Elite Competition, Religiosity, and Anti-Americanism in the Islamic World

7/30/2019 Elite Competition, Religiosity, and Anti-Americanism in the Islamic World

1/19

American Political Science Review Page 1 of 19 May 2012

doi:10.1017/S0003055412000135

Elite Competition, Religiosity,and Anti-Americanism in the Islamic WorldLISA BLAYDES Stanford UniversityDREW A. LINZER Emory University

The battle for public opinion in the Islamic world is an ongoing priority for U.S. diplomacy.

The current debate over why many Muslims hold anti-American views revolves around whetherthey dislike fundamental aspects of American culture and government, or what Americans do ininternational affairs. We argue, instead, that Muslim anti-Americanism is predominantly a domestic, elite-led phenomenonthat intensifies when thereis greater competition between Islamist and secular-nationalistpolitical factions within a country. Although more observant Muslims tend to be more anti-American,paradoxically the most anti-American countries are those in which Muslim populations are less religiousoverall, and thus more divided on the religioussecular issue dimension. We provide case study evidenceconsistent with this explanation, as well as a multilevel statistical analysis of public opinion data fromnearly 13,000 Muslim respondents in 21 countries.

Since September 2001, survey researchers havequestioned citizens of the Islamic world abouthow they view Americans, U.S. policy, and

American values and culture. The depths of anti-Americanism revealed by these surveysespeciallywithin the Arab worldhave been the subject of muchdiscussion (Abdallah 2003; Fuller 2002; Gentzkow andShapiro 2004). The consequences of this anti-Americansentiment are seen as delegitimizing American values,increasing sympathy for Americas enemies, and weak-ening Americas influence in foreign affairs (Keohaneand Katzenstein 2007; Nam 2003; Shore 2005). Lessnoticed, however, is that in many countries, Muslimsactually tend to view the United States quite favorably.Even within the Arab world, there is a considerableamount of cross-national variation in levels of anti-Americanism. Why, then, do some Muslims harborsuch intense dislike for America, whereas others aremore neutral or even supportive of the United Statesas a global actor?

Anti-Americanism has been defined as any hos-tile act or expression that becomes part and par-cel of an undifferentiated attack on the foreign pol-icy, society, culture and values of the United States(Rubinstein and Smith 1988, 36) and more broadly asa psychological tendency to hold negative views ofthe United States and of American society in general

Lisa Blaydes is Assistant Professor, Department of PoliticalScience, Stanford University, Encina Hall West, Suite 100, Stanford,

CA 94305 ([email protected]).Drew A. Linzer is Assistant Professor, Department of PoliticalScience, Emory University, 102 Tarbutton Hall, 1555 Dickey Drive,Atlanta, GA 30322 ([email protected]).

The authors thank Christopher Anderson, Ceren Belge, GiacomoChiozza, Tom Clark, Jorge Dominguez, James Fearon, Justin Grim-mer, Nahomi Ichino, Shanto Iyengar, David Laitin, Monika Nalepa,ChrisReenock, Ed Schatz, Paul Sniderman,Jeffrey Staton, JonathanWand, Carrie Wickham, and the audiences of the Notre Dame Kel-logg Institute, the MIT Works-in-Progress Seminar Series, the UCSan Diego Comparative Politics Speaker Series, and the StanfordComparative Politics Workshop. Jana Marie Hutchinson, Ugur Pece,Jeremy Voss, and Meredith Wheeler provided exemplary researchassistance. The Pew Global Attitudes Project bears no responsibilityfor the interpretations presented or conclusions reached based onour analysis of the data.

(Katzenstein and Keohane 2007, 12). Commentatorshave argued at length over why Muslims in particularmight adopt such a stance, with most explanations set-

tling into one of twocategories. The first is that Muslimswho dislike the United States do so on the basis ofcore differences over societal norms and values. Paz(2003, 53), for example, writes that Islamiststhosewho advocate the formal integration of Muslim socialand religious precepts into governmentview conflictwith the United States as a war of cultures and thatthe nature of Islamist anti-Americanism is culturalrather than military or political. Many other schol-ars, however, regard this view as a myth (Espos-ito and Mogahed 2008, 140). Instead, they contendthat Muslims dislike the United States not for whoAmericans are, but rather for what Americans do (e.g.,Abdallah 2003; Cole 2006; Shore 2005; Tessler2003). As described by Makdisi (2002, 538), Anti-Americanism is a recent phenomenon fueled by Amer-ican foreign policy, not an epochal confrontation ofcivilizations.

Despite their disagreement over the root causes ofanti-Americanism, both theories share a presumptionthat individuals form their opinions about the UnitedStates primarily as a direct reaction to what the UnitedStates is or does. Although this may be true in part,it neglects the important intermediary role played bypolitical elites in determining what information aboutthe United States individuals hear, how they interpretthis information, and how they incorporate it into their

political perspective (Zaller 1992). Like people every-where, Muslims are open to persuasion on the issueof anti-Americanism and susceptible to elite influencethrough themass media (Lynch 2007; Nisbet andMyers2011).

In this article, we propose a theory of anti-Americanism that transcends the conventional whatAmerica does versus who America is debate. Wetrace the source of Muslim anti-Americanism to theintensity of domestic political competition between acountrys Islamist and secular-national factions. Lead-ers on both sides of this dominant cleavage in theMuslim world can credibly claim opposition to the

1


2/19

Anti-Americanism in the Islamic World May 2012

United States. When the struggle for political controlbetween these two groups escalates, elites of both typeshave incentives to ramp up anti-American appealsto boost mass support. More generally, when oppos-ing politicians share similar stances on salient issues,heated competition can lead to more extreme rhetoricas politicians attempt to ideologically outbid theiropponents. As elites anti-American pronouncements

increase, larger numbersof Muslims hear, consider, andare led to adopt anti-American attitudes. In contrast,where conflict is minimal, individuals are less exposedto anti-American messages and subsequently reportlower levels of anti-Americanism. This is not to say thatAmerican actions or the United States unique positionin global political affairs is irrelevant to explaining anti-Americanism. Rather, it is that domestic elites find inthe behavior of America and Americans justificationsfor anti-American claims.

This logic explains whyseemingly paradoxicallyalthough religious Muslims are more anti-Americanthan theirsecular compatriots, anti-American attitudesare most prevalent in more secular countries where the

political division between religious and nonreligiousindividuals is the greatest. Because the split betweenreligious and secular-nationalist types in a societychanges slowly over time, our theory also accounts forwhy Muslim countries tend to return to a steady-statelevel of anti-American sentiment even after politicalshocks to public opinion, as well as why far greatervariationexists across Muslim communities thanwithinMuslim communities over time.

Scholars have termed domestic sources of anti-American attitudes instrumental anti-Americanism,reflecting the efforts of a Muslim political elitethat instigates and manipulates hostility toward the

United States in order to mobilize domestic support(Rubinstein and Smith 1988, 41). Our theory offers anexplanation for the conditions under whichsuchinstru-mental anti-Americanism is more or less likely to befound; applying a multilevel statistical model to pub-lic opinion survey data from nearly thirteen thousandMuslim respondents in 21 countries, we are the firstto systematically test forand find empirical evidenceofinstrumental anti-Americanism at work.

Increasing scholarly understanding of the roots ofanti-American sentiment in the Muslim world has im-portant and far-reaching political implications. Oneand a half billion Muslims make up one-fifth of thetotal world population, and favorable attitudes to-

ward the United States are rarer in the Muslim worldthan anywhere else (Kohut and Stokes 2006). Many ofthe worlds most intractable conflicts involve Muslim-majority countries, and the ability of the United Statesto exercise soft power to influence the trajectoriesand outcomes of these disputes is of considerable im-portance.1 One would be hard pressed to think of atime when Muslim attitudes toward the United Statescarried greater political import. As Arab citizenries

1 Nye (1990, 2004) describes soft power as the ability to attain policyobjectivesthrough cooptation, persuasion, and attraction rather thancoercion or through the use of side payments.

across the Middle East engage in both peaceful andmore violent protest against dictatorssome of whomenjoyed close ties to the United Statesa reassessmentof U.S.-Arab and U.S.-Muslim relations is underway,withthe potentialto criticallyreshape Americasabilityto promote its values and interests in the region.

SOURCES OF MUSLIM OPPOSITION

TO THE UNITED STATES

The grievances that motivate many Muslims to expressanti-American sentiments have been linked both tospecific actions taken by the United States in foreignpolitical and economic affairs and to Americas grow-ing global cultural influence since the endof World WarII (Rubinstein andSmith1988).2 The questionthatcon-temporary scholars continue to debate is which amongthis amalgam of discontents (Kohut and Stokes 2006,23) matters most for explaining anti-American atti-tudes. We briefly review the two sets of long-term fac-tors generally seen as providing the basis for Muslimanti-Americanism.

U.S. Policy and the Foreign Backlash

Several key U.S. foreign policy developmentsin the sec-ond half of the 20th century have provided fodder forthe spread of anti-American sentiment in the Islamicworld. The first relate to Americas involvement in theinternal political and economic affairs of countries withlargeMuslim populations. During the ColdWar, Amer-ica propped up dictators seen as friendly to the UnitedStates and worked to topple regimes with leanings to-ward the Soviet Union. American overidentificationwith unpopular local dictators like the Shah of Iran

or Sadat in Egypt was one factor contributing to massanti-American sentiment (Parker 1988, 53).American government agencies, such as USAID,

also intervened in the economic planning of severaldeveloping Muslim countries (Thornton 1988, 10). Be-cause aid was disbursed conditionally, a perceptionemerged that the United States used foreign assis-tance to enslave a country politically. For example,American involvement in Pakistani political and eco-nomic life is widely deplored, and secular elites tendto describe their ill feelings toward the United Statesin terms of capitalist or imperialist exploitation(Kizilbash 1988, 59, 63). Muslim leftists, such as Egyp-tian economist Samir Amin, argue that American eco-

nomicintervention in theMiddle East is part of a globalimperialist strategy on the part of the United States.

Another key U.S. foreign policy development in thepostWorld War II era is American support for thestate of Israel. There is widespread anger on the partof many Muslims toward the United States for its po-litical support of Israel, which is invariably seen as

2 During the first half of the 20th century, the image of the UnitedStates in the Arab worlda key constituency in the broader set ofIslamic countrieswas generally positive, because Arabs saw Amer-icans less as imperialists and more as liberal, benevolent educatorsand missionaries (Makdisi 2002).

2


3/19

American Political Science Review

coming at the expense of Palestinian interests. Thefate of Jerusalemthe third most holy city in Islamand the location of important Muslim holy sitesisalso of importance to Muslims around the world. De-scribing the Arab world, Hammond (2007, 57) writesthat views of the United States today are first andforemost conditioned by American policy vis-a-vis theIsrael-Palestinian conflict and the degree to which the

United States is seen backing Israel to the detriment ofthe Palestinians.

Islamism and Cultural Anti-Americanism

One of the most important political developments toemerge in the Muslim world over the last four decadeshas been the growth of support for what many termIslamismthe idea that both society and politicsshould be infused with a greater religious sensibility.This has been coupled with the rise of the mosquemovement. According to Mahmood (2004, 4447), thismovement emerged in response to the perceptionthat religious knowledge, as a means for organizing

daily life, had become increasingly marginalized undermodern structures of secular governance . . . . Piety ac-tivists [sought] to imbue each of the various spheresof contemporary life with a regulative sensibility thattakes its cue from the Islamic theological corpus ratherthan from modern secular ethics.

Within the Muslim world, many of the strongestproponents of anti-American attitudes are found inIslamist politicalcircles (Faath and Mattes2006). Fuller(2002, 5455) comments that most in the Muslimworld feel themselves besieged by the West . . . . Is-lamist movements today provide a key source of iden-tity to peoples intent on strengthening their social co-

hesion against Western cultural assault. Some Islamistactivists believe that the United States represents aprimary threat to Muslim society and is to blame fora variety of domestic and international political prob-lems (Ajami 2003). Others see the United States asthe neo-Mongol power lurking behind the apostategovernments that they seek to topple (Doran 2002,183). In Arab countries, Islamists promote the idea ofa global Western conspiracy against the Arabs andtheArabandMuslim world, which providesthe Islamistswith their main justification and motive for developingthe image of the American enemy (Paz 2003, 53).

EXPLAINING ANTI-AMERICANISM

The unpopularity of certain U.S. policies and neg-ative reaction to the spread of American cultureare both important reasons why Muslims around theworld might hold anti-American attitudes. Neverthe-less, many Muslims are not anti-American. In 2007,the Pew Global Attitudes Project (GAP), a cross-national opinion poll sponsored and directed by thenonpartisan Pew Research Center, asked respondentsin 21 countries with significant Muslim populations ifthey had a very favorable, somewhat favorable, some-what unfavorable, or very unfavorable opinion of theUnited States. Large proportions of Muslims around

the world told researchers that they in fact had a favor-able view of the United States (Figure 1).3 To explainanti-Americanism, it is not enough to demonstrate thatpotential grievances exist: A valid theory must alsoaccount for the observed variation in anti-Americansentiment not only across Muslim individuals but evenmore strikingly across Muslim countries.

We propose that Muslim attitudes toward America,

like most political attitudes, are shaped by the messagesand arguments to which individuals are exposed bypolitical elites and the mass media. In countries whereelites are the most outspoken in their opposition to theUnited States, anti-American sentiment will be great-est. We argue, moreover, that local political factorsoffer one explanation for why elites in some countriesmay be more aggressive in promoting anti-Americanattitudes. Where there is more intense political compe-tition between Islamist and secular-nationalist groups,domestic elites have much stronger incentives to ex-ploit grievances against the United States for politicalgain.

Unlike existing theories, we describe a specific mech-

anism through which domestic political factors influ-ence elite messages about the United States and con-sequently shape public opinion. Our theory also ex-plains a notable empirical regularity: Levels of oppo-sition to the United States in Muslim countries areextremely consistent from year to year (Pew GlobalAttitudes Project 2009). Most variation, instead, isfound across countries in their long-term, baseline lev-els of anti-Americanism. As observed by Schatz (2008,9), attitudes about the United States do not changein lock-step with U.S. policies. Rather, they becomesedimented in domestic cultures, institutions, and con-texts. Even the increase in Muslim anti-Americanism

following the 2003 U.S. invasion of Iraq was, as de-scribed by Chiozza (2007, 125), a momentary reactionto the exceptional circumstances of the Iraq War ratherthan a structural shift in the popular perceptions of theUnited States. By 2004, Muslim perceptions of theUnited States had returned to prewar, 2002 levels, andhave remained mostly steady since that time.

Elite-led Opinion Formation

As a phenomenon of public opinion,anti-Americanismdoes not emerge solely as a response to U.S. actionsand values, but is also shaped by the way local politi-cal leaders describe and discuss the United States. As

Zaller (1992) describes, the opinions people expressare a combination of their own personal experiencesand the balance and intensity of the elite attitudes towhich they are exposed and are predisposed to ac-cept. Individuals who identify with particular opinionleaders will tend to align their viewpoints with thoseelites.4 For those without strong political attachments,

3 The 21 study countries include 68% of the worlds Muslim popula-tion (Pew Research Center2009).Of thecountries with the10 largestMuslim populations, only Iran and Algeria were not surveyed.4 On the role of elite persuasion in opinion formation, also seeKinder (1998); Druckman and Lupia (2000); and Gabel and Scheve(2007).

3


4/19


FIGURE 1. Distribution of Favorable and Unfavorable Opinions toward America among Muslims in21 Pew GAP Study Countries in 2007 (Question 16A)

Ivory Coast

Mali

Senegal

Ghana

Uganda

India

KenyaBangladesh

Nigeria

Kuwait

Tanzania

Ethiopia

Lebanon

Indonesia

Egypt

Morocco

Jordan

Pakistan

Palestinian ter.

Malaysia

Turkey

100 80 60 40 20 0 20 40 60 80 100

20

20

30

31

36

38

44

45

49

52

52

53

70

72

78

79

80

82

90

90

90

80

80

70

69

64

62

56

55

51

48

48

47

30

28

22

21

20

18

10

10

10

Percent antiAmerican Percent proAmerican

Notes: Light gray bars denote more extreme opinions on either side. Percentages exclude respondents with no opinion.

what matters is the prevalence of different argumentsin their social or mass media environment. FollowingZallers one-message model, the more insistentlythat elites promote anti-Americanism, the more thatindividuals tend to adopt anti-American attitudes. Incountries where elites project a positive stance towardthe United States, individuals should be relatively morepro-American as well.

Domestic elites, communicating through local massmedia, are a primary source of political news and

opinion about the United States. A secondary streamof information is provided by transnational satellitenews networks such as Al Jazeera, Al Arabiya, andCNN, which each have their own editorial slant. Ex-posure to Al Jazeera is especially significant becauseof the channels perceived anti-American stance (e.g.,Ajami 2001). Survey evidence indicates that watchingAl Jazeera is associated with more negative attitudestoward the United States (Gentzkow and Shapiro 2004;Nisbet and Myers 2011). At the country level, however,Muslim anti-Americanism has remained fairly stableover time (Pew Global Attitudes Project 2009), even asAl Jazeera hasexpandedits reach. This finding suggests

that the influence of transnational news networks onanti-Americanism can only be one part of a broaderexplanation.

If our argument about domestic, elite-led opinionformation is correct, then the crucial question becomeswhy political debate in certain countries is more orless saturated with anti-American messagesand fromwhich segments of elite leadership to which sets offollowers. To understand how anti-American attitudesenter into mass opinion, we describe (1) the domi-

nant political cleavages in Muslim society, which af-fect how individuals choose to align with competingelites; (2) the positions held by elites in each compet-ing bloc; and (3) how the intensity of political com-petition between opposing groups affects the strengthwith which political elites advance anti-Americanclaims.

The ReligiousSecular Cleavage

In large parts of the Islamic world, the substance of po-litical debate surrounds a secularreligious issue cleav-age akin to the leftright ideological dimension that

4


5/19


describes policy preferences in most Western democ-racies. Debate concerning the nature of the governingregime and the role of the state in society is centralto the secularreligious divide. Roy (1994, 23) writesthat, despite the protests of some political actors, anypolitical action amounts to the automatic creation ofa secular space or a return to traditional segmenta-tion where a more traditional space refers to one

with a greater religious sensibility. Hunter (1995, 327)argues that Muslim-majority countries are character-ized by a rift between the more Westernized and themore traditional segments in both social and politicalaffairs.

The modern historical basis for this cleavage datesto the mid-19th century (Owen 2010). Hunter (1998,7576) explains that as European powers increasinglygained in economic and political prominence relativeto the Muslim world, MuslimWestern relations thathad previously been characterized as a competitionbetween equals evolved into that between the dom-inating and the dominated. Within Muslim societies,a debate arose as to whether stricter adherence to reli-

gious principles was the culprit or the solution forthis change. From the 1920s through the 1970s, propo-nents of the former argument dominated positions ofpolitical power, as most Muslim societies underwenta process of state-directed secularization and culturaland political nationalization (Hunter 1998, 85). Sincethen, however, Islamism has emerged as the primaryideological rival to secular nationalismthough thepace at which secular nationalism has given way toIslamism varies considerably across states.

A range of explanations for the rise of Islamismhave been proposed. Some argue that secular nation-alism failed to successfully incorporate dissatisfied so-

cial groups and classes (Sutton and Vertigans 2005).In the Arab states, increasing support for Islamismhas also been associated with the Arab defeat in the1967 Arab-Israeli War (e.g., Ajami 1992); accordingto Hammond (2007), Islamists and secular nationalistscontinue to argue vociferously about how best torespond to that defeat. Murphy (2002) argues that, forMuslim individuals, identification with Islam counter-acts feelings of inferiority toward the West, provid-ing a renewed sense of cultural dignity. Others suchas Wickham (2002, 6) see the growth of Islamism asdue more to the successful mobilizing efforts of op-position [Islamic] elites than as a natural result ofaccumulated grievances. Whatever the exact cause,

Browers (2009, 1) observes that even when secularnationalists and Islamists confront a common oppo-nent in the form of corrupt, authoritarian regimes, theyhave still shown themselves to be each others worstenemy.

The Value of Anti-American Rhetoric

Although contemporary anti-Americanism is primar-ily associated with Islamism, in fact proponents ofboth Islamism and secular-nationalism publicly criti-cize the United States. For secular leaders, who histor-

ically positioned themselves as anticolonial and anti-imperialist, anti-American attitudes reflect a percep-tion of America as encroaching on the independenceand sovereignty of countries in the Muslim world.Reetz (2006, 186) writes of Pakistan that left-leaninganti-American arguments play a central role in thepublic debate, especially in the largely independentprintmedia. In the Arab world, Hammond (2007, 205)

describes both camps as virulently anti-American.In Indonesia, Bowen (2007, 245) finds anti-Americansentiments in the pronouncements of both Islamistsand non-Islamists, the latter tending to focus on whatthey see as Americas neo-imperialist economic ambi-tions. In Egypt, both secularists and the Muslim Broth-erhood have orchestrated anti-American demonstra-tions. (Mitchell 2004, 98).

Because of the political, economic, and culturalhegemony of the United States, large segmentsof Muslim society are receptive to anti-Americanrhetoricfrom whichever side it comes. Opinion sur-veys indicate that most Muslims believe Americansare not religious enough and that the religious beliefs

that they do hold drive the United States to make baddecisions in the world (Kohut and Stokes 2006, 93).Although many individuals across the Muslim worldenjoy American movies, television, and music, theyalso view globalization and the spreading influence ofAmerican culture as potential threats to local beliefs(Esposito and Mogahed 2008; Faath and Mattes 2006;Hammond 2007; Kohut and Stokes 2006).

Against this backdrop, anti-American appeals canrepresent a successful instrumental strategy for politi-cal elites seeking to gain popular support. During theCold War, the association between anti-Americanismand communism gave anti-American rhetoric a cer-

tain strength, but at the same time placed limitson its ability to appeal broadly. More recently, anti-Americanism has become what Krastev (2004, 6)describes as an all-purpose ideologycompellingyet vague enough to be harnessed by any numberof political groups for cynically designed politicalstrategies seeking to mobilize supporters for politi-cal gain.5 Similarly, Schatz (2008, 1213) views anti-Americanism as a symbolic resource used by polit-ical activists to link their on-the-ground . . . goals tolarger-scale (even global) imperatives, thereby legit-imizing and enhancing the significance of their politicalmovements.

The Intensity of Elite Competition

Despite its potential appeal, not all Muslim politicalleaders pursue anti-American claims or do so withequal force. Many have little to gain by doing so.In Muslim communities where religious leaders al-ready monopolize local political control, competition

5 Krastev (2004) also sees anticorruption and antiterrorism rhetoricas having many of the same properties as the instrumental use ofanti-Americanism to mobilize supporters.

5


6/19


for converts is no longer as fierce. The battle forlocal supremacy has already been won by those whoare more religious, leaving neither side with strong in-centives to invoke grievances against the United Statesto recruit supporters.

Where competition is more intense, both Islamistsand secularists will seek to lay claim to anti-Americanism, because it remains an issue with which

both sides possess a credible association. From a histor-ical perspective, secularists associated with left-leaningpolitical movements positioned themselves in opposi-tion to theUnited States as part of thepolitical environ-ment of the Cold War. Since the end of the Cold War,Islamists have also sought to leverage anti-Americansentiment for political endsand have enjoyed greatpolitical success with these efforts. Although Islamistsmay currently enjoy an advantage over secular factionson the issue of anti-Americanism (as our empirical re-sults indicate), neither side can afford to abandon thishighly salient, all-purpose issue altogether.

There are clear parallels to the logic of issue own-ership. Political entrepreneurs engage in a type of ide-

ological outbidding to avoid the costs of ceding atough stance on the United States to their opponents.In the American context, scholars of election cam-paigns have argued that candidates should focus voterattention on issue areas that play to their strengths orhighlight a weak area for a political opponent (Petro-cik 1996; Spiliotes and Vavreck 2002). In many cases,however, politicians intentionally trespass on issues as-sociated with an opponent (e.g., Damore 2004; 2005).As explained by Pfau and Kenski (1990), elites seekto neutralize potential vulnerabilities and will engageon an issue to prevent opponents from positioningthemselves in a way that may pose a future political

threat.Our argument also parallels research on the politicaleconomy of religion. Scholars of religious marketssuggest that, when religious groups become dominantin a local setting, religious adherence declines. Forexample, U.S. cities with higher rates of religious di-versity have elevated levels of adherence, increasedSunday school attendance, a larger number of childrenin Catholic school, and more priestly ordinations (e.g.,Finke and Stark 1988; Stark and McCann 1993). Gill(1998) argues that in Latin American countries wherethe Catholic Church faced competition from Protes-tantism, the Churchseeking to maximize member-ship and financial resourceshad an incentive to sup-

port democracy to avoid losing poor parishioners withmore democratic preferences. We believe that the logicof the lazy monopolist applies to competition forsupport within the Islamic world as well.

Implications for Observed Levelsof Anti-Americanism

If anti-Americanism is a manifestation of elite opin-ion leadership, then anti-American attitudes shouldbe most widespread in countries where elites fromacross the political spectrum have incentives to pro-

mote grievances against the United States. We expectthat intense competition between political elites alongIslamistsecular lines provides these incentives.

One factor contributing to the intensity of elite com-petition along secularreligious lines is the baselinelevel of religiosity in a countrys Muslim population.The effect of this variable on anti-Americanism atthe country level should be non-monotonic: In highly

religious and highly secular contexts, there is a lessbalancedand hence, less competitivepolitical envi-ronment, so levels of anti-Americanism should be rel-atively low. In the contemporary Muslim world, how-ever, we do not observe countries with entirely secu-lar populations. As a result, the empirical predictionswe yield refer only to the second half of an invertedU-curve, suggesting that as religiosity becomes morewidespread, elite competition diminishes, so levels ofanti-Americanism are expected to fall.

At the country level, we also predict that the sizeof a countrys Muslim population as a share of thetotal population will be associated with greater anti-Americanism. The reasons are consistent with our

theory, but we are primarily concerned that thisvariable might confound an empirical relationship be-tweencompetitiveness and levelsof anti-Americanism.Where Muslims dominate domestic politics, the stakesof secularIslamist competition increase, and messagesvoiced by Muslim leaders become more prevalent inpolitical discourse. Conversely, countries with fewerMuslims are less anti-American overall (that is, includ-ing the non-Muslim population), which may affect theopinions of Muslims in those countries.

The theory is associated with additional predic-tions about individual-level differences in average lev-els of anti-Americanism within countries. If a coun-

trys Islamist leaders are especially anti-American intheir rhetoric, then their religiously observant follow-ers will receive and internalize those considerations,reporting stronger anti-American views. To the ex-tent that secular-nationalist elites follow suit in ex-pressing anti-American sentiments, their followers willecho such attitudes. Because Islamist political activistshave emerged as some of the most vocal opponentsof the United States, we expect the pious followersof Islamist political elites to demonstrate greater anti-Americanism. Likewise, we expect that individualswho follow international news more closely will reportstronger anti-American attitudes, because popular out-lets such as Al Jazeera contain content that amplifies

criticism of the United States. Those who prefer tofocus on domestic news may not be exposed to thisadditional influence. Taken together, the most anti-American individuals, on average, should be those whoare both highly religious and highly attuned to interna-tional affairs.

Finally, if the underlying mechanism that we haveproposed as linking elite behavior to mass attitudesis correct, then we should find that depictions of theUnited States in the domestic mass media will bepredominantly negative in highly competitive coun-tries, but neutral or even positive in less competitivecountries.

6


7/19


TABLE 1. Survey Items Used to Create the Anti-Americanism Scale

Item Question Wording

Q16a Please tell me if you have a very favorable, somewhat favorable, somewhat unfavorable or veryunfavorable opinion of the United States?

Q16b Please tell me if you have a very favorable, somewhat favorable, somewhat unfavorable or veryunfavorable opinion of Americans?

Q27 Which of the following phrases comes closer to your view? Its good that American ideas and customs arespreading here, or its bad that American ideas and customs are spreading here.

Q28 Which of these comes closer to your view? I like American ideas about democracy, or I dislike Americanideas about democracy.

Q29 Which comes closer to describing your view? I like American ways of doing business, or I dislikeAmerican ways of doing business.

Q30 Which is closer to describing your viewI like American music, movies and television, or I dislikeAmerican music, movies and television.

Q31 Which comes closer to describing your view? I admire the United States for its technological andscientific advances, or I do not admire the United States for its technological and scientific advances.

Source: Pew Global Attitudes Project (2007).

MULTILEVEL STATISTICAL ANALYSIS OFMUSLIM PUBLIC OPINION

We begin our analysis by considering the determi-nants of anti-Americanism at both the national andindividual level using public opinion survey data col-lected by the Pew Global Attitudes Project (2007). The2007 GAP study interviewed nearly 13,000 Muslim re-spondents in 21 countries, spanning a geographically,economically, and culturally diverse range (Figure 1).Although individual attitudes are not the only way inwhichanti-Americanism can be expressed, we considerresponses to the GAP study to provide the most cross-nationally consistent source of information about per-ceptions of America at a single point in time.6 Country-level sample sizes range from 34 in Kenya to 1,930 in

Pakistan.

Measurement

Anti-Americanism describes a wide range of antag-onisms and grievances directed at the United States,its government, policies, culture, and even individualcitizens. For this reason, isolating the concept in survey-based research may be difficult (Katzenstein andKeohane 2007). In addition, as in most surveys, weexpect a large amount of measurement error in re-sponses to any single question about the United States.Averaging across a series of related items can signif-

icantly improve the reliability of survey-based mea-sures, while allowing for a more complete opera-tionalization of the notion of anti-Americanism (An-solabehere, Rodden, and Snyder 2008). The Pew GAPstudy contains a battery of seven questions that weuse to create an anti-Americanism scale. These items

6 Pew has fielded a Global Attitudes survey in every year from 2002to 2010, but the 2007 study that we analyze is the most compre-hensive of the entire series. Of the 21 study countries, only Jordan,Pakistan, and Turkey were surveyed in all nine years. Even whencertain countries were surveyed in multiple years, the wording of theGAP questionnaire often changed.

elicit individuals opinions of the United States, Amer-icans, and the spread of American customs, ideas about

democracy, ways of doing business, cultural products,and technological and scientific advances (Table 1). Werescale each item to range from 0, indicating completefavorability of the United States, to 1, indicating themaximum level of anti-American sentiment, and aver-age across responses. This is our dependent variable.Country-level means vary from 0.23 in Ivory Coast to0.84 in Turkey.7

At the individual level, we employ two distinct mea-sures of respondents secularIslamist political predis-positions.The firstis based on individuals self-reportedlevel of religiosity, which classifies respondents accord-ing to the religioussecular issue dimension that isdominant in the Muslim world. The GAP survey asksthree questions pertaining to Muslims level of reli-gious commitment: their frequency of prayer, fasting,and the importance of religion in their life (Table 2).Responses to these questions demonstrate consistentpatterns: Of the 87% of Muslims who say that religionis very important, 91% fast at least through Ramadan,and 68% pray five times a day. Praying five times aday is a particularly good standard by which to judgean individuals level of religiosity because, although themid-day, afternoon, sunset, and evening prayers tendtotake place when most individuals are awake, the dawn,orfajr, prayer occurs whenmost individuals are asleep.8

Willingness to rise for the dawn prayer demonstrates

a high level of religious commitment. We thereforecreate a dichotomous variable for piety, considering

7 At the country level, average levels of anti-Americanism using thisscale correlate at 0.89 with the percent anti-American in Figure 1.The scale also validates against country-level disapproval of the jobperformance of the leadership of the United States in the Gallup,Inc. (2008)World Poll, witha correlation of 0.80. Cronbachs fortheseven-itemscale is 0.78, indicatinga highlevel of internal consistency.8 Take, for example, prayer times in Cairo for a typical day in Oc-tober. The first four prayers were at 11:41 am, 2:59 pm, 5:28 pmand 6:45 pmall times during the waking hours of most individuals.The dawn prayer, on the other hand, was scheduled to take place at4:28 am.

7


8/19


TABLE 2. Survey Items Used to Measure Religiosity (Q114, Q116, and Q117), thePerceived Intensity of SecularIslamist Conflict (Q75), Identification with Either Group (Q75b),and Media Exposure (Q37)

Item Question Wording

Q114 How often, if at all, do you pray: hardly ever, only during religious holidays, only on Fridays, only on Fridaysand religious holidays, more than once a week, every day at least once, or every day five times?

Q116 How often, if at all, do you fasthardly ever, some days during Ramadan, during most or all days ofRamadan, or during all of Ramadan and other religious holidays?

Q117 How important is religion in your lifevery important, somewhat important, not too important, or not at allimportant?

Q75 Do you think there is a struggle in our country between groups who want to modernize the country andIslamic fundamentalists, or dont you think so?

Q75b If answered yes to Q75: Which side do you identify with more in this struggle, the groups who want tomodernize the country or Islamic fundamentalists?

Q37 Which of the following two statements best describes you: I follow international news closely only whensomething important is happening or I follow international news closely most of the time, whether or notsomething important is happening?

Source: Pew Global Attitudes Project (2007).

individuals who answer yes to all three questions to be

the most highly religious.9The second measure of individuals political alle-

giances uses responses to a GAP item asking respon-dents whether they identify more closely with groupswho want to modernize the country or Islamic fun-damentalists. Although this question would appear toproduce an apt measure, it is only asked of respondentswho first report awareness of a struggle in our countrybetween groups who want to modernize the countryand Islamic fundamentalists (Table 2). Because thissubgroup does not comprise a representative sample,our inferences from this variable, although informa-tive, are less generalizable than those from the variablemeasuring piety.

To assess individuals media exposure, we examineresponses to a survey question asking whether respon-dents follow international news closely only whensomething important is happening or most of thetime, whether or not something important is happen-ing. In most countries, the proportion who indicatefollowing international news closely most of the timeis between 40% and 60%.

At the country level, we require a measure of the in-tensity of conflict between Islamistand secular factions.This conceptual variable is difficult to operational-ize in a cross-nationally consistent manner, becausemany countries in our sample do not have traditionally

consolidated democratic institutions. For example, wecannot measure political competition as the ratio ofIslamist to secular politicians in parliament or their as-sociated vote margins. Even where elections are held,they may not reflect the balance of political powerbetween secular and Islamist organizations, because

9 We chose not to use a more finely grained measure of religiositybecause it was unclear from a theoretical perspective how to weightthe relative importance of each component of our measure beforeaggregating. To validate ourpiety measure, we fit a two-cluster latentclass model to these responses, which confirmed the categorizationof survey respondents into a more and less religious division (Linzerand Lewis 2011).

political elites may not be seeking office, per se, as an

end goal. Indeed, Islamist activists frequently see theirobjectives as being much wider ranging than politicalrepresentation through existing institutions; their goalsoften focus on a broader Islamization of society.

We overcome this challenge in two ways, both ofwhich take further advantage of questions includedin the Pew GAP survey. In 17 of the 21 study coun-tries, Muslim respondents were asked if they perceiveda struggle between modernizing forces and Islamicfundamentalists (as described earlier). We assume thatsuch perceptions represent an accurate assessment ofthe actual level of struggle between the two groups.Our first country-level measure of competitiveness istherefore the proportion of Muslims in each countrywho respond affirmatively to this question. Second,we create a country-level measure of religiosity as theproportion of a countrys Muslim respondents who arehighly religious, as defined at the individual level. Val-ues of this variable range from 36% in Turkey to 90%in Kuwait.

We obtained a measure of the share of each coun-trys population that is Muslim from the Pew ResearchCenter (2009) Report on the Size and Distribution ofthe Worlds Muslim Population which is a study con-ducted independently from the Pew GAP survey. Mus-lim population shares range from 7% in Kenya to 99%in Morocco.

A Hierarchical Model of Anti-Americanism

To test both the individual- and country-level im-plications of our theory, we model responses tothe dependent variableattitude toward the UnitedStatesusing a Bayesian hierarchical model con-taining individual-level and country-level components(Gelman and Hill 2007). The structure of the surveydata places individual respondents within countries.The hierarchical model simultaneously estimates theeffects of individual-level predictors of anti-American

8


9/19


10/19


TABLE 3. Hierarchical Linear Model Coefficient Estimates and Posterior Standard Deviations

Individual-level effects Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Piety 0.11 (0.03) 0.11 (0.03) 0.11 (0.03) 0.10 (0.03)Identify with Islamists 0.17 (0.04) 0.17 (0.03)Follow international news 0.04 (0.01) 0.03 (0.01) 0.04 (0.01) 0.04 (0.01) 0.04 (0.01) 0.04 (0.01)Age/100 0.11 (0.04) 0.10 (0.04)Male 0.00 (0.01) 0.00 (0.01)Income level 0.05 (0.01) 0.05 (0.02)

Secondary education 0.04 (0.02) 0.04 (0.02)University education 0.06 (0.02) 0.06 (0.02)Policy satisfaction 0.12 (0.03) 0.12 (0.03)

Country-level effects

Constant 0.25 (0.18) 0.42 (0.19) 0.13 (0.21) 0.28 (0.22) 0.18 (0.16) 0.34 (0.20)ReformerIslamist struggle 0.52 (0.21) 0.47 (0.21) 0.79 (0.18)Total religiosity 0.49 (0.16) 0.43 (0.19) 0.58 (0.18)Proportion Muslim 0.27 (0.12) 0.27 (0.11) 0.29 (0.13) 0.28 (0.12) 0.16 (0.10) 0.17 (0.12)GDP per capita, log 0.03 (0.03) 0.02 (0.03) 0.01 (0.04) 0.03 (0.03) 0.01 (0.03) 0.06 (0.03)

0.24 0.24 0.23 0.23 0.23 0.230 0.13 0.12 0.15 0.14 0.10 0.121 0.12 0.12 0.15 0.15 0.11 0.122 0.04 0.04 0.04 0.04 0.04 0.04Notes: The dependent variable is an individuals level of anti-Americanism on a 01 scale. Positive coefficients indicate variables that

have an increasing effect on levels of anti-Americanism. Dataset includes 12,831 respondents in 21 countries. Coefficients on theindividual-level variables are the average effects k across countries.

21 study countries.15 The combined effect of religios-ity and high media attentiveness, by adding 1 + 2,is approximately 0.15 points in a typical country. Insubstantive terms, this individual-level effect is nearlyone-quarter of the total range of anti-Americanism ob-served across countries (0.23 to 0.84). When we mea-sure Muslims political predispositions by their alle-giance with Islamists or modernizers, the estimatedindividual-level effect on anti-Americanism is even

largerfully 0.17 points in a typical country (Models 3and 4). Again, however, this finding is based on the sub-set of respondents who believed there to be a strug-gle in their country between fundamentalists andmodernizers.

At the country level, the effect of secularIslamist conflict on levels of anti-Americanism is evenmore pronounced. Anti-Americanism is much morewidespread in countries with higher perceived levelsof struggle between secular and Islamist elites, as wellas in countries with lower overall levels of religios-ity among the Muslim population (Figure 2). Recallthat as countries converge on 100% religious types,there is less intense secularIslamist competition. In

Model 1, we estimate that an increase of just 10% inthe proportion perceiving a reformerIslamist struggleis associated with an additional 0.05 points of anti-Americanism at the country level. Across the entirerange of this variable, our model predicts a differenceof 0.3 points on the 01 anti-Americanism scale. Themagnitude of the effect of overall levels of religiosityon anti-Americanism (Model 2) is substantively justas large. These results are consistent with our asser-

15 The exceptions are Mali, Morocco, and Indonesia, where the esti-mated effects are again indistinguishable from zero.

tion that anti-Americanism in the Muslim world has aninstrumental basis and is closely related to countriesdomestic politics. They also provide an explanation forthe low levels of anti-Americanism observed amongMuslims in Ethiopia and West Africa and are robust toalternative explanations discussed later.

The models further indicate that anti-Americanismis more widespread among Muslims in countries whereMuslims comprise a larger share of the population.Where Muslims represent a smaller proportion of thepopulation, anti-Americanism is lower, on average.The relationship between competitiveness and levelsof anti-Americanism does not seem to be a spuriousassociation due to some countries having larger sharesof Muslims than others. The estimated effect of thisvariable is of similar size to that of competitivenessacross its observed rangeapproximately 0.25 points onthe anti-Americanism scale. Finally, a countrys level ofper capita GDP does not confound either of these twopredicted effects. There is no consistently positive ornegative estimated effect of wealth on countries levelof anti-Americanism.

To more fully interpret the results of Models 1 and 2,

we compare the predicted level of anti-Americanismfor nonreligious, low-media-aware individuals to thatof highly-religious, high-media-aware individuals indifferent types of countries. We vary the percentageof Muslims in a country who see a reformerIslamiststruggle from 10% to 75%, which is slightly greaterthan the observed range in our study. We then considerhypothetical countries in which the Muslim populationshare is near its minimum (10%), mean (65%), andmaximum (100%). Because logged per capita GDPhas a relatively small effect, we hold it fixed at its meanvalue, approximately $1,000. In the first group, x1 = 0andx2 = 0, so at each combination of the country-level

10


11/19


FIGURE 2. Average Level of Muslim Anti-Americanism Compared to Two Measures of the Intensityof SecularIslamist Political Competition: Proportion of Muslims Seeing a Struggle between GroupsWho Want to Modernize the Country and Islamic Fundamentalists (left); and Proportion of MuslimsWho Report a High Level of Religious Piety (right)

0.0 0.2 0.4 0.6 0.8

0.2

0.4

0.6

0.8

1.0

Proportion seeing a reformerfundamentalist struggle

AverageunfavorabilityofUnitedStates

BangladeshEgypt

EthiopiaIndonesia

Ivory Coast

Jordan

Kuwait

Lebanon

Malaysia

Mali

Morocco

Nigeria

Pakistan

Palestinian ter.

Senegal

Tanzania

Turkey

0.2 0.4 0.6 0.8 1.0

0.2

0.4

0.6

0.8

1.0

Proportion highly religious

AverageunfavorabilityofUnitedStates

BangladeshEgypt

EthiopiaIndonesia

Ivory Coast

Jordan

Kuwait

Lebanon

Malaysia

Mali

Nigeria

Pakistan

Palestinian ter.

Senegal

Tanzania

Turkey

covariates, the expected level of anti-Americanism isy = 0 + 1z1 + 2z2 + 3z3. In the second group, x1 =1andx2 = 1, so the expected level of anti-Americanismis y = ( 0 + 1 + 2)+ 1z1 + 2z2 + 3z3, where 1and 2 are the estimated means of the random effects1j and 2j. Anti-American attitudes are most preva-

lent where reformers and Islamists are most engaged instruggle, and in countries that are predominantly Mus-lim (Figure 3a). Recalculating the model predictionsusing a countrys level of religiosity as the measure ofcompetitiveness (Model 2) produces the same result(Figure 3b).

As the perceived level of competition between secu-lar and Islamist groups increases, so does the preva-lence of unfavorable attitudes toward the UnitedStates. From left to right, the combined effects ofreformerIslamist struggle and Muslim populationshare can explain variation of 60% of the entire pos-sible range of anti-Americanism (01). Adding in theindividual-level effects, as shown, raises this to 75%.

Because high overall levels of religiosity are associ-ated withlessintense secularIslamistcompetition,Fig-ure 3b illustrates exactly how more religious countriesare less anti-American even though more religious in-dividuals are more anti-American.

We now consider a series of individual-level de-mographic and attitudinal factors that might also ex-plain anti-American attitudes and could confoundthe estimated effects of religiosity and media atten-tiveness. Indeed, there are many predictors of anti-Americanism at the individual level that matter in ad-ditionto theones identifiedby our theory (e.g., Chiozza2009; Gentzkow and Shapiro 2004; Nisbet and Myers

2011)and it is probable that the effects of these vari-ables vary by country. We therefore enter into Models 1and 2 measures of respondents age, gender, income,education level, and policy satisfaction.16 The resultsappear in Table 3 as Model 5 and Model 6. We find thatin most countries, age is associated with greater anti-

Americanism and that greater wealth, education, andsatisfaction with politics are all associated with lesseranti-Americanism. Yet none of these effects diminishthe estimated effects of religiosity or media aware-ness.17

To what extent is it possible that a countrys levelof reformerIslamist struggle is caused by aggregatelevels of anti-Americanism and not vice versa? Mostscholars point to the long-term historical origins of thesplit withinthe Muslim world over optimal regimetype.Owen (2010, 238), for example, describes the strugglebetween promoters of Islamism and secularism as hav-ing a deep structural cause with its origins in the late19th andearly 20th centuries. If a countrys level of con-

flict on this dimension has a long-term historical basis,

16 The demographic variables are GAP questions 107, 108, 118, and123. We rescale income from zero (minimum) to one (maximum)within each country. We measure education as a trichotomy for lessthansecondarylevel, at leastsecondary level,and someuniversity ed-ucation. Policy evaluationsare measured as the respondents averagesatisfaction with the direction of the country (Q7), perception of thecurrent economic situation (Q11), assessment of whether childrenwill grow up better off (Q13), and evaluation of the performance ofthe national government (Q23a).17 Our estimates of the individual-level effects of age and education,in particular, are consistent with the findings of Nisbet and Myers(2011).

11


12/19


FIGURE 3. Predicted Levels of Anti-Americanism by National Context: Percentage of a CountrysPopulation that is Muslim and (a) Level of Conflict between Secular and Islamist Elites or(b) Proportion Highly Religious

(a) Effects of secularIslamist competition

Muslims 10% of population

Proportion seeing secularIslamist struggle

LevelofAntiAmericanism

0.1 0.2 0.3 0.4 0.5 0.6 0.7

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1




0.1 0.2 0.3 0.4 0.5 0.6 0.7

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Highly religious,high media awareness

Not highly religious,low media awareness




0.1 0.2 0.3 0.4 0.5 0.6 0.7

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

(b) Effects of levels of religiosity




0.4 0.5 0.6 0.7 0.8 0.9

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1




0.4 0.5 0.6 0.7 0.8 0.9

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Highly religious,high media awareness

Not highly religious,low media awareness




0.4 0.5 0.6 0.7 0.8 0.9

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

reverse causality is unlikely. This suggests that anti-Americanism has been epiphenomenal to ideologicalpolarization on the mosque-state dimension.

Testing Alternative Explanations

The theoretical argument and empirical findings thusfar focus on how the domestic political environmentserves as a primary driver of anti-American sentiment

in the Muslim world. This section considers alterna-tive contextual explanations for anti-Americanism anddescribes the results of a series of empirical tests thatoperationalize these alternative hypotheses.

Does Muslim anti-Americanism vary with greaterU.S. cultural and economic presence in a country? Onthe one hand, higher levels of cultural and economic in-teraction could increase feelings of connectedness andfellowship between the two societies (Chiozza 2009);on the other hand, American business practices andcultural products might be seen as invasive or exploita-tive. We thus enter into the model a measure of the di-rect exposure of Muslims in each country to the United

States, calculated as the (logged) percapita dollar valueof a countrys imports of goods from the United Statesin 2007 (Table 4).18 In our sample, wealthier coun-tries import greater amounts of merchandise from theUnited States,and thisvariable is highly correlated withlevels of per capita GDP. As with wealth in Model 1,we find no consistent effect of U.S. imports on overalllevels of anti-Americanism (Model 7). Nor does thisvariable confound the effects of secularMuslim con-flict or Muslim population share.19

A second possible factor that might increase or de-crease anti-Americanism is the amount of U.S. for-eign economic aid received by a country. Poorercountries, such as those in sub-Saharan Africa, re-ceive significantly larger amounts of U.S. assistance,making citizens of those countries potentially more

18 Data on the total dollar value of imports from the United Statesin 2007 are available from the Foreign Trade Division, U.S. CensusBureau; http://tse.export.gov (accessed March 7, 2012).19 In each of Models 711, the results are similar if we use levelsof religiosity as the measure of elite competition rather than theperceived extent of struggle.

12


13/19


TABLE 4. Tests of Alternative Country-Level Hypotheses and Robustness Checks for Model 1 inTable 3

Individual-level effects Model 7 Model 8 Model 9 Model 10 Model 11

Piety 0.11 (0.03) 0.11 (0.03) 0.12 (0.03) 0.11 (0.03) 0.11 (0.03)Follow international news 0.04 (0.01) 0.04 (0.01) 0.04 (0.01) 0.04 (0.01) 0.04 (0.01)

Country-level effects

Constant 0.13 (0.11) 0.05 (0.11) 0.17 (0.13) 0.12 (0.11) 0.06 (0.14)

ReformerIslamist struggle 0.45 (0.20) 0.48 (0.19) 0.41 (0.20) 0.42 (0.19) 0.44 (0.17)Proportion Muslim 0.25 (0.11) 0.25 (0.10) 0.23 (0.10) 0.23 (0.11) 0.24 (0.11)U.S. imports per capita, log 0.01 (0.02)U.S. foreign economic aid 0.17 (0.16)Proximity to Jerusalem 0.02 (0.02)Military expenditure 0.00 (0.02)Level of political rights 0.02 (0.02)

0.24 0.24 0.24 0.24 0.240 0.14 0.13 0.13 0.13 0.131 0.11 0.12 0.12 0.11 0.122 0.04 0.04 0.04 0.04 0.04Notes: Positive coefficients indicate variables that have an increasing effect on levels of anti-Americanism, on a 01 scale. Thesurvey sample contains 12,831 respondents in 21 countries.

appreciativeor, at least, more sympathetictowardthe United States, leading to more favorable attitudes.However, Muslims might also see foreign aid from theUnited States as an affront to a countrys sovereignty.We test the effects of this variable in Model 8.20 Af-ter domestic sources of anti-Americanism have beenaccounted for, levels of U.S. foreign aid have no sig-nificant effect on opposition to the United States. Theestimated effects of domestic competition and Muslimpopulation share are robust in this model specification.

Commitment to a just resolution of the Arab-Israeliconflict remains important to many Muslims, partic-ularly given the religious and cultural importance of

Jerusalem within the Islamic faith. The involvementof the United States as a mediator in this conflictmight be expected to increase anti-American senti-ments. Because the conflict may be more salient forcitizens of frontlinestateslike Jordan,Syria, and Egypt,a countrys geographic distance to Jerusalem mightserve as a proxy for concern about this issue. Yet dis-tance to Jerusalem is not a significant predictor of anti-American sentiment (Model 9).21 This variable alsoserves as a control for the existence of an Arab orMiddle Eastern effect because many consider theLevant to be the geographic center of the MiddleEast.

We last consider the possible effects of mili-tarism and degree of democracy on levels of anti-Americanism. Muslims living in highly militarized soci-eties may be more aware of geostrategic considerationswhile simultaneously viewing their countries in more

20 U.S. foreign economic aid in 2007 is measured in historicaldollars in units of billions and is taken from the U.S. Over-seas Loans and Grants Obligations and Loan Authorizations(Greenbook) database, U.S. Agency for International Development;http://gbk.eads.usaidallnet.gov (accessed March 7, 2012).21 Proximity to the conflict in Israel is measured as the distance inthousands of miles from each countrys capital to Jerusalem.

direct competition with the United States for regionalor global influence. Yet we findno relationship betweencountries domestic military expenditures and levels ofanti-Americanism (Model 10).22 Similarly, it is possiblethat Muslims living in a country with a democraticpolitical culture more similar to that of the UnitedStates might have more affinity for America in gen-eral. Countries levels of political rights, however, arenot associated with levels of Muslim anti-Americanism(Model 11).23 In addition, in both Models 10 and 11,the estimated country-level effect of secularIslamistcompetition on anti-Americanism does not change.24

COMPARATIVE CASE STUDY EVIDENCE

We now turn to comparative case studies and mediacontent analysis of news reports and editorials aboutthe United States in three predominantlyMuslim coun-tries: Turkey, Morocco, and Senegal. To establish thatthe causal mechanism we have described is consistentwith the experiences of these countries, we assess both(1) the nature and intensity of secularIslamist polit-ical competition in each country and (2) the publicportrayal of the United States by local leaders andother political elites. In choosing these three countriesfor in-depth analysis, we follow the guidance of King,

Keohane, and Verba (1994, 140) that the best inten-tional design selects observations to ensure variation

22 Population data and military expenditures as a percent of GDPare obtained from the World Bank (2008) World Development In-dicators for the year 2007.23 Our measure of political rights is the 2007 Freedom House score.24 We investigated whether the number of active-duty U.S. troopsstationed in each country might be associated with levels of anti-Americanism, but this variable exhibited almost no cross-nationalvariation in our sample in 20067. According to the military person-nel statisticsof the U.S. Departmentof Defense InformationAnalysisDivision, no more than a few dozen U.S. troops were stationed incountries other than Egypt, Kuwait, and Turkey.

13


14/19


in the explanatory variable. . . without regard to thevalues of the dependent variable. Turkey and Senegalare among the most dissimilar countries in the Mus-lim world with respect to their level of contestationbetween secular and Islamist political groups: Turkeyis highly competitive along that dimension, whereasSenegal is not (Figure 2). In the 2007 GAP study,71% of Turkish Muslims perceived a struggle between

modernizersandIslamic fundamentalists, whereas only36% were highly religiously observant. In Senegal, bycontrast, only 14% of Muslims observed a reformerfundamentalist struggle, and 83% were highly religious.Morocco is an intermediate case. The population of allthree countries is nearly entirely Muslim.

We expect that depictions of the United States in theTurkish press will be overwhelmingly negative, whereasthe United States will receive more balanced treatmentin the Moroccan press and relatively positive treat-ment in the Senegalese mass media. Preliminarily, weobserve that, in line with this expectation, levels of anti-American sentiment are extremely high in Turkey andquite low in Senegal, with Morocco falling in between

(Figure 1). We do not deny that mass opinion about theUnited States affects elite opinion and is affected by itin return. That said, if our analysis did not find eliteopinion patterns in the manner we have described, wewould count this as evidence against our theory.

Senegal

The vast majority of Senegalese are highly religious,identifying with one of a handful of Islamic Sufi broth-erhoods that dominate Senegalese associational life(Clark 1999).25 In recent years, an Islamist movementhas emerged that calls for the adoption of Islamic law

in Senegal (Loimeier 1996), but its influence has re-mained fairly limited.Like many African countries that gained indepen-

dence in the 1950s and 1960s, Senegal inherited a sec-ular state structure after its period of colonial rule byFrance. The governing apparatus in the newly inde-pendent Senegalese state came to be dominated bya narrow elite of Francophone technocrats and in-tellectuals. Nominally Muslim but secular in outlookand orientation, this elite was bound together by whatOBrien (2003, 53) calls the powerful secularizinginfluence of French language, culture, and education.A university educationin Frenchbecame a prerequisitefor employment in the state bureaucracy, even though

less than 20% of the population spoke French withany degree of fluency and less than 1% used Frenchexclusively.

The system of political compromise that emergedin Senegal brought together the secular Francophoneelite with the leaders of the Sufi brotherhoods. Recog-nizing the need for an intermediary between the stateand the masses, the secular elite worked closely withthese Islamic leaders (e.g., Haynes 1996; Villal on 1995),

25 Sufism is generally described as mystical Islamic belief and prac-tice.

frequently demonstrating their acquiescence and, insome cases, submission (Dieye 2009). The leaders ofthe Sufi orders, by allowing the secular elite to run thestate, created a system of coexistence for the religiousbrotherhoods that obviated the need for political com-petition within the set of Sufi groups. Yet tension stillexists between the secular elite and the Sufi Broth-erhoods. Forms of symbolic confrontation (OBrien

2003) over issues like family law and the inclusion ofreligious instruction in education continue to demon-strate the relevance of the secularreligious conflict inSenegalese daily life. Despite this conflict, the citizensof Senegal are overwhelmingly pious Muslims livingin a country whose political system is dominated by anarrow, secular bureaucratic elite.

Internationally, Senegal has cultivated and currentlyenjoysaverystrongrelationshipwiththeUnitedStates.Over the past decade, numerous American politiciansand dignitaries have visited Senegal and have beenreceived favorably. There is little to be gained in thecontext of Senegalese politics by criticizing the UnitedStates; as such, Muslims in Senegal are among the most

pro-American in the world.

Turkey

Mainstream anti-Americanism in Turkey first emergedduring the Cold War, but was largely confined to theTurkish left, who strongly opposed the nature and ex-tent of U.S. involvement in Turkish political affairs. Inthe 1960s and1970s, the Turkish right aligned itself withthe United States against international communism(Criss 2002). With the decline of Cold War bipolar-ity, however, the salience of the religioussecular issue

dimension intensified in Turkey, starting in the 1980sand continuing through the 2000s (Taspinar 2005).According to Hale (2002, 178), the political poles

in contemporary Turkey are Kemalist secularism andpolitical Islamism, and this cleavage has supersededthe traditional leftright socioeconomic divide. Usingpublic opinion data from the World Values Survey,Kalaycioglu (1999) found that in the 1990s individu-als religiosity corresponded closely with party prefer-ences, whereas social class and economic satisfactionmattered little. The intensity of the religioussecularcleavage in Turkey is reflected not only in Turkeysformal party organizations but also in the substance ofhighly prominent national debates about the role of re-

ligion in Turkish public and political lifefor example,concerning policies prohibiting women from attendingschools or universities while wearing the Islamic head-scarf.

Anti-Americanism in Turkey is embraced by nearlyall segments of Turkish society, andbothsecular nation-alists and Islamists engage in stridently anti-Americanrhetoric (Guney 2008; Taspinar 2005). Pollock (2005)describes anti-Americanism in Turkey as a combi-nation of old leftism and new Islamism where justabout every politician and media outlet (secular andre-ligious) preaches an extreme combination of America-and Jew-hatred that . . . voluntarily goes far further

14


15/19


than anything found in most of the Arab world. Is-lamists, such as the elite associated with the Justice andDevelopment Party, have been relentlessly and pub-licly negative in their portrayal of the United States(Cagaptay 2008). Likewise, anti-imperialist rhetoric isa main theme of secular nationalists, who argue thatTurkey is under a lethal threat from both the UnitedStates and religious Muslims (Akyol 2008). Secularist

rallies, some of which draw millions of supporters, fre-quently feature demonstrators carrying anti-Americanplacards (Somer 2007; Zaman 2007). Secular nation-alist intellectuals also argue that American policies inthe Middle East are part of a neo-colonial effortto establish hegemony in the region: For example, aU.S. bombing in Iraq triggered a major earthquake inTurkey in 2003, the United States is keeping Turkey outof Iraq to ensure Turkey is not able to exploit Iraqs oilresources, and Osama bin Laden and Saddam Husseinwere both on the American payroll (Cagaptay 2004).26

Pollock (2005) similarly relates that the Islamist news-paper Yeni Safak, and the mainstream secular paperH urriyet, are both rife with conspiracy theories, in-

cluding how U.S. forces in Iraq have been harvestingthe organs of dead Iraqis for sale in the United Statesand how secret American nuclear testing was actuallyresponsible for the Southeast Asian tsunami.

Morocco

Religious and secular political organizations exist si-multaneously in contemporary Morocco. The domi-nant religiously oriented groups include both militantand moderate Islamist groups as well as a broad-basedSufi movement that mirrors the vocabulary and struc-tures of Moroccos monarchy (Zeghal 2008, xix). Sec-

ular Moroccansmany of whom graduated from aparallel Francophone educational system and are as-sociated with left-leaning political organizationsareinfluential but fewer in number than Moroccans with amore religious worldview (Zeghal 2008, 61;80).

The relative balance between secular and religioustrends has led scholars to argue that there exists aMuslim consensus in Morocco in which Islam an-imates the nations spiritual life and anchors its so-cial existence (Entelis 1989, 1112). At least threesignificant political groups exist within this consensus:moderate Islamists, radical Islamists, and individualsassociated with local Sufi movements (Zeghal 2008,xix). Supporters of a moderate interpretation of Islamenjoy a near majority, as evidenced by public opinionpolling (Sater 2010, 1). Radical Islamists disagree withtheir more moderate counterparts primarily on tacticalpolitical matters. Islamist activists of both types havebeen sharply critical of Western-influenced elites whoare believed to be steering the country toward secular

26 Turkeys media environment has evolved in the last 20 years, af-fecting how elites influence public opinion. Although in the past, theTurkish citizenry was largely rural with a single television channeland few media outlets, increasingly Turkish citizens receive a varietyof elite perspectives (Taspinar 2005).

values and Westernization that distance it from Araband Islamic roots (Shahin 1994, 169).27

In Morocco, there exists a complexand some ar-gue uniquely Moroccanrelationship between reli-gion and political authority.28 Moroccos king, whoclaims to be descended from the prophet Mohammed,is the commander of the faithful, and although stateinstitutions have a secular quality, the monarchy it-

self remains a key religious institution in the country(Munson1993, 121).Thus, although the Moroccanstatemay be viewed as not sufficiently religious by the Is-lamist right, supporters of the monarchy associate theking with religious authority.

Although the majority of Moroccans have convergedin their support of religion in one form or another,there nonetheless exists a politically influential secular-minded minority of the Moroccan political elite. In ad-dition, thereappearsto be significant social distancebe-tween religious and secular types in Moroccan society.Secular forces have historically identified with leftistpolitical organizations that have roots in the Arab so-cialist tradition (Zeghal 2008, 63). Moroccan socialist

parties have typically drawn support from organizedlaborers, urban migrants, bureaucrats, and universitystudents (Waterbury 1970, 1967), though are increas-ingly losing ground to Islamist political organizationson university campuses and in urban areas (Boukhars2011, 92). Yet despite a decline in popularity of thesecular left, elites associated with this perspective con-tinue to enjoy considerable influence both in media andgovernment.

How do these political cleavages relate to the in-tensity of anti-American sentiment? According to onejournalist, anti-Americanism in Morocco is espousedby everyone from Islamist traditionalists to urban so-

phisticates (Charney 2005). This trend is evidencedby both large anti-U.S. protests organized by Islamicactivists and by strong anti-American sentiment ex-pressed by secular elites. In this context, Moroccansechotheir medias sentiments about the United States,suggesting elites have a particularly important role toplay in influencing public opinion (Charney 2005). Theexistence of a broad, religious Muslim consensus witha politically significant secularist presence in Moroccoshouldaccording to our theorytranslate into lowerlevels of anti-American than observed in Turkey buthigher levels than seen in Senegal.

Comparative Media Content Analysis

We examined the balance between positive and nega-tive portrayals of the United States in two leading na-tional newspapers in each study country over a periodof two to five months in advance of the 2007 GAP sur-vey. This interval was long enough to produce samplescontaining approximately 50 to 100 articles from each

27 This position is seen clearly in the rhetoric of influential ShaykhAbd al-Salam Yassin who has argued that the Moroccan nation istorn between two worlds, the religious and the secular (Shahin1994, 170).28 See Munson (1993, 53) for one perspective.

15


16/19


FIGURE 4. Balance in the Tone of MediaMentions (Positive versus Negative) of theUnited States in Political and Economic Affairsin Senegal, Morocco, and Turkey

Senegal Morocco Turkey

Positive Neutral Negative

Percentofartic

les

0

10

20

30

40

50

60

70

46

33

2125

43

32

15

28

57

of the six sources.29 We chose the newspapers basedon the size of their circulation and the condition thatthey be domestically produced. Each one was majorenough to offer fully searchable online archives for theperiod under study. In Turkey, we examined Zaman, amoderate-conservative Islamist daily that is generallysympathetic to the policies of the AKP, and its pri-mary competitor, the Kemalist-secularist H urriyet. InMorocco, we searched Aujourdhui le Maroc and LeMatin du Sahara et du Maghreb; both widely circulatedFrancophone dailies. In Senegal, we examined issues ofthe government-owned Le Soleiland the independentSud Quotidien, the two largest Francophone dailies in2007 (Banks, Muller, and Overstreet 2007).

For articles discussing the role of the United States ineither international or domestic political or economic

affairs, we recorded whether the tone of the article waspredominantly positive, negative, or neutral. These ar-ticles included both straight news items and editorials.The most common positive mentions of the UnitedStates referred to American-backed development pro-grams, trade, or the role of the United States in in-ternational diplomacy. Negative mentions tended tofocus on objections surrounding the wars in Iraq andAfghanistan or other criticisms of U.S. foreign policy.We coded as neutral those cases in which Americawas referenced in neither a critical nor laudatory man-ner (for example, announcements of meetings betweenSenegalese leaders and U.S. officials).30

We tabulated the percentage of news articles in eachcountry describing the United States in a positive, neg-ative, or neutral manner (Figure 4). In Turkey, wheresecularIslamistcompetition is most intense,more thanhalf of the articles referenced the United States in anunfavorable manner. The tone of articles in the mod-erate H urriyet (61% negative) was more critical than

29 The study period for each country was Turkey: January 1March 31, 2007 (151 articles); Morocco: April 1May 31, 2007 (182articles); and Senegal: January 1May 31, 2007 (109 articles).30 A complete coding protocol and dataset, including the names,dates, topics, and perspectives of coded articles, are available fromthe authors on request.

those in Zaman (52%). The opposite was the case inSenegal, where nearly half of the articles portrayedthe United States in a favorable light. The content ofLe Soleil, which we might expect to be more closelyattuned to elite attitudes, was far more positive to-ward the United States (54%) than the independentSud Quotidien (34%). Although this difference in toneshould affect mass attitudesafter all, nearly one-third

of Senegalese Muslims do hold an unfavorable opinionof the United Statesit also matches our expectationthat Senegalese political elites are not predominantlyanti-American in their public pronouncements. Finally,we found a balanced set of perspectives toward theUnited States in the major Moroccan press, with a tiltin the anti-American direction. Morocco is situatedbetween Senegal and Turkey, with little difference intone between Le Matin and Aujourdhui le Maroc.Consistent with our expectations, in countries with lowsecularIslamist competition, people are exposed toone type of media environment, and in places withhigh competition people are exposed to a very differ-ent type of media environment. Although our media

analysis does not address the issue of causal direction,a different pattern of evidence might have falsified ourtheory.

CONCLUSION

Negative perceptions of the United States arewidespread in the Islamic world, but they are not uni-versal. We have argued that levels of Muslim opposi-tion to the United States are associated with the degreeof domestic political competition in a given countrybetween secular and religious groups. As competition

intensifies, it becomes increasingly advantageous forelites to foment anti-American sentiment for their ownpolitical gain. The outcome of this elite-led process iswhat we contend survey researchers are detecting, atleast in part, when they ask individual Muslims theiropinion of the United States.

The intensity of political competition alongreligioussecular linesthe key explanatory vari-able that we proposeexplains a substantively largeamount of the cross-country variation in anti-American sentiment across the Islamic world. A me-dia content analysis of the balance of pro- and anti-American messages in three Muslim countries withvarying levelsof competition providesfurther evidence

in support of the country-level mechanism we describe.At the individual level, we find consistent and robusteffects of religiosity and media attentiveness on Muslimanti-Americanism across a wide range of countries.

Explaining why many Muslims dislike America canprovide insight into the utility of the policy optionsavailable to the United States in mitigating Americanunpopularity abroad. Previous scholarly work has sug-gested that, to the extent anti-Americanism is basedon attributes of America and Americans, it will beless likely to moderate over time, because these con-cerns are based on deep-seated attitudesin contrastto concerns about American policies that may be more

16


17/19


mutable (Thornton 1988, 13). As the Muslim reactionto the events of 2003 makes clear, the actions of theUnited States in the Islamic world do affect percep-tions of and support for the United States as a globalactor,even if only for a short time. At the very least, theUnited Statesthroughits positionin the worldsystemand interventionism overseasprovides a generalizedenvironment of grievance that allows for political mo-

bilization against America to take place.Yet it is unclear how far a more balanced approachto American foreign policy making would go towarderadicating anti-Americanism in the Islamic world.Our results indicate that to the extent Muslim anti-Americanism is a domestic phenomenon, a certain de-gree of pessimism is warranted toward the potential ofAmerican actions to lessen negative perceptions of theUnited States in the Islamic world. Foreign elites con-tinue to have their own motivations for promoting neg-ative views about the United States, which are relatedto incentives surrounding local political mobilization.A core assumption made by advocates of an enhancedpublic diplomacy campaign is that anti-Americanism

stems from poor strategic communication on thepart of the United States (Nisbet and Shanahan 2008).The results of this study suggest that Muslim publicsare highly responsive to messages from domestic elitesand that any American-led effort to offer a counter-narrative would have to compete with local mediaenvironments that may be firmly oriented toward theinstrumental advancement of anti-American attitudes.

REFERENCES

Abdallah, Abdel Mahdi. 2003. Causes of Anti-Americanism in theArab World: A Socio-Political Perspective. Middle East Reviewof International Affairs 7 (4): 6273.

Ajami, Fouad. 1992. The Arab Predicament: Arab Political Thoughtand Practice since 1967. Cambridge: Cambridge University Press.

Ajami, Fouad. 2001. What the Muslim World is Watching. NewYork Times Magazine, November 18.

Ajami, Fouad. 2003. The Falseness of Anti-Americanism. ForeignPolicy (September/October): 5261.

Akyol, Mustafa. 2008. The Kemalist Crusade against Imperialism(aka the EU). Turkish Daily News, May 10.

Ansolabehere, Stephen, Jonathan Rodden, and James M. Snyder.2008. The Strength of Issues: Using Multiple Measures to GaugePreference Stability, Ideological Constraint, and Issue Voting.

American Political Science Review 102 (2): 21532.Banks, Arthur S., Thomas C. Muller, and William R. Overstreet,

eds. 2007. Political Handbook of the World. Washington, DC: CQPress.

Boukhars, Anouar. 2011. Politics in Morocco: Executive Monarchy

and Enlightened Authoritarianism. London: Routledge.Bowen, John R. 2007. Anti-Americanism as Schemas and Diacriticsin France and Indonesia. In Anti-Americanisms in World Politics,eds. Peter Katzenstein and Robert Keohane. Ithaca, NY: CornellUniversity Press, 22750.

Browers, Michaelle. 2009. Political Ideology in the Arab World: Ac-commodation and Transformation. Cambridge: Cambridge Uni-versity Press.

Cagaptay, Soner. 2004. Where Goes the U.S.-Turkish Relation-ship? Middle East Quarterly 11 (4): 4352.

Cagaptay, Soner. 2008. The Most Anti-American Nation.Newsweek, November 15.

Charney, Craig. 2005. Morocco: The Price of Anti-Americanism.Salon.com, January 7.

Chiozza, Giacomo. 2007. Disaggregating Anti-Americanism: AnAnalysis of Individual Attitudes toward the United States.

In Anti-Americanisms in World Politics, eds. Peter Katzensteinand Robert Keohane. Ithaca, NY: Cornell University Press, 93126.

Chiozza, Giacomo.2009.Anti-Americanismand the American WorldOrder. Baltimore: Johns Hopkins University Press.

Clark, Andrew F. 1999. Imperialism, Independence, and Islam inSenegal and Mali. Africa Today 46 (3/4): 14967.

Cole, Juan. 2006. Anti-Americanism: Its the Policies. AmericanHistorical Review Forum 111 (4): 112029.

Criss, Nur Bilge. 2002. A Short History of Anti-Americanism and

Terrorism: The Turkish Case.Journal of American History 89 (2):47284.

Damore, David F. 2004. The Dynamics of Issue Ownership inPresidential Campaigns. Political Research Quarterly 57 (3): 39197.

Damore, David F. 2005. Issue Convergence in Presidential Cam-paigns. Political Behavior27 (1): 7197.

Dieye, Abdoulaye. 2009. Secularism in Senegal: Withstanding theChallenge of Local Realities. Evanston, IL: Institute for the Studyof Islamic Thought in Africa, Buffett Center, Northwestern Uni-versity.

Doran, Michael. 2002. The Pragmatic Fanaticism of al Qaeda: AnAnatomy of Extremism in Middle Eastern Politics. Political Sci-ence Quarterly 117 (2): 17790.

Druckman, James N., and Arthur Lupia. 2000. Preference Forma-tion. Annual Review of Political Science 3: 124.

Entelis, John. 1989. Culture and Counterculture in Moroccan Politics.Boulder, CO: Westview.Esposito, John, and Dalia Mogahed. 2008. Who Speaks for Is-

lam? What a Billion Muslims Really Think. New York: GallupPress.

Faath, Sigrid, and Hanspeter Mattes. 2006. Between Anti-Americanism, Criticism of America and Americanism. In Anti-

Americanismin the Islamic World, ed. SigridFaath.London: Hurstand Company, 21938.

Finke, Roger, and Rodney Stark. 1988. Religious Economies andSacred Canopies: Religious Mobilization in American Cities,1906. American Sociological Review 53 (1): 419.

Fuller, Graham E. 2002. The Future of Political Islam. ForeignAffairs 81 (2): 4860.

Gabel, Matthew, and Kenneth Scheve. 2007. MixedMessages: PartyDissent and Mass Opinion on European Integration. EuropeanUnion Politics 8 (1): 3759.

Gallup, Inc. 2008. Global V

Documents

Elite Competition, Religiosity, and Anti-Americanism in the Islamic World