-acteristics of students prior to their entryinto selective colleges such as socioeco-

nomic background and motivations mayaccount for many of the positive effects thatare usually attributed to attendance at selec-tive colleges (Dale and Krueger 1999Pascarella and Terrenzini 1991) Since thelabor market probably values many of thesame traits as do college admissions officersthe correlation between entry into competi-tive colleges and labor market success maybe inflated

Affirmative action programs may influencethe college choice of minority students forseveral reasons beyond giving an edge inadmissions decisions Perhaps most impor-tant is financial aid because students from dis-advantaged backgrounds may be reluctant orunable to assume debt Also these studentsmay have different perceptions of academicprograms and campus social life (Chapman1984 Hanson and Litten 1982 HosslerSchmit and Vesper 1999 Manski and Wise1983 McDonough 1997)

Assuming that students especially minori-ty students are looking at potential collegeswith the question of whether ldquostudents likethemselvesrdquo are present public policies thateliminate or modify affirmative action pro-grams may be presumed to affect perceptionsof the relative attractiveness of different col-leges and universities The prospect of fewerldquoin-grouprdquo classmates or a general sense thatminorities are no longer welcome maychange the perceived climate of campusesthat have ended affirmative action policies(Gorov 1997) More broadly changes in affir-mative action policy could influence howhigh school counselors advise students andhow universities organize recruitment efforts(Traub 1999) In the immediate aftermath ofthe ban on affirmative action some minoritystudents especially those with multiple choic-es may enroll in out-of-state colleges espe-cially historically black ones In fact such aspike in out-of-state enrollments occurredamong Washington residents according toour analysis of data from the US Departmentof Educationrsquos National Center for EducationStatistics (httpncesedgovipedsdataasp)In the fall of 1998 just before WashingtonState banned affirmative action 76 studentsfrom Washington enrolled as freshmen in his-torically black colleges In 2000 that enroll-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 6: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 111

ment rose to 86 and then dropped to 62 in2001 and 72 in 2002

THE TRANSITION FROM HIGHSCHOOL TO COLLEGE

In our study the transition from high schoolto college was indexed by the ratio of collegefreshmen in four-year Washington State col-leges and universities to the number ofWashington State high school seniors in thepreceding year Although this measure issubject to some leakage from high schooldropouts (the measure of high school enroll-ment is taken in October of the senior year)and interstate mobility for higher educationthe index provides a consistent and reliablemeasure of the trend in the transition to col-lege over time Data on high school enroll-ments are collected and published by theWashington State Office of the Superinten-dent of Public Instruction (OSPI) and theUS Department of Educationrsquos IntegratedPostsecondary Education Data System(IPEDS) provides comparable enrollmentdata for students in all colleges and universi-ties We extracted the relevant data from theOSPI (httpwwwk12wausdataadmin)and IPEDS (httpncesedgovipeds) websites

Both the high school and college dataseries are based on administrative reportsfrom schools or colleges on the number ofstudents who are enrolled by year in schooland by raceethnicity and gender Althoughthere are slight inconsistencies in measure-ment in the race and ethnic categories themajor limitation is the lack of annual informa-tion on interstate mobility for college stu-dents The IPEDS form collects data on stateof origin and state of college enrollment ineven-numbered years (1996 1998 and soon) but only on a separate reporting formthat does not include race and ethnicity Thepercentages of Washington State natives whoattended out-of-state colleges and the reverseflowmdashout-of-state students who attendedWashington State collegesmdashwere roughlycomparable in the 1990s but there has beena modest increase in the outflow in recent

years1 We assume that these year-to-yearfluctuations in interstate mobility have notbiased our estimates of the transition to col-lege by race and ethnicity

The second limitation of the IPEDS data isthe increasing proportion of students who donot report their raceethnicity (or write insomething different from the standard cate-gories) on their college application formsAbout 5 percent of all first-year students whoenrolled at four-year Washington collegesand universities in 1999 were classified ashaving an ldquounknown or nonreportedrdquo racethere were virtually no missing data on raceand ethnicity for high school seniors In thefollowing analysis we estimate the transitionfrom high school to college for (reported)white students alone and for the sum ofldquowhitesrdquo and ldquounknown racerdquo together Evenif nonresponse to the race question were ran-dom the composition of the populationwould ensure that most of those with anunknown race were white Our expectationis however that whites are more likely not toreport their race because of their fear ofreverse discrimination

Our primary focus in this study is on stu-dents who begin their postsecondary studiesas freshmen in four-year colleges and univer-sities although about half the students whogo on to higher education in WashingtonState enroll in community colleges or othertwo-year institutions Many of these commu-nity college students are academically orient-ed and eventually transfer to four-year institu-tions It is difficult however to sort out theldquocollege-boundrdquo community-college stu-dents from the larger number who are simplytaking a class or two to upgrade their voca-tional skills or for self-fulfillment For exampleless than a quarter of first-time community-college students were enrolled in high schoolduring the previous year

We simplified the process by examiningjust the transition from high school senior tobeing a freshman in a four-year college oruniversitymdashwhich yields a consistent thoughincomplete picture of the overall processThere are three primary divisions of four-yearcolleges and universities in Washington StateThe state has two major research universities(UW and Washington State University WSU)

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 7: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

112 Brown and Hirschman

four additional four-year public universitiesand colleges (Western Washington CentralWashington Eastern Washington andEvergreen) and a number of private colleges(including Seattle University University ofPuget Sound Pacific Lutheran Gonzaga andWhitman)

CHANGES IN THE TRANSITIONRATE

The number of seniors who were enrolled inpublic and private high schools inWashington State rose from about 68000 inthe late 1990s to about 76000 in 2003About 4 out of 5 high school seniors inWashington State were non-Hispanic whitesbut the overall proportion of minority stu-dents increased by a couple of percentagepoints in recent years At the end of the1990s about 8 percent of seniors in the statewere Asian and Pacific Islanders a little lessthan 7 percent were Hispanic 4 percent wereAfrican American and 2 percent wereAmerican Indian

In the late 1990s about 16000 freshmenwere enrolled in Washington State collegesand universities which is about one fourththe number of high school seniors in the pre-vious year (assuming that approximately thesame number of students leave and enter thestate for higher education) This figure is anunderestimate of the proportion of highschool seniors who will eventually attendfour-year colleges because it ignores the siz-able number of community college studentswho transfer to four-year colleges Figure 1shows the trend in the transition from highschool senior to college freshman by race andethnicity from 1994 to 2002

Figure 1 shows three distinct racial andethnic patterns in the transition from highschool to college A little more than 1 in 5white Washington high school seniors go onto college the following year (with atten-dance rising by a couple of percentage pointsif the ldquorace unknown and otherrdquo students areconsidered white) Attendance is the highestfor Asian American studentsmdashabout 1 in 3high school seniors enroll in college the fol-

lowing year The third racialethnic pattern isthe below-average transition rate from highschool senior to college freshman for the tra-ditionally underrepresented minoritiesAfrican Americans American Indians andHispanics Only about 1 in 7 of these minori-ty high school seniors attend four-year col-leges in Washington the following year

An observable though relatively small andtemporary decline in the ldquotransition raterdquofrom high school to college appeared as of1999mdashthe year following the passage of I-200mdashfor all groups except whites (andwhites and othersunknown) A moredetailed examination is presented in Table 1which shows the number of college freshmenand transition rates from high school to col-lege by race and ethnicity and type of highereducation

The number of freshmen in four-year col-leges and universities in Washington State(based on IPEDS data) is presented byraceethnicity and by year in the left-handpanels of Table 1 for the average of the mid-1990s (1994ndash97) for the individual yearsimmediately before and after the passage of I-200 and for the average of 2001ndash02 In theright-hand panels are the transition ratesmdashthe proportions of high school students whoattended college the following year Beloweach panel are measures of relative changeindexed as the percentage change for eachinterval The appropriate index for assessingthe impact of I-200 is not the change in theabsolute number of college freshmen but thechange in the rate of transition from highschool to college These are the results shownin the right-hand columns of Table 1 with thetransition rates indexed by the ratio of collegefreshmen to high school seniors in the pre-ceding year for each racialethnic group

The absolute number of college freshmenand the transition rates are shown for all four-year colleges and for the three major subdivi-sions of higher education the research uni-versities (UW and WSU) other public four-year colleges and private four-year collegesAll high school students are administrativelyclassified by raceethnicity (probably onforms completed by parents teachers orschool administrators) however studentswho apply to college may write in any race or

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 8: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 113

ethnicity or leave the item blank As we notedearlier we assumed that most college stu-dents with an ldquoother or unknownrdquo raceeth-nicity are probably white and the last panelin Table 1 shows the combined white (non-Hispanic) plus ldquoothersrdquo student population

Examining the change (the absolute num-ber of freshmen and the rate of transitionfrom high school to college) from 1998 to1999 in the context of a longer span of yearsmay suggest whether I-200 had an effectabove and beyond what may be expectedfrom typical year-to-year variation The pat-tern for white non-Hispanic students (or theldquowhite plus otherrdquo students) establishes thebenchmark by which the other groups will bemeasured The freshmen college enrollmentfigures include all students (in state and outof state) while the denominators include onlyin-state high school seniors Fluctuations inthe relative number of students who go outof state for college compared to those fromother states who come to colleges and uni-versities in Washington may account for

some ldquonormalrdquo year-to-year variation in thetransition rates

The trend in freshman college enrollmentsamong white students has been remarkablysteady The absolute number of white collegefreshmen grew from about 11000 in themid-1990s to almost 13000 in 2003 and thetransition rate hovered at about 21 percent to22 percent with a slight overall decrease in2000 and in research universities in 1999 Ifwe assume that the combined white andldquoother and unknownrdquo raceethnicity groupbetter reflects the actual experience of whitestudents then their transition rate from highschool senior to college freshman is about 23percent to 25 percent

As was reported in the news media at thetime there were measurable declines in thenumber of minority freshman in 1999 rela-tive to 1998 (as well as earlier and later years)These declines were most evident at publiccolleges especially the research universitiesFor example the number of African Americanfreshman at the two research universities fell

Figure 1 The Transition from High School to College The Ratio of College Freshmen to HighSchool Seniors in the Previous Year by RaceEthnicity Washington State 1994ndash2002

Source OSPI and IPEDS see the text

Note The values for Hispanics in 1994 and 1995 are not available

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 9: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

114 Brown and Hirschman

Table 1 Trends in Freshman Enrollment and Transition Rates by Race and Ethnicity and Typeof College in Washington State 1994ndash2002

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Non-Hispanic White1994ndash97 Average 11070 4483 3137 3451 022 009 006 0071998 11723 4667 3435 3621 022 009 006 0071999 11944 4397 3820 3727 022 008 007 0072000 12656 4686 4164 3806 022 008 007 0072001ndash02 Average 12961 4742 4199 4020 023 008 007 007Percentage Change

From 1994ndash97 to 1998 59 41 95 49 -29 -45 05 -37From 1998 to 1999 19 -58 112 29 10 -66 103 21From 1999 to 2000 60 66 90 21 21 26 50 -16From 2000 to 2001ndash02 24 12 08 56 21 09 05 53

Black1994ndash97 Average 336 171 88 77 013 007 003 003

1998 386 208 100 78 014 007 004 0031999 336 154 92 90 012 005 003 0032000 406 208 101 97 013 007 003 0032001ndash02 Average 446 225 117 105 015 007 004 003Percentage Change

From 1994ndash97 to 1998 151 218 143 10 35 96 28 -92From 1998 to 1999 -130 -260 -80 154 -145 -273 -96 134From 1999 to 2000 208 351 98 78 103 233 02 -16From 2000 to 2001ndash02 99 82 153 77 116 99 172 95

Hispanic1996ndash97 Average 545 257 143 146 015 007 004 0041998 611 301 156 154 014 007 004 0041999 551 187 181 183 012 004 004 0042000 549 201 167 181 012 005 004 0042001ndash02 Average 754 285 225 244 015 006 005 005Percentage Change

From 1996ndash97 to 1998 121 173 95 55 -30 15 -53 -88From 1998 to 1999 -98 -379 160 188 -136 -405 112 139From 1999 to 2000 -04 75 -77 -11 -09 70 -82 -16From 2000 to 2001ndash02 373 418 347 348 227 267 204 205

American Indian1994ndash97 Average 248 92 86 70 018 007 006 0051998 265 108 106 51 017 007 007 0031999 215 70 102 43 014 005 007 0032000 222 79 86 57 014 005 005 0042001ndash02 Average 253 88 109 57 016 005 007 004Percentage Change

From 1994ndash97 to 1998 67 174 229 -271 -48 47 96 -350From 1998 to 1999 -189 -352 -38 -157 -162 -330 -06 -129From 1999 to 2000 33 129 -157 326 -39 50 -216 233From 2000 to 2001ndash02 137 108 262 -09 147 117 273 00

continued

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 10: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 115

from 208 in 1998 to 154 in 1999 a decline ofmore than 25 percent The transition rateswhich ldquoadjustrdquo for changes in the ldquosupplyrdquo ofstudents (high school seniors the yearbefore) also show significant declines in theproportions of minority students (includingAsian and Pacific Islanders) who went on tocollege in 1999

In absolute terms minority transition ratesdeclined by 2 to 3 percentage points from1998 to 1999 In relative terms the changesin transition rates represent double-digit per-centage declines of 14 percent to 16 percentin one year (less for Asian Americans)Changes of this magnitude stand out fromthe normal year-to-year variation Yet formost minorities the transition rates bouncedback to their ldquonormalrdquo levels in 2000 or short-ly thereafter The depressed transition ratesappear to have been only temporary

In addition to the unmistakable decline in

the likelihood of minority high school seniorsgoing on to college in Washington Statethere are important variations by type of col-lege in 1999 The significant decline in thetransition rates from 1998 to 1999 for theunderrepresented minorities appears to havebeen limited to the research universitiesMore detailed data (not shown here) indicatethat the largest declines in minority studentenrollments occurred at UW There were alsodeclines in minority enrollments at WSU butthis was part of an overall decline in enroll-ments of all racial and ethnic groups at WSUin 1999mdashperhaps because of the upswing inenrollments in 1998 following WSUrsquos appear-ance in the Rose Bowl

Although the ratio of IPEDS college fresh-man to OSPI high school seniors is not anideal measure of the transition from highschool to college the potential biases in theindex are not specific to one ethnic group

Table 1 Continued

Transition Rate Ratio of College Freshmen Number of Freshmen to Washington High School Seniors

All 4-Year Research Other Public Private All 4-Year Research Other Public PrivateRace and Ethnicity Colleges Universities Colleges Colleges Colleges Universities Colleges Colleges

Asian and Pacific Islander1994ndash97 Average 1704 1100 235 370 037 024 005 0081998 1854 1209 220 425 036 023 004 0081999 1887 1212 294 381 033 022 005 0072000 2148 1413 298 437 036 024 005 0072001ndash02 Average 2267 1518 321 429 038 025 005 007Percentage Change

From 1994ndash97 to 1998 88 99 -62 149 -21 -11 -156 34From 1998 to 1999 18 02 336 -104 -69 -83 223 -180From 1999 to 2000 138 166 14 147 90 116 -30 98From 2000 to 2001ndash02 55 74 76 -19 34 52 53 -40

Non-Hispanic White and Other1994ndash97 Average 11709 4685 3358 3667 023 009 007 0071998 12944 5261 3792 3891 024 010 007 0071999 13120 5133 4006 3981 024 009 007 0072000 13723 5260 4378 4085 024 009 008 0072001ndash02 Average 14239 5397 4478 4364 025 009 008 008Percentage Change

From 1994ndash97 to 1998 105 123 129 61 14 30 36 -27From 1998 to 1999 14 -24 56 23 05 -33 47 14From 1999 to 2000 46 30 91 26 07 -13 53 -12From 2000 to 2001ndash02 38 26 23 68 34 23 19 65

Source Office of the Superintendent for Public Instruction Washington State and the US Department ofEducations Integrated Postsecondary Educational Data System

Note The number of freshman students includes both in-state and out-of-state students The figures for theyear after I-200 are in bold face

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 11: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

116 Brown and Hirschman

nor are they likely to have changed signifi-cantly over the span of years representedhere Therefore we interpret the decline inminority students in 1999 following the pas-sage of I-200 as the impact of the ending ofaffirmative action The declines in 1999 weremuch larger than the ldquonormalrdquo year-to-yearvariation were limited to minority studentsand were found only at the research universi-ties Fortunately we have access to data onapplications and admission of students to UWfrom 1998 to 2003 which allows for a moredetailed examination of the impact of I-200on enrollment in the major research universi-ty in the state

A CLOSER LOOK AT UW

UW is the dominant educational institution inthe state both quantitatively and qualitative-ly With almost 5000 freshmen in 2000 itenrolled almost one third of the first-year stu-dents in four-year colleges in the state Inaddition to its reputation as the flagship insti-tution of the public educational system UW isalso considered to be the most prestigiousinstitution for undergraduate education inthe state Because of UWrsquos unique position inthe higher educational system of the state aswell as the finding that the impact of I-200 onminority enrollments is primarily limited toresearch universities we examine the trend inminority enrollments at UW before and afterthe implementation of I-200

Overall American Indians Hispanics andAfrican Americans constituted 126 percent ofhigh school seniors in Washington during the1997ndash98 academic year but only 82 percentof the entering in-state UW freshmen in thefall of 1998 In 1999 following the passage ofI-200 there was no change in the ethniccomposition of high school seniors but theshare of underrepresented minorities fell to57 percent of the UW freshmen class Itappears that I-200 had an impact but skep-tics may reasonably ask for an analysis thatconsiders alternative explanations as well as adetailed account of how I-200 affected fresh-men enrollments Our analysismdashbased on theUW admissions and records databasemdashcon-tains information on applications admissions

and the enrollment of freshmen and providesa window on the sequence of steps from highschool senior to college freshman

APPLICATION ADMISSION ANDMATRICULATION

The transition from high school senior to col-lege freshman begins with a studentrsquos deci-sion to apply to college The second step isthe institutional decision by the college toadmit or reject the studentrsquos application Thethird step is the studentrsquos enrollment (matric-ulation) as a college freshman These stepsare diagrammed for in-state students at UWin Figure 2 The first and most consequentialstep is the application ratemdashfewer than 12percent of seniors in Washington State highschools apply for admission to UW Becauseof the vastness of the potential pool of out-of-state applicants this model (and the analysis)is limited to in-state students

Of the applicant pool of 8614 high schoolseniors in 2000 UW admitted 855 percentThis UW admissions policy may be describedas mildly selective for in-state students amuch smaller proportion of out-of-state stu-dents is admitted The final step in Figure 2 isthe sum of the individual decisions amongthe successful applicants who decide which (ifany) college to attend About 55 percent ofthe admitted pool of 7365 students accept-ed the UW admission offer to the freshmanclass in 2000 The entering class of 4090 in-state UW freshmen in the fall of 2000 repre-sented about 57 percent of Washington highschool seniors in the previous year

This simple model illustrates that decisionsby students (and their families) can be muchmore consequential than those of institutionsThe standard model of affirmative action andthe debate over the policy focus almost exclu-sively on the institutional decisions of col-leges This image is probably shaped by theexamples drawn from the most selective col-leges and universities in the nation Moststate universities in the United States areprobably similar to UW in that they havemildly selective admission practices with themajority of qualified in-state students being

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 12: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 117

accepted In such circumstances changes inpublic policy hold considerable potential toinfluence the transition to college by affectingthe decisions of students as well as the deci-sions of institutions With this framework inmind we consider the potential reasons forthe decline in the transition from high schoolsenior to UW freshman in 1999 after the pas-sage of I-200

Table 2 elaborates the model in Figure 2for four years or periods (1998 1999 2000and the average of 2001ndash03) by racialethnicgroups The first panel shows the transitionrates from high school senior to UW appli-cant the middle panel shows the UW admis-sion rates and the final panel shows theenrollment rates (the actual number of highschool seniors applicants admitted fresh-man and enrolled freshman are shown inAppendix Table A) There are slight differ-ences in the racial and ethnic classification ofhigh school seniors (from OSPI) and the UWracial and ethnic classification Asians andPacific Islanders are combined in the OSPI

data but are separate categories in the UWdatabase Two additional categories appear inthe UW database otherunknown (race notreported) and international (the number ofin-state international students are too few tobe reported here)

Students are not required to report theirraceethnicity on their college applications anda growing number of students feel that it is notin their interest to do so Those who did notreport raceethnicity rose from 117 percent ofthe applicants in 1998 to 153 percent in 1999but then declined after UW specifically askedthese students to supply a race or ethnicity Insome of the analyses reported later we com-bined whites and othersunknowns

About 12 percent of Washington Statehigh school seniors apply to UW but thisoverall average is a composite of varied racialand ethnic patterns A high proportion of thestatersquos Asian and Pacific Islander high schoolseniors (almost all are Asian Americans)mdashabout 27 percent to 30 percentmdashapply foradmission to UW A little more than 9 percent

Figure 2 The Transition from High School to College Freshmen The Winnowing of 72003Washington State High School Seniors to 4090 In-State Freshmen at the University ofWashington

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 13: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

118 Brown and Hirschman

Table 2 The Transition from Washington High School Senior to College Applicant fromApplicant to Admitted Freshman and from Admitted to Enrolled Freshman University ofWashington 1998 to 2003 (percentage)

From High School Senior to UW ApplicantPercentage of Seniors Who Apply

RaceEthnicity 1998 1999 2000 2001ndash03

White 95 92 91 91White and OtherUnknowna 113 115 112 109Black 84 67 73 83Hispanic 70 60 53 62American Indian 64 56 51 53Asian AmericanPacific Islanderb 303 270 284 304Total 122 121 120 119

From Applicant to Admitted FreshmanUW Admission Rate

RaceEthnicity 1998 1999 2000 2001ndash03

White 793 821 856 801Black 837 704 833 772Hispanic 867 801 828 809American Indian 820 765 904 885Asian Americanb 823 855 878 838Pacific Islanderb 973 833 857 785OtherUnknowna 789 842 833 781Total 802 826 855 807

From Admitted to Enrolled FreshmanPercentage Who Enroll

RaceEthnicity 1998 1999 2000 2001ndash03

White 504 503 522 525Black 482 511 577 541Hispanic 651 526 534 596American Indian 537 538 573 557Asian Americanb 729 725 724 710Pacific Islanderb 500 500 750 643OtherUnknowna 521 541 452 436Total 554 551 555 558

Sources Office of the Superintendent for Public Instruction Washington State and UW admis-sions and records database

a Assuming that almost all college applicants with raceethnicity not reported are whiteb The OSPI high school data combine Asian American and Pacific Islanders in one category but

the UW data have separate categories here the categories are combined for application but sep-arated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 14: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 119

of white high school seniors apply to UW butthis proportion could be about 11 percent ifwe assume that almost all those who did notreport their race and ethnicity (theothersunknown category) are white

Underrepresented minority high schoolseniors (African Americans Hispanics andAmerican Indians) are less likely to apply toUW than are white high school seniors Thedifferences are fairly modest in absoluteterms but are consequential in affecting thecomposition of college freshmen as we shallshow Most important there was a drop-off inthe percentage of minority high school stu-dents who applied to UW in 1999 relative to1998 The application rate for AfricanAmerican students rose in 2000 and by2001ndash03 had nearly returned to the 1998rate of a little more than 8 percent The appli-cation rates for Hispanics and AmericanIndians declined in 1999 and again in 2000but rebounded modestly in 2001ndash03Although Asian Americans had the highestlevel of application to UW they also experi-enced a dip in their application rate in 1999These patterns suggest a ldquodiscouragementrdquoeffect for all applications by minority studentsafter the passage of I-200

The second step of the process is the univer-sity admissions ratemdashindexed here as the per-centage of applicants who are admitted Theoverall UW admission rate of freshman appli-cants is determined by a host of factors includ-ing the number and qualifications of in-stateand out-of-state applicants the absorptivecapacity of the institution and the expectedldquoyieldrdquo (the percentage of admitted studentswho will matriculate) For the years representedhere the UW admission rate for in-state stu-dents was about 80 percent to 86 percent withonly modest variations by race and ethnicity

In 1998 when race and ethnicity could betaken into account the admission rates of theunderrepresented minorities were slightlyhigher (82 percent for American Indians 87percent for Hispanics and 84 percent forAfrican Americans) than for whites (79 per-cent) In 1999 (after I-200) there was adecline of about 14 percentage points in theadmission rates of Washington State AfricanAmerican students (from 84 percent to 70percent) and smaller declines for the other

underrepresented minorities In the sameyear there were modest rises in the admis-sions rates for whites and Asians In 2000there was little variation in admission rates byrace and ethnicity There have been modestdecreases in the admission rates for all groupssince 2000 but little in the way of clear pat-terns The admissions rates of theotherunknown applicants are parallel tothose of the white applicants

The final step in the process is matricula-tionmdashthe percentage of admitted applicantswho actually enroll Since many studentsapply to multiple colleges and universitiesthe ldquoyieldrdquo of enrolled freshmen to admis-sions depends on many factors that cannotbe measured here At UW a little more thanhalf the in-state students (55 percent to 56percent) who are admitted become UWfreshmen Since 1999 the rates for theunderrepresented groups in WashingtonState have been fairly close to those ofwhitesmdashperhaps a few percentage pointshigher The one group with a distinctive pat-tern is Asian Americans students whoseenrollment rate tops 70 percent of those whoare admitted

DECOMPOSITION OF THECHANGES IN TRANSITION RATE

Measuring the intervening steps in the transi-tion from high school senior to college fresh-man does not by itself show the relativeimportance of each of these processes Onemethod of estimating the relative impact ofthe three processes on the overall change inthe transition rate from high school senior toUW freshman is the demographic method ofthe ldquodecomposition of differences in ratesrdquo(Kitagawa 1955 Preston Heuveline andGuillot 2001) The standard decompositionformula is as follows

∆T = ∆AP ADmdash

ENmdash

+ ∆ADAPmdash

ENmdash

+ ∆ENAPmdash

ADmdash

The percentage point change in the transi-tion rate from high school senior to collegefreshman (∆T) equals the sum of three quan-tities

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 15: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

120 Brown and Hirschman

1 the change in the application rate (∆AP)times the average admission rate (AD

mdash )times the average enrollment rate (EN

mdash )2 the change in the admission rate (∆AD)

times the average application rate (APmdash )

times the average enrollment rate (ENmdash )

3 the change in the enrollment rate (∆EN)times the average application rate (AP

mdash )times the average admission rate (AD

mdash )where ∆T is the percentage-point change

in the transition rate from high school seniorto college student from t to t + 1 ∆AP is thepercentage-point change in the applicationrate from t to t + 1 ∆AD is the percentage-point change in the admission rate from t tot + 1 ∆EN is the percentage-point change inthe enrollment rate from t to t + 1 and AP

mdash ADmdash and EN

mdash are the average (mean) admis-sion application and enrollment rates for tand t + 1

In Table 3 we present the decomposition ofthe 1998 to 1999 1999 to 2000 and 2000 to2001ndash03 changes in rates (percentage-pointchanges) of Washington State high schoolseniors who enrolled as freshmen at UW Thetransition rates are the ratio of UW freshmen tohigh school seniors in the preceding year Thefirst panel shows the rates for each year 19981999 2000 and 2001ndash03 and then thechange in the rates from 1998 to 1999 from1999 to 2000 and from 2000 to 2001ndash03 (inthe top right-hand panel) As with the earliertables we computed a ldquowhite and otherrdquo pop-ulation which includes the sum of whites andunknownother race UW freshmen in thenumerator and OSPI white seniors in thedenominator

Overall about 1 in 20 high school seniorsin Washington State enters UW as a freshmanthe following year More precisely the pro-portion rose slightly from 54 percent in 1998to 55 percent in 1999 to 57 percent in 2000and then declined slightly to 54 percent in2001ndash03 About 17 percent to 18 percent ofAsian American high school seniors in thestate enter UW as freshmen Close to the stateaverage are whites (or whites and unknowns)with about 4 percent to 5 percent of seniorsenrolling as UW freshmen the following yearConsistently below the white level are AfricanAmericans Hispanics and American Indianswhose rates of transition from high school

senior to college freshman hover from slight-ly above 2 percent to slightly below 4 per-cent Since underrepresented minorities havehigher attrition during high school these fig-ures are underestimates of the interethnicgaps in attaining a higher education

From 1998 to 1999 the rates of transitionfrom high school senior to UW first-year stu-dent dropped for all groups except whitesAlthough these declines are only 10 percent-age point for African Americans 14 percent-age points for Hispanics 15 percent forAsians and Pacific Islanders and 05 percent-age point for American Indians these are sig-nificant changes from the previous year Forexample the change in the transition ratefrom high school senior to UW freshman of34 percent to 24 percent for AfricanAmericans represents a relative decline of 29percent For the two-year interval from 1998to 2000 it appears that the transition rate forAfrican American and Asian American stu-dents bounced back to the 1998 level butthere was a slower recovery for AmericanIndians and Hispanics Only for the mostrecent period 2001ndash03 are the transitionrates roughly comparable to the 1998 levels

In the lower panels of Table 3 we decom-pose the percentage-point change for eachgroup for the 1998 to 1999 interval (and the1999 to 2000 and the 2000 to 2001ndash03 inter-vals) into three transitions application ratesadmission rates and enrollment rates (basedon the transition rates presented in Table 2)The components of change can be positive ornegative2 The 98 percentage-point drop inthe African American transition rate from1998 to 1999 is the sum of three compo-nents -065 owing to the lower applicationrate -050 owing to the lower admission ratesby the university and a counterbalancingchange of +017 points because admittedAfrican American students were slightly morelikely to accept admission in 1999 than in1998 The decline in enrollment of Hispanicsof 139 percentage points was due partially todeclines in the application rate (-045) butalso to declines in rates of admission (-025)and enrollment (-068) Declines in applica-tion and admission rates also lowered theAmerican Indian rate of transition from highschool senior to UW first-year student

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 16: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 121Ta

ble

3D

eco

mp

osi

tio

no

fth

eC

han

ge

inth

eR

ate

of

Was

hin

gto

nSt

ate

Hig

hSc

ho

ol

Sen

iors

Enro

llin

gas

Fres

hm

enat

the

Un

iver

sity

of

Was

hin

gto

n(U

W)

by

Rac

ean

dEt

hn

icit

y19

98to

1999

19

99to

2000

an

d20

00to

2001

ndash03

Perc

enta

geof

Hig

hSc

hool

Seni

ors

Enro

lling

atU

WPe

rcen

tage

Poin

tC

hang

e

Aver

age

of19

9819

9920

00to

Race

Eth

nici

ty19

9819

9920

0020

01ndash0

3to

1999

to20

0020

01ndash0

3

Whi

te3

83

84

13

80

00

3-0

2W

hite

and

Oth

erU

nkno

wn

45

49

49

44

03

00

-04

Blac

k3

42

43

53

5-1

01

10

0H

isp

anic

39

25

23

30

-14

-02

07

Am

eric

anIn

dian

28

23

26

26

-05

03

00

Asi

anP

acifi

cIs

land

er18

116

718

018

0-1

51

40

0To

tal

54

55

57

54

01

02

-03

Dec

omp

ositi

onof

the

Rate

ofU

WEn

rollm

ent

ofH

igh

Scho

olSe

nior

s

1998

to19

99C

hang

eA

ttrib

utab

leto

1999

to20

00C

hang

eA

ttrib

utab

leto

2000

to20

01ndash0

3C

hang

eA

ttrib

utab

leto

Race

Eth

nici

tyAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Appl

icat

ion

Adm

issio

nEn

rollm

ent

Tota

lAp

plic

atio

nAd

miss

ion

Enro

llmen

tTo

tal

Whi

te-0

12

014

000

002

-00

50

160

140

260

00-0

26

002

-02

4W

hite

and

Oth

erU

nkno

wn

011

020

004

034

-01

40

15-0

02

-00

1-0

14

-03

00

01-0

43

Blac

k-0

65

-05

00

17-0

98

027

049

035

111

046

-02

6-0

22

-00

2H

isp

anic

-04

5-0

25

-06

8-1

39

-03

30

080

04-0

22

043

-00

60

290

66A

mer

ican

Indi

an-0

33

-01

80

01-0

50

-02

40

410

160

330

11-0

06

-00

8-0

02

Asi

anP

acifi

cIs

land

er-2

01

058

-00

3-1

45

083

044

008

135

123

-08

1-0

42

000

Tota

l-0

04

016

-00

20

09-0

07

019

004

016

-00

1-0

32

003

-03

0

Sour

ce

Tabl

e2

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 17: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

122 Brown and Hirschman

The decomposition of the overall change inthe transition rate from 1999 to 2000 and from2000 to 2001ndash03 shows an overall pattern ofrecovery From 1999 to 2000 the observeddeclines for the African American and Asian andPacific Islander populations that were registeredin 1999 had largely been erased The depressedrates for Hispanics and American Indians wereslower to recover although they did so in themost recent period

The decomposition exercise provides evi-dence that changes in the rates of applicationare an important part of the explanation forthe relative declines of minority enrollmentsafter the passage of I-200 At first glance itmay seem that the white application ratedecreased as well but if we can assume thatmost of those who did not report a race (theldquounknownsrdquo) are white there was no overallchange in the application rate of the majoritypopulation to UW from 1998 to 2000Increases in minority application rates werealso important components of the recoveryfrom 1999 to 2000 and from 2000 to2001ndash03

What are the practical implications ofthese findings To present this analysis to uni-versity administrators we computed theimplications of these results as the outcomesof three alternative strategies or scenarios thatmay be considered to increase the number ofunderrepresented minority students The firststrategy would be to increase the applicationrate of minority high school seniors to UW tothe average of all Washington State seniorsmdash120 percent in 2000 The second strategywould be to admit 100 percent of all studentswho apply The third strategy would be topersuade 100 percent of all accepted stu-dents to matriculate None of these strategiesis realistic in the sense that it could beachieved The goal of this ldquowhat ifrdquo exercise isto discover the potential of these threeldquoleversrdquo to increase the number of underrep-resented students (details are available fromus on request)

The first scenario which raises the applica-tion rates for these groups to the simple aver-age of all Washington seniors (120 percent)and with the other two processes the admis-sion rates and the enrollment rates remain-ing unchanged would result in a dramatic

increase in the number of minority freshmenThe number of African American freshmenwould increase by two thirds and the num-ber of Hispanic and American Indian studentswould more than double The second sce-nario to increase the admission rate to 100percent of all applicants would have the leastimpact in raising the number of minority stu-dents With admission rates for in-stateminority students about 70 percent to 80 per-cent changes in racial differences in admis-sion rates would scarcely affect the composi-tion of the first-year students at UW The thirdscenariomdashhaving all admitted studentsmatriculatemdashwould have a more substantialimpact on minority studentsrsquo enrollmentsapproximately doubling their numberbecause only half the admitted applicants(minorities and the majority) currently enrollAlthough this strategy would be efficacious inincreasing minority enrollments it is alsoprobably the most difficult to accomplish

Although minority admission rates haverecovered to their pre-1999 levels there arestill substantial majorityndashminority gaps in theoverall transition rates from high schoolseniors to UW freshmen Since admission andmatriculation rates vary only slightly by raceand ethnicity the primary explanation is dif-ferences in application rates For the mostrecent period (2001ndash03) minority applica-tion rates were 2 to 5 percentage pointsbelow the whiteunknown level Even a smallgrowth in the proportions of qualified minori-ties who apply is likely to translate to a sub-stantially larger number of minorities in thefirst-year university enrollments

HAS THE QUALITY OF THEAPPLICANT POOL CHANGED

The decomposition analysis identified declinesin minority college applications as the primarymechanism for the lower number of AfricanAmerican Hispanic and American Indian fresh-man at UW after the passage of I-200 Thisanalysis assumes that the processes of applica-tion admission and enrollment are indepen-dent of each other However studentsrsquo plansfor and applications to college are probably

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 18: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 123

affected by expectations about success Highschool students who are in doubt about theirqualifications for admission to selective collegesmay be less likely to apply to the most compet-itive colleges and universities

This logic suggests an alternative explana-tion for the observed change in minority appli-cation rates to UW following I-200 If the num-ber and composition of the pool of high schoolseniors changed during the late 1990s perhapsapplications went down because of an increas-ing fraction of ldquoless competitive studentsrdquoamong the potential pool of college applicantsIf these less qualified students had applied thenthe admission rate (admittedapplied) wouldhave declined with no change in the numberof enrolled minorities

One possible way to address this alternativeexplanation albeit indirectly is to examine thenumber and qualifications of the pool of appli-cants We compared the ldquoqualityrdquo of potentialapplicants by examining the scores of SAT tak-ers in Washington State (presumably highschool seniors) by race and ethnicity Therewere no significant changes in either the ratioof test takers to high school seniors or averagescores by race and ethnicity (detailed data avail-able on request)

A more direct picture of the changing com-position of UW applicants comes from the UWadmissions and records database The UWadmissions office constructs an ldquoadmissionindexrdquo score based on high school grade pointaverage and SAT scores for each applicant Thisindex which ranges from 0 to 99 is used tomake the first cut of admissions decisionsStudents who score above a certain level tendto be admitted if they have all the other pre-requisites and at the other end of the distribu-tion those with an admission index below acertain line are routinely denied admissionApplicants in the middle of the range undergomore scrutiny before an admission decision ismade Table 4 shows the absolute number andpercentage change of all UW applicants from1998 to 2001ndash03 and for those who were ratedhighly in the admissions index (66 and higher)

The overall declines in minority applicationrates from 1998 to 1999 were paralleled bydeclines in the number of highly rated appli-cants for the same years There were alsodecreases in the number of less qualified and

average rated applicants however the ldquodis-couragementrdquo that led to the lower rate ofapplication was not limited to students with lit-tle chance of being admitted As minority appli-cants increased in subsequent years the num-ber of top-rated applicants also increased

Our reading of this evidence is that most ofthe decline in minority applicants especiallyamong Washington State students was amongstudents who stood a good chance of beingadmitted to UW Many of these students werecompetitive at other colleges perhaps out ofstate and may have decided not to apply to UW

CONCLUSION

The contemporary debate over affirmativeaction is confusing to the public partly becauseof heated rhetoric as both proponents andopponents try to take the high moral groundThe goals of eliminating discrimination forginga color-blind society and maintaining the lega-cy of Martin Luther King Jr are invoked byadvocates for and against affirmative actionUnderlying the debate are real disagreementsover how to define discrimination and how toredress the unfairness and injuries caused by dis-crimination

Changes in public policies can sometimescreate opportunities for ldquonatural experimentsrdquothat allow researchers to measure the impact ofaffirmative action In the present case the endof affirmative action after the passage of I-200eliminated the ability of colleges and universi-ties in Washington State to take race intoaccount in admission and financial-aid deci-sions Although we do not know if present orpast patterns of discrimination had any influ-ence on college attendance by minorities inWashington State we can assume that therewere no dramatic changes in the conditionsthat affected the transition from high school tocollege in Washington from 1998 to 1999including any patterns of discrimination Thiswould leave changes in policy as the most like-ly cause of any significant change in collegeenrollment rates from 1998 to 1999

There was a substantial decrease in the num-ber of African American Hispanic andAmerican Indian freshmen (and percentage of

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 19: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

124 Brown and HirschmanTa

ble

4Pe

rcen

tag

eC

han

ge

of

In-S

tate

Fres

hm

anA

pp

lican

tsan

do

fH

igh

lyR

ated

Ap

plic

ants

toth

eU

niv

ersi

tyo

fW

ash

ing

ton

by

Rac

ean

dEt

hn

icit

y19

98to

2003

Race

and

Ethn

icity

Am

eria

nH

isp

anic

Asi

anPa

cific

Whi

teBl

ack

Indi

anA

mer

ican

Am

eric

anIs

land

erO

ther

Unk

now

nTo

tal

Num

ber

ofAp

plic

ants

1998

514

723

329

410

01

528

3797

58

322

1999

503

518

926

685

150

024

128

58

394

2000

516

622

723

383

164

228

121

28

614

2001

to20

03(a

vera

ge)

523

626

232

387

182

936

104

28

815

Perc

enta

geC

hang

e(A

llAp

plic

ants

)19

98to

1999

-22

-18

9-9

5-1

50

-18

-35

131

80

919

99to

2000

26

201

-12

4-2

49

516

7-5

72

620

00to

2001

ndash03

13

153

388

44

114

274

-14

02

3

Perc

enta

geC

hang

e(o

fHig

hly

Rate

dAp

plic

ants

)a19

98to

1999

-09

-18

8-0

8-3

47

-42

-38

935

82

219

99to

2000

66

338

-15

359

411

436

4-4

76

020

00to

2001

ndash03

27

-84

559

-11

111

413

3-1

63

20

Sour

ce

Uni

vers

ityof

Was

hing

ton

adm

issi

ons

and

reco

rds

data

base

a

Scor

ed66

orhi

gher

ona

scal

eof

0to

99

base

don

high

scho

olG

PAan

dth

eSA

T

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 20: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 125

high school graduates) who enrolled in the fallof 1999 This decline was almost entirely con-centrated at the two research universities UWand WSU The drop in first-year enrollments atWSU affected all groups including non-Hispanic whites perhaps because of the unusu-ally high 1998 enrollments following WSUrsquosparticipation in the Rose Bowl the year beforeThe decrease in the underrepresented minori-ties at UW however stood in contrast to thelack of change in first-year enrollments amongnon-Hispanic whites

If a substantial number of the minority stu-dents at UW in the 1990s were there solelybecause of preferential admissions policies thepassage of I-200 would have reduced theirnumber because more qualified majority stu-dents took their places However changes inthe rates of admission of applicants played asecondary role in the decline of minority stu-dents from 1998 to 1999 The rates of admis-sion of minority applicants were only marginal-ly lower in 1999 than in 1998

The decrease in minority enrollmentsoccurred in large part because of the drop inapplications by minority students We haveinterpreted the decline in the application rateto UW in 1999 and 2000 as a discouragementeffect following the passage of I-200 This inter-pretation is consistent with the broader litera-ture on campus climate Minority students whoare more socially integrated are less likely to feelalienated and drop out or transfer (Cabrera etal 1999 DrsquoAugelli and Hershberger 1993Hurtado and Carter 1997) Changes in the per-ceptions of campus climate are important fac-tors as high school seniors make decisionsabout college applications

An alternative interpretation is that thedecline in minority application rates was pri-marily among high school seniors who decidednot to apply because they thought that theywould be unlikely to be admitted in a race-blind admissions process However this inter-pretation is inconsistent with the stability in thenumber and average scores of WashingtonState minorities who took SAT examinations in1998 and 1999 Moreover the decrease inminority applicants was across the boardincluding many students whose high schoolgrades and SAT scores would have ensuredtheir admission to UW

If the publicity over I-200 discouraged someminority high school seniors from applying toUW why was this response not evident at otherstate universities and colleges or private col-leges in the state The prestige of UW its sizeand its reputation as a demanding (and per-haps impersonal) educational environmentmay have discouraged many students fromapplying Other colleges in the state were prob-ably seen as providing a less intimidating tran-sition to higher education Countervailingforces including the encouragement of familyfriends and teachers may have led many stu-dents to see UW as a positive challenge withgreater opportunities and the reward of a pres-tigious degree

Although minority high school seniors arenot alone in seeing UW as a potentially intimi-dating environment they may be more hesi-tant than majority students to take the leap toapply because they have fewer family membersand friends who had been undergraduates at amajor research university It seems reasonable toassume that affirmative action may have servedas one of the countervailing forces for minoritystudents who may have been apprehensiveabout the prospect of attending UWAffirmative action may have meant a personalcontact (by letter or telephone) or encourage-ment from a high school counselor that servedto overcome the studentsrsquo natural reluctance toapply to the most competitive university in thestate

When affirmative action was eliminated in1999 the fundamental conditions remainedthe same No new barriers were erected thatmade it more difficult to apply to UW and theadvantages and disadvantages remained asbefore What was missing however was a pol-icy that provided a welcoming and positive faceon the prospect of attending UW in spite of itssize and reputation With the loss of an impor-tant positive image created by affirmativeaction a small number of college-bound minor-ity high school seniors probably realigned theirset of attractive college destinations and left UWoff the list

The impact of ending affirmative action inWashington State on minority studentsrsquo appli-cations to UW appears to have been tempo-rary After one or two years of depressedapplication rates the number of freshman

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 21: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

126 Brown and Hirschman

applications of African American AmericanIndian and Hispanic students has risen orexceeded their prendashI-200 levels With anawareness of the problem UW mounted anoutreach program to high schools and com-munity groups to encourage more freshmanapplications from prospective minority stu-dents These efforts may have eased theimpact of I-200

The findings reported here for the state ofWashington may have relevance for manyother regions and universities in the UnitedStates In particular the finding that affirma-tive action programs can affect minorityapplication rates stands in contrast to thestandard interpretation that affirmative actionprograms are important only because of theirpresumed effect on admission rates This iscertainly the case for selective private andpublic colleges and universities At these insti-tutions only a small fraction of qualifiedapplicants are admitted and variations in eth-nic-specific admission rates can have a majorimpact on the racial and ethnic compositionof students

The UW experience may be more repre-

sentative of public higher education in moststates where there is only a moderate degreeof selectivity in admissions decisions In thesesettings the majority of qualified applicantsfrom every racial and ethnic group arealready admitted so the use of affirmativeaction criteria in admissions decisions hasonly a modest impact on the number ofminority students For these universitiesincreasing the application rate from minoritystudents may be a more effective strategySince application rates tend to be low (andminority application rates are even lower)even a small rise in the minority applicationrate can lead to a substantial increase inminority college enrollments with no changein admission rates Although affirmativeaction policies may be one means of encour-aging minority studentsrsquo applicants theremay be other strategies including outreachefforts and an inclusive public image that mayaccomplish the same end If the political tidescontinue to erode traditional affirmativeaction programs universities and collegesmay have to find more creative strategies forencouraging a diverse student body

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 22: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 127

Appendix Table A High School Seniors in Washington State and In-State Applicants Admissions and EnrolledFreshmen at the University of Washington (UW) by Race and Ethnicity 1998 to 2003

Washington State High School Seniors

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 54409 54875 56974 56888 57437 58994Black 2790 2840 3112 3074 3051 3302Hispanic 4226 4410 4432 4757 5162 5755American Indian 1561 1511 1624 1590 1630 1654Asian and Pacific Islandera 5157 5636 5888 5922 6103 6401Total 68143 69272 72030 72231 73383 76106

In-State Applicants to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 5147 5035 5166 4930 5480 5297Black 233 189 227 222 309 254Hispanic 294 266 233 288 308 374American Indian 100 85 83 91 94 75Asian Americana 1528 1500 1642 1659 1871 1958Pacific Islandera 37 24 28 29 33 45OtherUnknown 975 1285 1212 1592 895 640International 8 10 23 18 11 16Total 8322 8394 8614 8821 9001 8647

In-State Students Admitted to UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 4079 4136 4421 4092 4397 4100Black 195 133 189 178 243 185Hispanic 255 213 193 243 259 283American Indian 82 65 75 83 82 65Asian Americana 1258 1283 1441 1459 1531 1609Pacific Islandera 36 20 24 24 25 35OtherUnknown 769 1082 1010 1239 703 500International 3 4 15 13 8 7Total 6677 6936 7368 7331 7248 6784

In-State Freshman Enrolled at UW

RaceEthnicity 1998 1999 2000 2001 2002 2003

White 2056 2082 2309 2196 2203 2206Black 94 68 109 112 116 100Hispanic 166 112 103 162 138 168American Indian 44 35 43 43 47 38Asian Americana 917 930 1044 1034 1077 1152Pacific Islandera 18 10 18 18 11 25OtherUnknown 401 585 457 598 265 201International 0 0 7 7 4 1Total 3696 3822 4090 4170 3861 3891

Sources Office of the Superintendent for Public Instruction Washington State and the UWadmissions and records database

a The OSPI high school data report Asian American and Pacific Islanders in one category but the UW admissions and record database has separate categories The categories are combinedfor application but separated for admission and enrollment

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 23: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

128 Brown and Hirschman

NOTES

1 The percentages of students from out ofstate were 124 in 1992 145 in 1994 141in 1996 128 in 1998 167 in 2000 and140 in 2002 The percentages of WashingtonState high school students who went to out-of-state colleges were 143 in 1992 148 in1994 156 in 1996 175 in 1998 223 in2000 and 201 in 2002 (Knapp et al 2005)

2 The decomposition results are presentedto two significant digits to the right of thedecimal point while the summary percent-age-point changes are rounded to the nearesttenth of a percentage point in the top panels

REFERENCES

Bobo Lawrence and James R Kluegel 1993ldquoOpposition to Race-Targeting Self-InterestStratification Ideology or Racial AttitudesrdquoAmerican Sociological Review 58443ndash64

Bowen William G and Derek Bok 1998 The Shape ofthe River Long-Term Consequences of ConsideringRace in College and University AdmissionsPrinceton NJ Princeton University Press

Cabrera Alberto F Amaury Nora Patrick T TerenziniErnest Pascarella and Linda Serra Hagedorn1999 ldquoCampus Racial Climate and theAdjustment of Students to College AComparison Between White Students andAfrican-American Studentsrdquo Journal of HigherEducation 70134ndash60

Chapman David W 1984 ldquoA Model of StudentCollege Choicerdquo Journal of Higher Education52490ndash505

DrsquoAugelli Anthony R and Scott L Hershberger1993 ldquoAfrican American Undergraduates on aPredominantly White Campus AcademicFactors Social Networks and CampusClimaterdquo Journal of Negro Education 6267ndash81

Dale Stacy Berg and Alan B Krueger 1999ldquoEstimating the Payoff to Attending a MoreSelective College An Application of Selection onObservables and Unobservablesrdquo WorkingPaper No 7322 National Bureau of EconomicResearch Cambridge MA

Davila Florangela and Justin Mayo 2002November 24 ldquoI-200 Didnrsquot Erase Color onCampusrdquo Seattle Times pp 1 A18

Dobbs Michael 2004 November 22 ldquoUniversitiesRecord Drop in Black Admissionsrdquo WashingtonPost p A1

Duncan Otis Dudley 1969 ldquoInheritance of Poverty

or Inheritance of Racerdquo Pp 85ndash110 in OnUnderstanding Poverty Perspectives from theSocial Sciences edited by Daniel P MoynihanNew York Basic Books

Featherman David L and Robert M Hauser 1978Opportunity and Change New York AcademicPress

Folger John K and Charles B Nam 1967 Educationof the American Population A 1960 CensusMonograph Washington DC US GovernmentPrinting Office

Fryer Alex 2002 November 25 ldquoAffirmative ActionFight Shifts from Ballot Box to CourtroomrdquoSeattle Times pp 1 A12

Gorov Lynda 1997 December 23 ldquoAffirmativeAction Ban Alters Terrain in Californiardquo BostonGlobe p A1

Gose Ben and Peter Schmidt 2001 September 7ldquoRuling Against Affirmative Action Could AlterLegal Debate and Admissions PartnersrdquoChronicle of Higher Education Available on-lineat httpchroniclecomfreev48i0202a03601htm

Hanson Katherine H and Larry H Litten 1982ldquoMapping the Road to Academe A Review ofResearch on Women Men and the College-Selection Processrdquo Pp 73ndash97 in TheUndergraduate Woman Issues in EducationalEquity edited by Pamela J Perun LexingtonMA D C Heath

Hirschman Charles and Morrison Wong 1986 ldquoTheExtraordinary Educational Attainment of Asian-Americans A Search for Historical Evidence andExplanationsrdquo Social Forces 651ndash27

Holzer Harry and David Neumark 2000 ldquoAssessingAffirmative Actionrdquo Journal of EconomicLiterature 38483ndash568

Horn Catherine L and Stella M Flores 2003Percent Plans in College Admissions AComparative Analysis of Three Statesrsquo ExperiencesCambridge MA Civil Rights Project HarvardUniversity

Hossler Don Jack Schmit and Nick Vesper 1999Going to College How Social Economic andEducational Factors Influence the DecisionsStudents Make Baltimore MD Johns HopkinsUniversity Press

Hurtado Sylvia and Deborah Faye Carter 1997ldquoEffects of College Transition and Perceptions ofthe Campus Racial Climate on Latino CollegeStudentsrsquo Sense of Belongingrdquo Sociology ofEducation 70324ndash45

Jaynes Gerald and Robin M Williams Jr eds 1989A Common Destiny Blacks and American SocietyWashington DC National Academy Press

Kane Thomas 1998 ldquoRacial and Ethnic Preferencesin College Admissionrdquo Pp 431ndash56 in The Black-

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 24: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

Affirmative Action in Washington State 129

White Test Score Gap edited by ChristopherJencks and Meredith Phillips Washington DCBrookings Institution

Kitagawa Evelyn M 1955 ldquoComponents of aDifference Between Two Ratesrdquo Journal of theAmerican Statistical Association 501168ndash94

Knapp Laura G Janice E Kelly-Reid Roy WWhitmore Shiying Wu Seungho Huh BurtonLevine Marcus Berzofsky and Susan G Broyles2005 NCES Enrollment in PostsecondaryInstitutions Fall 2002 and Financial StatisticsFiscal Year 2002 (NCES 2005-168) WashingtonDC National Center for Education Statistics

Lelyveld Joseph ed 2001 How Race Is Lived inAmerica New York Times Books

Lomibao Stephanie Matt A Barreto and Harry PPachon 2004 ldquoThe Reality of Race-NeutralAdmissions for Minority Students at theUniversity of California Turning the Tide orTurning Them Awayrdquo Los Angeles TomaacutesRivera Policy Institute University of SouthernCalifornia

Manski Charles F and David A Wise 1983 CollegeChoice in America Cambridge MA HarvardUniversity Press

Mare Robert D 1995 ldquoChanges in EducationalAttainment and School Enrollmentrdquo Pp155ndash214 in State of the Union America in the1990s Volume 1 Economic Trends edited byReynolds Farley New York Russell SageFoundation

Marin Patricia and Edgar K Lee 2003Appearance and Reality in the Sunshine StateThe Talented 20 Program in FloridaCambridge MA The Civil Rights Project atHarvard University

Martinez Gladys and Andrea E Curry 1999ldquoSchool EnrollmentmdashSocial and EconomicCharacteristics of Students October 1998(Update) Current Population Reports (P20-521) Washington DC US Census BureauDetailed tables available on-line at httpwwwcensusgovpopulationsocdemoschooreport98tab01pdf

McDonough Patricia M 1997 Choosing Colleges

How Social Class and Schools StructureOpportunity Albany State University of NewYork Press

Ogletree Charles J 2004 All Deliberate SpeedReflections on the First Half-Century of Brown vBoard of Education New York W W Norton

Pascarella Ernest T and Patrick T Terenzini 1991How College Affects Students Findings andInsights from Twenty Years of Research SanFrancisco Jossey-Bass

Preston Samuel H Patrick Heuveline and MichelGuillot 2001 Demography Measuring andModeling Population Processes Malden MABlackwell

Reskin Barbara 1998 The Realities of AffirmativeAction in Employment Washington DCAmerican Sociological Association

Skrentny John David ed 2001 Color LinesAffirmative Action Immigration and Civil RightsOptions for America Chicago University ofChicago Press

Solmon Lewis C and Paul Wachtel 1973 ldquoEffectsof Income on Type of College AttendedrdquoSociology of Education 4875ndash90

Tienda Marta and Sunny Niu 2004 ldquoCapitalizingon Segregation Pretending NeutralityCollege Admissions and the Texas Top 10Lawrdquo Available on-line at httpwwwtexas-top10princetonedureportsforthcomingcapitalizing_on_segregationpdf

Traub James 1999 May 2 ldquoThe Class of Prop209rdquo New York Times Magazine p 44

University of California 2003 ldquoUnder-graduateAccess to the University of California After theElimination of Race-Conscious Policiesrdquo Officeof the President Student Academic ServicesAvailable on-line at httpwwwucopedusaspublishaa_final2pdf

Wegner Eldon L and William H Sewell 1970ldquoSelection and Context as Factors Affectingthe Probability of Graduation from CollegerdquoAmerican Journal of Sociology 75665ndash67

Winter Greg 2004 April 13 ldquoAfter Ruling 3Universities Maintain Diversity in AdmissionsrdquoNew York Times p A22

Susan K Brown PhD is Assistant Professor of Sociology at the University of California Irvine Hercurrent research examines immigrant incorporation in the United States She is a co-investigator ofImmigration and Intergenerational Mobility in Metropolitan Los Angeles a project studying theadult children of immigrants

Charles Hirschman PhD is Boeing International Professor in the Department of Sociology andthe Daniel J Evans School of Public Affairs at the University of Washington His research interestsare immigration and ethnicity in the United States and social change in Southeast Asia He cur-rently directs the University of WashingtonndashBeyond High School project a longitudinal study of edu-cational attainment and the early life course of young adults

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342

Page 25: The End of Affirmative Action in ashington State and W ... PUBS/A107.pdf · African American, Latino, and American Indian freshmen at the University of ashington (UW) in 1999, following

130 Brown and Hirschman

This article is a revised version of a paper that was presented at the annual meeting of the PopulationAssociation of America May 9ndash11 2002 in Atlanta Georgia The research was supported by a grantfrom the Andrew Mellon Foundation tiled ldquoThe Impact of Changes in Affirmative Action Policy on theTransition from High School to College in Washington Staterdquo We are grateful to Gerald Gillmore HughParker Tim Washburn and Philip Hoffman for their advice on University of Washington administrativedata and their helpful comments on the manuscript to Harriet Zuckerman who believed that this studycould be done before we did and to Richard McCormick for his inspired leadership We also thank PaulLePore Daphne Kuo Robert Plotnick John Robert Warren Richard Morrill Carolyn Liebler JenniferHolsinger Jennifer Lee Jason Thomas and Julie Miller for their collaboration on this project and com-ments on earlier versions of the manuscript We are solely responsible for any errors of data analysis orinterpretation Direct correspondence to Susan K Brown Department of Sociology University ofCalifornia Irvine 3151 Social Science Plaza Irvine CA 92697-5100 e-mail skbrownuciedu or toCharles Hirschman Department of Sociology University of Washington Box 353340 Seattle WA98195 e-mail charlesuwashingtonedu

Delivered by Ingenta to University of WashingtonFri 16 Jun 2006 175342