This article was downloaded by: [The University of Manchester Library]On: 10 October 2014, At: 11:07Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The Journal of Economic EducationPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/vece20

Gender Economics Courses inLiberal Arts CollegesNancy J. BurnettPublished online: 25 Mar 2010.

To cite this article: Nancy J. Burnett (1997) Gender Economics Courses in Liberal ArtsColleges, The Journal of Economic Education, 28:4, 369-376

To link to this article: http://dx.doi.org/10.1080/00220489709597940

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information(the “Content”) contained in the publications on our platform. However, Taylor& Francis, our agents, and our licensors make no representations or warrantieswhatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions andviews of the authors, and are not the views of or endorsed by Taylor & Francis. Theaccuracy of the Content should not be relied upon and should be independentlyverified with primary sources of information. Taylor and Francis shall not be liablefor any losses, actions, claims, proceedings, demands, costs, expenses, damages,and other liabilities whatsoever or howsoever caused arising directly or indirectly inconnection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden.Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Gender Economics Courses in Liberal Arts Colleges

Nancy J. Burnett

Gender has become a “hot” research topic in recent years and has begun mak- ing its way into the classroom (Conrad 1992). Interest in gender issues has spread, but only a small proportion of economics departments beyond the few top national liberal arts colleges include courses in gender economics.

This article presents a simultaneous probit model of gender-related economics courses as well as women’s studies programs in the undergraduate, liberal arts curriculum. Liberal arts colleges are often perceived to be in the forefront of undergraduate pedagogy. I restricted the study to these colleges to avoid, as much as possible, the complications created by graduate programs. Liberal arts col- leges are generally private and not subject to state mandates.

DATA

Data for the academic year 1992-93 were collected from 132 top-ranked lib- eral arts colleges. This ranking and the geographical designation of each col- lege are from “America’s Best Colleges” ( 1993), that provides rankings of lib- eral arts colleges as either national or regional. National colleges are rated strictly above the regional schools. My sample included the first quartile of each category.’ There were 35 national colleges and 97 regional colleges in this sample. Of these 132 schools, 44 were in the northern region of the United States, 36 in the Midwest, 32 in the South, and 20 in the West. Summary sta- tistics describing the data are shown in Table I . (Raw data are available from the author upon request.)

Academic ranking of each institution was determined by “America’s Best Col- leges” (1993), with the rank of 1 as best. The top quartile of national liberal arts colleges in the sample had rankings from 1 (a tie between Williams and Amherst) to 44 (also a tie, between Union and the University of the South). Regional col- leges were of strictly lower academic standing than the national schools and were ranked independently by region, so the top academically ranked schools in each of the four geographic regions had the same ranking.

From the 1992-93 catalogue of each institution, information was gathered on the presence of a gender-related economics courses. There were 3 1 gender eco-

Nancy J. Burneit is an u.r.si.Uutu privt,.ssor of ecofioniics o r the Ufi iwnity of Wi.scon.sin. The author would like to thank the urionymous rcferees and a pcrrticulrrly helpjitl editor at this .jourtial as well ( i s 7i,m Goodwmiri and Eleanor Brown far their helpful commeiitx and Catheririe Brick ,for research as,sistarice.

Fall 1997 369

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

TABLE 1 Data Summary Statistics

Variable National Regional

Overall colleges colleges _____

GENDERECON ,2348 (.4239)

WOMSTUDIES ,4394 (.4963)

ACADEMICREP 119.0984 (62.1290)

(.2179) WOM ANECONFAC ,247 I

ECONFAC

WOMANFAC

NOR

sou

MID

RELIG

6.7500 (3.7627)

,3579 (. 1554)

.3257 (.4686)

,2424 (.4286) .2803

(.44915) ,5757

(.4942)

~

,7143 (.4583)

.9429 (.2355)

18.6857 (12.5435)

.2201 (. 1 138)

10.3429 (3.827)

,3245 (.0967)

,6857 (.47 10)

,0857 I (.2840)

. I 143 (.3228)

. I 143 (.3228)

.06 19 (.2409)

,2577 (.4374)

155.3298 (19,7094)

,2568 (.24 10)

5.4536 (2.7800)

,3699 (. 1706)

,1959 (.3969)

,2989 (.4578)

,3402 (.4738) .7423

(.4374)

Note: Means with standard deviations in parentheses.

nomics courses in the sample. A few of these courses treated gender issues as somewhat of a sideline. For example, Trinity College in Hartford, Connecticut, offered a course, Poverty in America, that dealt with women’s issues. That course description read “. . . within each of these topics special attention will be given to, women’s experience” (Trinity College 1992-93, 144, emphasis added). Although many economics courses may refer to gender issues, I included only those courses that devoted “special attention” to the topic.

There were 61 women’s studies programs (or majors) in the sample (“Ameri- ca’s Best Colleges” 1993). The presence of a woman’s studies program was cross-checked with NWSA Directory of Women’s Studies ‘Programs, Women s Centers, and Women’s Research Centers (1990). Institutions were interviewed by telephone whenever uncertainties arose.

Information regarding the economics* departments was also garnered from the college catalogues. Only full-time, regular- appointment faculty members (assis- tant professor and above) were in~luded.~ Any uncertainty regarding the status or gender of an individual was checked by telephone.

Data on the gender composition of the entire faculty of each surveyed college were found in Academe (1993). For consistency, only full-time faculty members of assistant professor rank or above were included in the sample. It is a com-

370 JOURNAL OF ECONOMIC EDUCATION

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

monly accepted truism that women are more often found in the adjunct (instruc- tor or lecturer) positions than are men. Eliminating these ranks restricts the sam- ple to tenure-track (or tenured) position^.^ Another reason for restricting the sam- ple in this way was to target an institution’s commitment to female faculty. Schools not reporting in Academe were telephoned for information regarding the 1992-93 academic year.

Religious affiliation of each institution was determined by using The World Almanac and Book of Facts (1993), and these data were cross-checked with “America’s Best Colleges” (1993).

MODEL

Economic theory suggests that all products are produced in response to supply and demand. In education, supply is in the hands of the faculty and administra- tion. It may be driven by the institution’s mission, an individual faculty member’s interests, tradition, or by generally accepted practicess Demand for courses aris- es from student and/or parental concerns. Judging student demand, a priori, is a difficult task because students cannot directly reveal their demand for a course until it is offered.h Judging demand is made doubly difficult because there may be a considerable lag between the expression of demand and the creation of a course. The simultaneous supply model presented here relies on both adminis- trative and departmental variables.

The presence of a gender-related economics course should be affected by the size and gender composition of the economics department. The larger the depart- ment, the larger the course variety would tend to be, all else being equal. Gender composition of the department should also matter given that women may be more likely to teach gender-related courses.

Academic ranking of the college may also affect the probability that gender- related coursework is offered, either in a women’s studies program or in the eco- nomics department. The higher-ranking institutions may be more aware of the importance of diversity and multiculturalism (as it applies to women’s issues). They also may be more overtly competitive and try to offer courses that reflect the newest developments in academics.

The presence of a women’s studies program or major reflects both institu- tional commitment to the study of gender and, to some extent, the demand fac- tors calling for gender coursework. The presence of a woman’s studies program or major should reduce the transactions costs involved in creating a gender eco- nomics course, so colleges with a women’s studies program might be more like- ly to offer a gender-related economics course.

General attitudes toward women and women’s issues may also be affected by geographic region or the religious affiliation of the institution. Colleges that maintain religious affiliations may be more bound by traditional theories of edu- cation, and may, therefore, tend toward the more traditional fields of study.

In the model, I attempt to explain, or predict, the presence of gender-related courses, either in the economics department or in the general college curriculum, with a two-equation system:

Fall 1997 371

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

GENDERECON = AACADEMICREP, WOMSTUDIES, (1)

WOMSTUDIES = AACADEMICREP, WOMANFAC, RELIC, (2)

ECONFAC, WOMANECONFAC, RELIC)

SOU, MID, NOR)

where

GENDERECON is a 0, 1 variable (1 = gender economics class); ACADEMICREP is the academic ranking of the institution; WOMSTUDIES is a 0, 1 variable (1 = women’s studies program); ECONFAC is the number of full-time economics department members; WOMANECONFAC is the percentage of female economics faculty; RELIC is a 0, 1 variable ( I = religious affiliation); WOMANFAC is the percentage of female faculty in the institution; SOU is a 0, 1 variable (1 = southern geographic area); MID is a 0, 1 variable (1 = midwestern geographic area); and NOR is a 0, 1 variable (1 = northern geographic area).

Following Rivers and Vuong (1988), I used a two-stage conditional estimator (2SCE).7 The system was run using an iterated, seemingly unrelated Marquardt- Levenberg technique. Because the model was over identified in the previous form, the variable defining the presence of a women’s studies program, WOM- STUDIES, was eliminated from the first equation by substituting the residuals (RES) from the second equation to form the 2SCE. The model became A

AGENDERCONJ = @W’I, P + RES’WOMSTUDIES, VGENDERCON, * (3)

[ 1 - @(X’j, P -t RES’WOMS-~~JD~ES,~)]~

A where

R E S ~ O M S T ~ D ~ E ~ , = WOMSTUDIES, - WOMSTUDIES,

AWOMSTUDIES,) = @(X’~f~)WoMSTUDIES~[ 1 - @(X’~,(X)]’ - WoMSTUDIES~ (5 )

(4)

where

X I , is the matrix of variables from equation (1) less WOMSTUDIES; XzI is the matrix of variables from equation (2); and @() is a standard normal cumulative distribution function.

RESULTS

Empirical results from the model are shown in Table 2. The R2, adj. R2, and MSE statistics show a good fit for both equations. Signs on the coefficients were appropriate: A positive sign suggested that an increase in that independent vari- able increases the probability that the event will occur.

Academic reputation was the most significant variable in either regression. It

372 JOURNAL OF ECONOMIC EDUCATION

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

TABLE 2 Predicting Gender Courses in Liberal Arts Colleges

~~ ~

Variable

Dependent variables GENDERECON WOMSTUDIES

INTERCEPT

ACADEMICREP

ECONFAC

WOMANECONFAC

RELlG

- I 3 3 (-1.76)

-.03 (-3.68)**

4.52 (3 .22 )**

.I9 (2.50)**

4.43 (3.36)**

-2.03 (-3.4 I )**

WOMANFAC

SOU

MID

NOR

R’ Adj. R’ MSE

.6 I2 ,585 ,072

2.36 (1.99)**

-.03 (&3.27)**

- . X I (-2.88)**

2.87 (3.06)**

.I4 (34 )

I .40 (3. I7)**

3 3 ( I .42)

,454 .443 . I33

Nore: Results fnund using the Marquardl-Levenberg technique of iterated. seemingly unrelated equations following Rivers and Vuong ( IYXX). with r statistics in parentheser. Error terms have been transformed tn he asymptotically

-Significant at the .OS Type I errnr level. Valid.

was measured where the ranking of 1 was best, so that an increase in the acade- mic reputation variable was a fall in academic standing. The negative sign on this variable supported the notion that the better the academic reputation of an insti- tution, the more likely it was to offer women’s courses-ither a women’s stud- ies program or a gender economics course.

A college that maintained a religious affiliation was less likely to offer gender- related courses of study. RELlG had a negative sign on its coefficient and was highly significant in both regressions.

Departmental variables clearly affected the presence of a gender economics course. Larger economics departments with higher levels of female representa- tion had higher probabilities of a gender economics course. Both of the coeffi- cients on these variables were positive and highly significant.

The presence of a women’s studies program was more likely when the fac- ulty had a large percentage of women. The positive sign and significance level

Fall 1997 373

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

of the coefficient on WOMANFAC bore this out. Women’s studies programs were also more likely in the Midwest and North than in the South or West, ceteris paribus.

Care must be taken, however, when interpreting coefficients from models with limited dependent variables8 They do not indicate changes in probabilities of an event given a unit change in the independent variable under consideration. The standard way to interpret the results from a probit model is to use the means of the independent variables in the partial derivatives of the slope parameter^.^ The incremental slope parameters, or partial derivatives, are contained in Table 3. These parameters indicate changes in probabilities of an event given a unit change in the independent variable. For example, an increase in academic stand- ing of 1 unit (a decrease of 1 in ACADEMICREP) will increase the probability of offering a gender economics course by 1.1 percent, whereas the probability of offering a gender economics course will rise by 7.3 percent if the economics fac- ulty increases in size by 1 member (ECONFAC increases by 1).

Analysis of R E S W O M S T U ~ ~ ~ ~ requires more caution still. Because it is the residual of the WOMSTUDIES prediction function, it can be positive only when a women’s studies program exists. It is large and positive only when there is a women’s studies program and the values of the explanatory variables are small, demonstrating a situation in which the institution is atypically positive in its atti- tudes toward women’s issues in the curriculum (as shown by the existence of a women’s studies program) given the level of observable indicator variables (lower levels of these variables). Therefore, R E S W ~ M ~ T U D I E ~ is strongly indica- tive of an institution’s attitude toward women’s issues in the curriculum. Its sig- nificance and positive sign (Table 2) indicate that these unobservable attitudes are extremely important to the existence of a gender economics course.

CONCLUSIONS

Highly ranked institutions, with an accepting attitude toward women, are the most likely to offer gender-related coursework. There is also a strong correlation between female representation on the faculty and these courses. The data from

TABLE 3 Incremental Slope Parameters (Partial Derivatives)

Variable GENDERECON WOMSTUDIES ~ ~

ACADEMICREP RESWOMSTUDIES ECONFAC WOMANECONFAC RELlG WOMANFAC sou MID NOR

-.Ol I I .76 I ,073 1.728 -.792

-.01 I

-.3 13 1 . 1 13 ,056 ,543 .206

~

374 JOURNAL OF ECONOMIC EDUCATION

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

this sample clearly show that these courses are clustered in colleges with high academic standings."' Colleges that retain religious affiliations are less likely to offer gender-related courses. Growing competition among colleges, caused by the recent decline in college-age youth, may increase the number of institutions offering gender-related coursework.

An institution with a large economics department that has a strong female con- tingent will be most likely to offer a gender-related economics course. This unsurprising result confirms standard thinking about course diversity. Women's studies programs are more likely in the Midwest or northern regions of the coun- try than they are in the South. This pattern substantiates generally accepted pre- conceptions about geographic regions and their attitudes toward women.

An extension of this work may be to investigate how these variables affect aca- demic achievement, or persistence, of female students in economics. What we now know is that institutional attitudes play a key role in the existence of women's issues in the curriculum. What role, if any, will those attitudes play in the acade- mic success of female students? Economics is generally known to be a less-pop- ular major with women than with men. Will that pattern continue as more atten- tion is paid to women's issues in the curriculum and to increasing female representation on the faculty?

NOTES

I . Results reported in this article are entirely consistent with a second data set made up of a strati- tied random sample of liberal arts institutions taken from all quartiles of each major category (national and regional). The data set discussed here produced slightly smaller estimation errors.

2. Some colleges (mainly those in the regional category) list business or accounting programs sep- arately from economics programs. I grouped those departments (or programs) together in an attempt to standardize the data across schools.

3. Most adjunct or lecturing appointments in economics are temporary replacements. Because peo- ple in these positions do not usually create courses. they do not usually add to course diversity. For that reason, departmental information w

4. Dividing ranks in this manner (assuming assistant professor and above are the tenurable ranks) is not always accurate, but i t is widely accepted.

5 . In economics. for instance, these generally accepted requirements consist of such courses as principles, intermediate micro, intermediate macro, and. usually, a statistics course. A few other courses such as money and banking or international economics are also extremely common in economics and departments may feel they are required. in some sense, to offer them.

6. Demand variables were not significant for these models. Variables such as the number of stu- dents. the percentage of female students, and the ethnic diversity of students were all investigat- ed and dropped.

7. As shown in Rivers and Vuong (1988). the ZSCE will be both consistent and efficient. The tech- nique used to estimate a system of two probit equations required an iterated estimation proce- dure. To ensure robustness, I used a variety of starting conditions. All results were identical. As is often the case with two-stage procedures, the standard errors required correction. Only the cor- rected standard errors are reported.

lculated omitting these individuals.

8. This discussion is derived from Judge et al. (1985). 9. The model can be interpreted by using the normal estimating function with the model results

from Table 2. Set all but one of the independent variables to their means and determine the change on the outcome probabilities, using the cumulative normal distribution function, given a unit change from the mean in the independent variable of choice. The nonlinearity of the cumu- lative normal distribution function makes this procedure more accurate around the means of the independent variables than elsewhere.

10. The data in the sample as well as data from a stratified random sample taken from all quartiles of colleges support these ideas

Fall 19Y7 375

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4

REFERENCES

Academe. 1993. The annual report on the economic status of the profession: 1992-1993.

America’s best colleges. 1993. U.S. News und World Report. Washington, D.C. Committee on the status of women in the economics profession. Annual Reports. Winter issues,

Conrad, C. A. 1992. Evaluating undergraduate courses on women in the economy. American Eco-

Judge, G. G., W. E. Griffiths, R. C. Hill, H. Liitkepohl, and T. C. Lee. 1985. The theory and practice

NWSA directory of women’s studies programs. women’s centers, and women’s research centers.

Rivers, D., and Q. H. Vuong. 1988. Limited information estimators and exogeneity tests for simulta-

The World almanac and book of facts 1993. New York: World Almanac: An imprint of Pharos Books,

Triniry College bulletiri 1992-93. Hartford, Conn.

(March/April):34-8I,

CSWEP Newsletters 1990, 1991, 1992, 1993, 1994, 1995, 1996.

nomic Review: Papers and Proceedings 82 (May): 565-69.

of econoinerrics (2nd ed.). New York: Wiley.

1990. New Jersey: A Publication of the National Women’s Studies Association.

neous probit models. Journal of Econometrics 39 (November): 347-66.

A Scripps Howard Company.

CALL FOR PAPERS AND PARTICIPATION

Academy of Economics and Finance (Formerly Midsouth Academy of Economics and Finance)

Montgomery Embassy Suites Hotel Montgomery, Alabama February 11-14, 1998

Proposals for papers, panels, and sessions in all areas of economics and finance are invited for the meeting. If editorial requirements are met, papers presented at the meeting are eligible for inclusion in the published Proceedings of the meeting. Participants for session moderator and discussant responsibilities are also welcome. Submission deadline is October I , 1997.

Program Chair: Kristin K. Howell Department of Economics and Finance University of North Carolina at Wilmington 601 South College Road Wilmington, NC 28403-3297

Executive Secretary: Paul Merkle Department of Economics Louisiana State University-Shreveport Shreveport, Louisiana 7 1 1 I5

Local Arrangement Chairs: David Lange and David Sollars Auburn University of Montgomery

7300 University Drive Montgomery, AL 361 I7

376 JOURNAL OF ECONOMIC EDUCATION

Dow

nloa

ded

by [

The

Uni

vers

ity o

f M

anch

este

r L

ibra

ry]

at 1

1:07

10

Oct

ober

201

4