You are on page 1of 12

The Social Science Journal 49 (2012) 127138

Contents lists available at SciVerse ScienceDirect

The Social Science Journal


journal homepage: www.elsevier.com/locate/soscij

Minority comparison model: Effects of Whites multiracial evaluation on symbolic racism and racialized policy preferences
Young Min Baek a,b, , Angela M. Lee a,c
a b c

University of Pennsylvania, USA Graduate School of Convergence Science and Technology, Seoul National University, South Korea School of Journalism at University of Texas at Austin, USA

a r t i c l e

i n f o

a b s t r a c t
Most Racial Studies primarily focus on African Americans without paying attention to nonblack minorities, and it fails to capture recent increase in racial diversity. Based on previous theories and empirical ndings, we propose a new model, minority comparison model, which accounts for theoretical shortcomings in Racial Studies. This model (1) captures psychological processes that compare blacks and nonblacks, and (2) explores the effects of whites multiracial evaluation on racialized policy preferences. Drawing on a 2008 national representative sample, this study nds that whites who have positive stereotype of nonblacks (e.g., Hispanics and/or Asians) but negative stereotype of blacks show substantially higher symbolic racism and stronger opposition to Afrmative Action, whereas whites who have positive stereotype of blacks but negative stereotypes of nonblacks have stronger opposition to expansive immigration policy. Our study offers new ways of understanding and accounting for symbolic racism in modern context, and shows how whites preferences in racialized policies are inuenced by multiracial evaluation. 2011 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

Article history: Received 23 May 2011 Received in revised form 2 September 2011 Accepted 29 September 2011 Available online 19 June 2012

1. Introduction Most studies of race in America focus exclusively on African Americans (blacks) and Caucasians (whites). While this is understandable for both historical and demographic reasons, the United States is no longer predominantly a nation of two colors, consisting of more people of different races, ethnicities, and cultural backgrounds than ever before. Consequently, existing theories in the eld need to move beyond dichotomies in their attempt to understand the role of race and ethnicity in American politics. With the steady increase of new immigrants from Asia and Latin America, African Americans are becoming only one of

Corresponding author at: Graduate School of Convergence Science and Technology, Seoul National University, 864-1, Iui-dong, Yeongtonggu, Suwon-si, Gyeonggi-do 443-270, South Korea. E-mail address: ymbaek@gmail.com (Y.M. Baek).

several minority groups who are compared to, or contrasted with, both each other and the white majority. Recently scholars of race have emphasized the importance of multiracial evaluation, dened as evaluation of racial/ethnic minorities in comparison to other races. For example, Asian Americans have long been situated in a racial position in relation to the blackwhite binary (Junn & Masuoka, 2008, p. 729), and Hispanic workers are racially sandwiched by employers, who place them somewhere between Black workers and White workers (Maldonado, 2006, p. 354, emphasis in original). In other words, a more racially diverse environment has demanded multiracial evaluation among white minds, which changes the dynamics of racial tension and also affects white citizens preferences in racialized politics (Kim, 1999, 2000). Building upon existing theories and ndings in the eld, and taking into account the multiracial nature of modern American society, we propose a new model that we term

0362-3319/$ see front matter 2011 Western Social Science Association. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.soscij.2011.09.006

128

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

the minority comparison model. Examining white Americans multiracial evaluations of both black and nonblack minorities, our model aims to explain whites symbolic racism (Kinder & Sanders, 1996) against blacks, as well as their preferences in black-oriented (i.e., afrmative action) and nonblack-oriented racialized policies (i.e., immigration policy) by performing secondary analysis of a 2008 national survey. To our knowledge, this is the rst study that investigates the inuence of multiracial evaluations on racial resentment and white racialized policy preferences. To demonstrate the advantage of multiracial evaluation, the minority comparison model connects theoretical insights from racial triangulation theory (Kim, 1999; Maldonado, 2006) to empirical ndings about the role of race in determining white Americans policy preferences and public opinion (Kinder & Sanders, 1996; McDonald, 2001). 1.1. Linking model minority discourse to empirical ndings in public opinion research Until fairly recently, blacks comprised the largest racial minority in the United States, making them an important and often decisive voting bloc. Because of their political importance, as well as past history of racial discrimination, blacks (and white opinions regarding them) have been studied extensively in political science literature. Nevertheless, with increases in the number and percentage of nonblack immigrants to the United States, political dynamics have changed over the past several decades. When we consider the American political system, numbers are crucial: the growth of nonblack minority groups present new challenges and opportunities for forming voting blocs with blacks and/or whites (Browning, Marshall, & Tabb, 1984; Sonenshein, 1997; Sonenshein & Pinkus, 2005). Moreover, when we consider quantitative research, numbers are also important in deciding whether a group is likely to be studied in part because studies of nonblack minorities such as Latinos, Asians, and Native Americans cannot warrant the same reliability as those of blacks (Entman & Rojecki, 2000, p. xxiv). For these and other reasons, most quantitative studies have focused on the binary racial relationship between white and black Americans, especially on how white perception or evaluation of blacks inuences whites preferences in candidates for elected ofce, or for racialized policies and issues such as afrmative action, social welfare, black poverty or urban crimes. Theories emanating from some critical racial scholars have been more advanced in conceptualizing multiracial relationships and their political effects on the formation, modication and/or consolidation of racial ideology in American politics. One promising approach is the racial discourse analysis of model minority (Guinier & Torres, 2002; Kim, 1999, emphasis added). Model minority discourse emphasizes the success stories of nonblack minorities, in particular Asian Americans. According to Kim (2000), model minority discourse produces a colorblind ideology that allows whites to maintain racial power not through the direct articulation of racial differences between whites and blacks, but rather by obscuring the operation of racial power, protecting it from challenge (p. 17). The core arguments of model minority discourse are

subtly racialized and contribute to the stability of existing racial order in American society. Simply put, the logic behind model minority arguments is summarized as follows: (1) Asians and blacks are both racial minorities; (2) Asians have the motivation to move up the social ladder and succeed with no help from anyone else (Petersen, 1966), but blacks have no such motivations; (3) Therefore, Asians success stories are evidence that American society is reasonably fair and open to all racial minorities. Model minority discourse is based on multiracial evaluation beyond black and white, but with major implications for both blacks and whites. Racially resentful whites can justify black poverty or even black inferiority by emphasizing blacks lack of motivation, and can do so not by comparing blacks with whites, but by comparing blacks with other nonblack minorities. Despite lack of quantitative studies that examine white multiracial perception and its policy implications, two empirical studies provide supportive evidence of Kims theory (1999). The rst is the theory of symbolic racism (Henry & Sears, 2002; Kinder & Sanders, 1996; Kinder & Sears, 1981). This theory provides an empirical clue that is telling of whites multiracial evaluation. While symbolic racism theorists do not consider multiracial evaluation by itself, two items on the symbolic racism scale provide important empirical cues as to when symbolic racism becomes more pronounced in blacknonblack minority comparison than in blackwhite comparison: Irish, Italian, Jewish, and many other minorities overcame prejudice and worked their way up. Blacks should do the same. (Emphasis added) Its really a matter of some people not trying hard enough; if blacks would only try harder they could be as well off as whites. (Emphasis added) As can be seen, the rst statement, as a measure of symbolic racism, compares blacks to nonblack minorities,1 and the second compares blacks to whites. Symbolic racism measures are empirically helpful in clarifying the difference between blackwhite and blacknonblack comparisons, as these two items are included in the American National Election Studies (University of Michigan), and thus demonstrate trends in nationally representative opinion. Fig. 1 shows the rate of white Americans agreement with both statements. Whites in general showed stronger agreement with the rst than the second statement, and this suggests that, at least since the mid-1980s, symbolic racism has become more pronounced when nonblack minorities were compared to blacks than when whites were compared to blacks.

1 The rst item compares one racial group (i.e., blacks) to other ethnic groups because Irish, Italian and Jewish are considered racially white in racial classication.

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138
4.5

129

4.03

4.01

3.99 3.88 3.9 3.93

3.75 3.56 3.47

3.76 3.66 3.57 3.42 3.36 3.45

3.5

3.55

1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 year of survey
vs. Other minorities vs. whites

Based on ndings introduced above, we argue white Americans have dual routes of racial evaluation: biracial and multiracial. Biracial evaluation compares blacks and whites, and multiracial evaluation compares blacks and nonblack minorities, such as Asians or Hispanics. While biracial evaluation dominated American society before the rise of multiracial and multiethnic immigration in the United States, multiracial evaluation has become more prominent over the past few decades. As racial evaluations move beyond black versus white, racial theories should also reect this change in American demographics. Building upon racial triangulation theory (Kim, 1999, 2000), we propose the minority comparison model. 1.2. Minority comparison model: multiracial evaluation among whites Minority comparison model aims to show that opinion toward racialized policies, such as afrmative action and immigration policy, is better explained by incorporating multiracial evaluations. The theoretical notion of racial triangulation articulated by Kim (1999) is useful as a conceptual framework for understanding the role of multiracial evaluation in racial phenomena. Racial triangulation, according to Kim (1999, p. 107), is a social psychological process for placing a racial group in relation to other racial groups in the eld of racial positions. While Kims studies (1999, 2000) mainly deal with Asian Americans, racial triangulation of Hispanics by white Americans is also observed in recent eld studies (e.g., Maldonado, 2006; Moss & Tilly, 2001). Racial triangulation of racial minorities serves to solidify existing racial hierarchies through two evaluation processes: racial valorization and civic ostracism. Racial valorization occurs when people evaluate a racial group in comparison to other racial groups. Model minority discourse is the perfect example of racial valorization (Kim, 1999; Tuan, 1998). Studies on racial and ethnic stereotypes among white Americans (Allport, 1954; Fiske, 1998) consistently demonstrate that Asians are stereotypically associated with diligence, competence and frugality, in contrast to blacks association with laziness, incompetence and poverty. Some whites valorize Asians or Hispanic Americans relative to blacks on work ethic and competence/ability, which solidies white supremacy over blacks and simultaneously suppresses black protest without provoking whiteblack racial tensions. Civic ostracism occurs when nonblack groups are evaluated or treated as foreigners who fail to assimilate into mainstream cultures, and are ostracized from society. Critical racial theorists have accumulated much historical evidence of dominant racial discourse targeting Asians and Hispanics. For example, Hispanics and Latinos are often not accepted as Americans because they are perceived as offspring of illegal immigrants from different ethnic cultures (Maldonado, 2006). Moreover, while Japanese Americans were forced into concentration camps during World War II, German Americans were not only free from physical connements, but also allowed to serve in the U.S. army (Tuan, 1998). Historical examples of ostracization of nonblack minorities indicate that Hispanics and Asians

Fig. 1. Opinion trends in white explanations of how blacks should overcome black poverty. Note: only self-identied white respondents are selected. A higher number denotes stronger agreement with the given statements (1 = strongly disagree, 5 = strongly agree). The rst question item mentions many other minorities (i.e., Irish, Italian, Jewish, and many other minorities overtime prejudice and worked their way up. Blacks should do the same), while the second compares only blacks and whites (i.e., Its really a matter of some people not trying hard enough; if blacks would only try harder they could be as well off as whites). The 1996, 1998, 2002 and 2006 surveys do not ask about respondents agreement with those two statements. Statistics are unweighted. Refusal and Dont know are treated as missing values. All mean differences across years between two questionnaires are statistically signicant (p < .001). Source: ANES (Cumulated Data File, self-identied white respondents only).

The second piece of research that supports the model minority theory is McDonalds study (2001) that compares whites explanations for black poverty and Hispanic poverty. McDonald nds that whites differently interpret poverty of the two minorities.2 Whites attribute black poverty to lack of motivation, but explain Hispanic poverty as due to no chance for education (McDonald, 2001, p. 567). In other words, whites blame the victim when explaining black poverty, but blame the system when explaining Hispanic poverty (McDonald, 2001, p. 563). Relating to multiracial evaluation, whites have differentiated evaluation schemes for blacks and Hispanics. Unfortunately, McDonald did not provide any theoretical explanations as to why whites use distinguished evaluation schemes, one for blacks and the other for Hispanics. In our view, plausible explanations should be imported from studies on white employers evaluations of Hispanic workers (Maldonado, 2006; Moss & Tilly, 2001). According to Maldonado (2006), Hispanics are generally thought to be less educated and have poor work skills, like black workers. However, white employers think that Hispanics have better work ethics because they are motivated to realize their American Dream, and thus are more reliable than blacks. In other words, Hispanics are frequently praised as a sort of model minority when compared to blacks because Hispanics have motivation to overcome economic hardship despite their low level of education or poor job skill.

2 Note that his ndings are only based on white respondents in Florida, and should thus be interpreted with caution.

Agreement among white citizens

130

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

Negative stereotype of blacks Negative stereotype of nonblacks (Hispanisc/Asians) Positive stereotype of nonblacks (Hispanisc/Asians) All-negative group

Positive stereotype of blacks Black-positive group

Nonblack-positive (Hispanic-positive / Asian-positive) group

All-positive Group

Fig. 2. Typology of minority comparison model.

are frequently considered cultural foreigners (Kim, 1999; Maldonado, 2006; Tuan, 1998), which carries a negative connotation. Civic ostracism may partly be explained by different minority groups length of stay in the United States, as it is possible that those who have been in the United States longer (e.g., blacks and Latinos) are more accepted by American society, whereas those who came later (e.g., Asians and Hispanics) are more seen and treated as foreigners. To better understand this phenomenon, we reviewed historical evolution of racial categories in the U.S. Census,3 and we found positive association between different minority groups length of stay in the United States and the aforementioned process of civic ostracism among different racial minority groups.4 To summarize, the racial valorization process aims to justify racial inequality faced by blacks by focusing on merits of nonblack minorities, while the civic ostracism process tries to rationalize racial inequality faced by nonblack minorities by focusing on their foreignness. Racial valorization and civic ostracism, according to Kim (1999, p. 107), often occur simultaneously to dominate both [black and nonblack minority] groups. Critical racial theorists, therefore, conclude that the racial triangulation process has arisen as a new racial discourse in multiracial environments for whites to justify and preserve existing racial hierarchies (i.e., white supremacy over black and nonblack minorities). Fig. 2 illustrates the overview of our minority comparison model. The main thrust of our model is that whites make relative comparisons between blacks and nonblack minorities when evaluating social policies. To facilitate understanding, let us focus on dichotomous evaluations (i.e., positivenegative) of: (1) white stereotype of blacks, and (2) white stereotype of nonblacks (see Fig. 2).

The all-positive group has positive stereotypes of both black and nonblack minorities; whereas the all-negative group has negative stereotypes of both black and nonblack minorities. On the other hand, the other two groups adopt multiracial evaluation because they have positive stereotypes of one racial group but negative stereotypes of the other: the nonblack-positive group has positive stereotypes of nonblacks (e.g., Asians or Hispanics) but negative stereotypes of blacks. Contrastingly, the black-positive group has negative stereotypes of blacks but positive stereotype of nonblacks. In other words, nonblack-positive and blackpositive group have differentiated evaluation schemes for racial minorities. According to the minority comparison model, blackpositive and nonblack-positive groups are interesting because the model assumes that their different racial valorization accounts for their preferences in racialized policies. For instance, if a policy is controversial for its inuence on blacks (e.g., afrmative action), the nonblack-positive group is not likely to support it because, echoing model minority discourse and empirical ndings of symbolic racism where white symbolic racism is more pronounced when evaluating blacks against nonblack minorities, it thinks blacks should not receive privilege that other nonblack minorities do not receive. In other words, the occurrence of multiracial evaluation process among nonblack-positive group members is similar to the racial valorization process (Kim, 1999), as illustrated earlier. Meanwhile, if a policy is controversial for its effect on nonblack minorities (e.g., immigration policy), blackpositive group is more likely to have negative opinions on the policy because it has positive views of blacks but negative views of nonblack minorities. Black-positive group members think that blacks have assimilated into American society because blacks share a long history with whites and have formed a common fate as members of American society, whereas Hispanics and Asians are cultural foreigners who have not assimilated (Kim, 1999; Tuan, 1998). In short, the multiracial evaluation process among those in the black-positive group resembles the civic ostracism process that was discussed earlier. 1.3. Research hypotheses and research questions Employing the minority comparison model, this study aims to assess the effects of multiracial evaluation on racialized policy preferences by testing three hypotheses. The rst hypothesis connects the symbolic racism scale (Henry & Sears, 2002; Kinder & Sanders, 1996; Kinder & Sears, 1981) to theoretical predictions derived from racial triangulation theory, especially the racial valorization process (Kim, 1999; McDonald, 2001). According to racial triangulation theorists, white Americans justify negatively racialized opinions toward blacks by comparing them to other model minorities (i.e., by emphasizing success stories or positive stereotypes of Asian Americans). If arguments of racial triangulation theorists are correct, and symbolic racism is present among some white Americans, then we predict:

3 U.S. Census Bureau claims its racial categories to be social-political constructs designed for collecting data on the race and broad population groups in the United States (see Anderson & Fienberg, 2001), and thus it can be argued that the ways in which U.S. Census classify racial groups is implicative of larger societal and political denition and interpretation of different racial groups in the United States. 4 In the U.S. Census, Black made its rst appearance in 1790, whereas other non-black minorities did not appear in the census until about 80 years later: Chinese (1870), American Indians (i.e., Native American, 1870), Japanese (1890), Hindu (i.e., South Asian, 1920), Korean (1920), Filipino (1920), and Hispanics (as an ethnicity, 2000).

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

131

Symbolic racism hypothesis: Individuals in the Nonblack-positive group will show higher levels of symbolic racism than those in the other three groups (i.e., all-positive, all-negative, and black-positive groups). The second and third hypotheses will investigate the effects of our minority comparison model on white Americans policy preference. As mentioned, some policies such as social welfare or afrmative action are black-oriented in that they are particularly controversial among blacks. According to our model, whites that oppose black-oriented policies often justify their resistance by emphasizing success stories of nonblack minorities. In other words, individuals in the nonblack-positive group (i.e., those who have a positive view of nonblack minorities but negative stereotype of blacks) will show greater opposition to blackoriented policies. Thus we hypothesize: Black-oriented policy hypothesis: Individuals in the Nonblack-positive group will show greater opposition to black-oriented policy than those in the other three groups (i.e., all-positive, all-negative, and black-positive groups). Unlike social welfare or afrmative action, policies that deal with illegal immigrants or the strengthening of border control mainly target nonblack racial/ethnic minorities (e.g., Hispanics or Asians), and are thus termed nonblackoriented policy in our study. The third hypothesis follows the same logic underlying the second hypothesis, although it predicts a contrasting pattern due to the nature of the political policies in question. Namely, opposition to an expansive immigration policy should be more pronounced among individuals in the black-positive group (i.e., those who have a positive view of blacks but negative views of nonblack minorities). Thus, the nal hypothesis follows: Nonblack-oriented policy hypothesis: Individuals in the Black-positive group will show greater opposition to nonblack-oriented policy than those in the other three groups (i.e., all-positive, all-negative, and nonblackpositive groups).

2.2. Measures5 2.2.1. Typology of multiracial evaluation Our minority comparison model contrasts white stereotypes of Hispanics and Asians with those of blacks. White stereotypes of racial minorities are measured by using two items that ask whether or not the three racial minorities (blacks, Hispanics, or Asians) are: (1) lazyhardworking, and (2) unintelligentintelligent. The two items are measured using seven-point scale and averaged to operationalize stereotype of racial minorities where a higher value indicates positive stereotypes of racial minorities. In summary, white stereotypes of Asians are the most positive (M = 5.13, S.D. = 1.24); white stereotypes of blacks the most negative (M = 4.07, S.D. = 1.17); and white stereotypes of Hispanics lie somewhere on the continuum (M = 4.50, S.D. = 1.11). Correlation coefcients among three racial minorities showed similar strata of racial minorities. While the correlation between stereotypes of blacks and stereotypes of Hispanics is .53 (p < .001), correlation between stereotypes of blacks and stereotypes of Asians is .30 (p < .001), and correlation between stereotypes of Hispanics and Asians is .49 (p < .001). In general, this nding is consistent with that of previous Racial Studies (Kim, 1999; McDonald, 2001; Sidanius, Devereux, & Pratto, 1992). We create two typologies of multiracial evaluation: (1) the Asianblack typology compares white stereotype of Asians and blacks; (2) the Hispanicsblack typology compares white stereotype of Hispanics and blacks. White stereotypes of all three racial minorities (i.e., blacks, Hispanics and Asians) are dichotomized by applying a mean-split: (1) positive stereotype of a racial minority (i.e., above the mean of stereotype), or (2) negative stereotype of a racial minority (i.e., below the mean of stereotype). In our minority comparison model, whites who have positive stereotypes of all racial minorities are put into the all-positive group; whites who have negative stereotypes of all racial minorities are put into the all-negative group; whites who have positive stereotype of Hispanics but negative stereotypes of blacks are put into the Hispanic-positive group; whites who have positive stereotypes of Asians but negative stereotypes of blacks are put into the Asian-positive; and whites who have positive stereotype of blacks but negative stereotype of Hispanics and Asians are put into the black-positive group. 2.2.2. Outcome variables 2.2.2.1. Symbolic racism scale. Symbolic racism scale is adopted to test the validity of multiracial evaluation in the minority comparison model. Since 1986, the ANES dataset has contained ve symbolic racism items (see Appendix A) (Henry & Sears, 2002; Kinder & Sanders, 1996), and this study constructs its symbolic racism scale by averaging these ve items (Cronbachs = .77; M = 3.59, S.D. = .93). 2.2.2.2. Black-oriented and nonblack-oriented policy opinion. This papers main hypotheses test the effects of multiracial evaluation on white racialized policy preferences, ergo chooses two racialized policies: Afrmative action

2. Method 2.1. Sample In order to reect the most updated trend, this study uses data from 2008 American National Election Studies (ICPSR Study #25383). All of the data were gathered via face-to-face interviews. The 2008 sample contains 2,323 respondents, but this study selects only self-identied white respondents (n = 1,186, 51%) because its primary aim is to investigate white Americans multiracial evaluation of racial minorities, and their effects on racialized policy preferences.

Specic question wordings are provided in Appendix A.

132

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

and Immigration Policy. While afrmative action causes chronic controversy in scientic and policy-making settings relating to the status of the black population (Kinder & Sanders, 1996; Sniderman & Carmines, 1997; Sniderman & Piazza, 1993), immigration policy incites social controversies relating to the recent growth of nonblack minorities, such as Hispanics or Asian immigrants (Junn & Masuoka, 2008; Kim, 2000). Support for afrmative action is measured by asking respondents how much they support the stance that federal government ought to see whether black people are getting fair treatment in jobs or not.6 Respondents are asked to rate their preferences on a ve-point scale (strongly opposestrongly support). In the 2008 survey, 30% of white respondents opposed the governments role to ensure fair jobs for blacks, whereas 23% showed supportive opinion. In general, of all the white Americans, opinion on afrmative action policy is skewed toward opposition (M = 2.96, S.D. = .91). Support for an expansive immigration policy is measured by asking people whether the government should increase, keep the same, or decrease the accepted number of new immigrants from foreign countries. Unlike the question on afrmative action policy, however, there are no racially explicit cues in the immigration policy question (e.g., specic reference to immigrants from Asian countries or Latin America).7 Respondents are provided with ve-point scale to express their opposition to the increase of accepted number of new immigrants: increase a lot (1)decrease a lot (5). In the 2008 ANES survey, 47% of white Americans thought the U.S. government should decrease the number of admitted immigrants either a lot or a little; and only 13% of white respondents thought that the government should increase its number of admitted immigrants. As with the afrmative action policy attitude, white Americans on average expressed opinions that were highly skewed toward the desire to decrease the number of admitted new immigrants (M = 2.47, S.D. = 1.06). 2.2.3. Socio-demographics To adjust for social demographic differences in multiracial evaluation, this study selects four variables: gender, age, annual income level, and educational achievement. In the 2008 survey, 44% of white respondents are male. Averaged age year of whites surveyed is 49.26 (S.D. = 17.29). Median level of annual household income is about 45,00049,999 dollars. Average year of education it 13.67 (S.D. = 2.23). 2.2.4. Political orientations To control for political orientations of respondents, this study adopts (1) self-reported measures of political

ideology, and (2) political party afliation. Sixteen percent of white respondents report that they are political Liberals; and 27% identify themselves as political Conservatives. Relating to party afliation, 30% of white Americans express their attachment to the Democratic Party, whereas 27% choose the Republican Party as their preferred party. 2.2.5. Control variables Since nal outcome variables are policy preferences, two control variables are adopted to reduce the danger of confounding the effects of multiracial evaluation with principle-related inuences in determining policy opinion. Based on previous studies that investigated racially conscious policies (see Chapter 6 in Kinder & Sanders, 1996), equality and limited government scales are chosen for statistical control. Equality scale is calculated by averaging the scores of six items ( = .68; M = 3.38, S.D. = .73), for example, If people were treated more equally in this country we would have many fewer problems (see Appendix A for all the other items). Limited government scale is estimated by asking whether government should provide more or less services using seven-point scale (M = 4.23, S.D. = 1.69). Since the minority comparison model might be inuenced by in-group evaluation, white citizens stereotype of its own racial category is also entered as a control variable. Consistent with stereotypes of racial minorities, the following two items are averaged: (1) lazyhardworking, and (2) unintelligentintelligent ( = .70; M = 4.98, S.D. = 1.06). Finally, since the dataset on which we rely to test hypotheses are gathered by interpersonal interviews, the race of interviewer (Davis & Silver, 2003) is also controlled (9% nonwhite interviewers). 2.2.6. Statistical method In the rst and second hypotheses, we predict that nonblack-positive group will manifest the strongest symbolic racism and opposition toward black-oriented policy. We used analysis of covariance (ANCOVA) with planned contrasts between nonblack-positive group and other three groups to test symbolic racism and black-oriented policy hypotheses. Similarly, we used ANCOVA with planned contrasts between black-positive group and other three groups because the nonblack policy hypothesis anticipates that black-positive group will manifest the strongest opposition toward nonblack-oriented policy. 3. Results 3.1. Testing symbolic racism hypothesis The rst hypothesis seeks evidence that white respondents symbolic racism is more pronounced among those in the nonblack-positive group (i.e., Asian-positive or Hispanic-positive group) than the black-positive, allpositive, and all-negative groups. Results of ANCOVA with planned contrast offers clear statistical support for Asianblack typology, but only marginally signicant support for Hispanic-black typology, after controlling for sociodemographics, political orientations, and other policyrelated attitudes (see Appendix B for specic signicance

6 The question contains a very clear and explicit racial cue, that is, blacks. As Sniderman and Carmines (1997) pointed out, racially loaded question wording (e.g., black) generally elicits greater opposition to the policy implementation than racially neutral question wording (e.g., poor people). 7 Although expressions such as immigrants from foreign countries may work as racially implicit cues in the same way that inner city residents cues black implicitly (Pefey & Hurwitz, 2002), empirical tests are required to verify whether such implicit assumption is valid.

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

133

Symbolic Racism scale among white citizens


(Minority Comparison Model)
Comparing Blacks and Hispanics (i.e., Hispanic-black typology)
3.8 3.8

Comparing Blacks and Asians (i.e., Asians-black typology)


Asian-positive
3.74

Hispanic-positive Symbolic Racism Scale


3.67

Symbolic Racism Scale

3.61

3.61

All-negative Black-positive
3.5

All-negative

3.59

Black-positive

Hispanic=Negative Hispanic=Positive Blacks=Negative

3.48

Asians=Negative
3.43

All-positive 3.4 3.4 Asians=Positive Blacks=Negative All-positive Blacks=Positive

Blacks=Positive

Fig. 3. Results of symbolic racism hypothesis. Note: Planned contrast detects marginally signicant evidence that whites symbolic racism is more pronounced among those in the Hispanic-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) than those in the other three groups (t = 1.93, p = .06, two-tailed test); and whites symbolic racism is more pronounced among those in the Asian-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) than those in the other three groups (t = 3.18, p < .01, two-tailed test).

tests of control variables). Fig. 3 provides the mean pattern of symbolic racism in minority comparison model. In Asian-black typology (right panel in Fig. 3), whites in the Asian-positive group (i.e., those who have positive stereotype of Asians but negative stereotype of blacks) show a substantially higher symbolic racism score (M = 3.74) than black-positive (M = 3.59), all-positive (M = 3.59), or all-negative group (M = 3.61). Planned contrast conrms that whites symbolic racism is more pronounced among Asian-positive group than any of the other three groups (t = 3.18, p < .01). In Hispanicblack typology (left panel in Fig. 3), whites in the Hispanic-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) show slightly higher symbolic racism (M = 3.67) than black-positive (M = 3.50), all-positive (M = 3.48), or allnegative group (M = 3.61). However, planned contrast only detects marginally signicant evidence that whites symbolic racism is more pronounced among Hispanic-positive group than any of the other three groups (t = 1.93, p = .06). In sum, there is supportive evidence that white symbolic racism is augmented among whites that have positive stereotype of nonblack minorities but negative stereotype of blacks, which is supportive of the symbolic racism hypothesis. 3.2. Testing black-oriented policy hypothesis The second hypothesis posits that whites who have positive stereotype of nonblack minorities but negative stereotype of blacks are most likely to oppose black-oriented policy (i.e., policies that ensure fair jobs for blacks) than any other whites (i.e., those in all-positive, all-negative, or black-positive groups). Similar

to the results obtained from the symbolic racism hypothesis, there is supportive evidence for Asian-black typology, but no evidence for Hispanic-black typology, after controlling for socio-demographics, political orientations, and other policy-related attitudes (see Appendix B for specic signicance tests of control variables). Fig. 4 provides the mean pattern of support for black-oriented policy in the minority comparison model. In Asian-black typology (right panel in Fig. 4), whites in the Asian-positive group (i.e., those who have positive stereotype of Asians but negative stereotype of blacks) show substantially less support for blackoriented policy (M = 2.82) than those in black-positive (M = 3.02), all-positive (M = 2.95), or all-negative (M = 2.90) groups. Planned contrast conrms that whites support for black-oriented policy is substantially weaker among Asianpositive group than any of the other three groups (t = 2.03, p < .05). In Hispanic-black typology (left panel in Fig. 4), however, whites in the Hispanic-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) are only the second to the least supportive of black-oriented policy out of the four groups (M = 2.94). Instead, whites in the all-negative group (i.e., those who have negative stereotypes of both blacks and Hispanics) are the least supportive among the four groups (M = 2.83). Planned contrast does not support blackoriented policy hypothesis in Hispanic-black typology (t = .10, p = ns). As Fig. 4 clearly shows, whites support for black-oriented policy is mainly determined by whether a white respondent has positive stereotype of blacks or not, regardless of his/her stereotype of Hispanics. To summarize, there is partial evidence that whites support for black-oriented policy weakens only among

134

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

Support for black-oriented policy among white citizens


(Minority Comparison Model)
Comparing Blacks and Hispanics (i.e., Hispanic-black typology)
3.1 Hispanics=Negative All-positive Support for black-oriented policy Hispanics=Positive Support for black-oriented policy
3.02

Comparing Blacks and Asians (i.e., Asian-black typology)


3.1 Asians=Negative All-positive Asians=Positive
3.02

2.97

2.95

2.94

Black-positive

Hispanic-positive

All-negative
2.9

Black-positive

2.83

2.82

2.8

2.8

All-negative

Asian-positive Blacks=Negative Blacks=Positive

Blacks=Negative

Blacks=Positive

Fig. 4. Results of black-oriented policy hypothesis. Note: Planned contrast fails to detect signicant evidence that whites support for black-oriented policy weakens among those in the Hispanic-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) than those in the other three groups (t = .10, p = ns, two-tailed test); however, planned contrast indicates that whites support for black-oriented policy signicantly weakens among those in the Asian-positive group (i.e., those who have positive stereotype of Hispanics but negative stereotype of blacks) than those in the other three groups (t = 2.03, p < .05, two-tailed test).

whites in the Asian-positive group (i.e., those who have positive stereotype of Asians but negative stereotype of blacks). However, consistent with our nding in the Asian-black typology, whites stereotype of Hispanics has no inuence on their support for black-oriented policy.

3.3. Testing nonblack-oriented policy hypothesis The nal hypothesis posits that whites who have positive stereotype of blacks but negative stereotype of nonblack minorities are more likely to oppose nonblack-oriented policy (i.e., policies in favor of new

Support for nonblack-oriented policy among white citizens


(Minority Comparison Model)
Comparing Blacks and Hispanics (i.e., Hispanic-black typology)
2.7 All-positive
2.66

Comparing Blacks and Asians (i.e., Asian-black typology)


2.7
2.67

All-positive Support for nonblack-oriented policy Support for nonblack-oriented policy All-negative
2.46

Hispanic-positive
2.44 2.39

All-negative

2.33

Asian-positive

2.16

Hispanic=Negative
2.08

Asians=Negative Asians=Positive Blacks=Negative

Black-positive

Blacks=Negative

Blacks=Positive

Hispanic=Positive

Black-positive

Blacks=Positive

Fig. 5. Results of nonblack-oriented policy hypothesis. Note: Planned contrast detects signicant evidence that whites support for nonblack-oriented policy weakens among those in the black-positive group (i.e., those who have positive stereotype of blacks but negative stereotype of nonblack) than those in the other three groups (t = 2.93, p < .01 in left panel; t = 2.73, p < .01 in right panel, two-tailed test).

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

135

immigrants) than other whites (i.e., those in all-positive, all-negative, and black-positive groups). In both Asianblack and Hispanic-black typology, there is supportive evidence of the nonblack-oriented hypothesis, controlling for socio-demographics, political orientations, and other policy-related attitudes (see Appendix B for specic significance tests of control variables). Fig. 5 shows the mean pattern of support for nonblack-oriented policy in the minority comparison model. In the Asian-black typology (right panel in Fig. 5), whites in the black-positive group show substantially less support (M = 2.16) for nonblack-oriented policy than whites in either nonblack-positive (M = 2.33), all-positive (M = 2.67), or all-negative group (M = 2.46). Planned contrast conrms that whites support for nonblack-oriented policy is substantially weaker among black-positive group than any of the other three groups (t = 2.73, p < .01). In the Hispanic-black typology (left panel in Fig. 5), whites in the black-positive group show substantially less support (M = 2.08) for nonblack-oriented policy than whites in either nonblack-positive (M = 2.44), all-positive (M = 2.39), or all-negative (M = 2.39) group. Planned contrast indicates that whites support for nonblack-oriented policy is substantially weaker among black-positive group than any of the other three groups (t = 2.93, p < .01). To summarize, there is fairly strong evidence that whites support for nonblack-oriented policy is signicantly weaker among black-positive group (i.e., those who have positive stereotype of blacks but negative stereotype of Hispanics and/or Asians). 4. Discussion and conclusion Empirical ndings in this study serve as evidence that multiracial evaluation, which compares blacks and nonblack minorities, is adequate in explaining white Americans symbolic racism against blacks, as well as their opinions on racialized policies. The three hypotheses are supported by data from a recent national representative survey (ANES 2008), indicating that our minority comparison model is successful in explaining why white Americans oppose both black-oriented and nonblackoriented racialized policies. Compared to those who have undifferentiated negative stereotypes of all racial minorities, white Americans who have positive stereotypes of nonblack minorities but negative stereotype of blacks manifest stronger symbolic racism and opposition to afrmative action policy. Moreover, this study presents novel ndings that offer explanations for why some white Americans oppose nonblack-oriented policies as opposed to others (e.g., immigration policy). Compared to those who have negative stereotypes of all racial minorities, white Americans who have positive stereotypes of blacks but negative stereotypes of nonblack minorities show stronger opposition to immigration policy. Findings presented in this study suggest that multiracial evaluation should be adopted in future studies of race and its effects on public opinion regarding racialized policies. Our ndings are unique and theoretically meaningful in several aspects. To our knowledge, this is the rst

empirical study demonstrating that multiracial evaluation indeed inuences white Americans policy preferences and racial attitudes. While studies on nonblack minorities have yet to be established in mainstream quantitative studies, consideration of nonblack minorities helps to clarify racial tensions between blacks and whites (e.g., opinion on afrmative action) and much more. This study provides telling evidence that white Americans racial resentment toward blacks and preference for black-oriented policies are explained substantially when evaluations of nonblack minorities are taken into consideration. Considering the steady and substantial growth of nonblack population in the United States, our ndings suggest that multiracial evaluation will have a much stronger inuence on white Americans policy preferences and U.S. politics overtime. While our ndings are limited to the U.S. population, we believe our minority comparison model may be generalizable to other parts of the world with comparable multiracial demographic. Taking Brazil for example, not only has it been found that white Brazilians also attribute low socio-economic status of black Brazilians to their lack of motivation and work ethics by praising nonblack minority immigrants (i.e., Asians) for their business achievements, but also white Brazilians ostracize nonblack minority immigrants due to their relatively short residence in Brazil and lack of assimilation (Lesser, 1999; Skidmore, 1974). Nevertheless, more empirical studies need to be conducted to further test the generalizability of our model in non-U.S. contexts. As aforementioned, the minority comparison model is constructed by integrating racial triangulation theory (one of the recent achievements in qualitative studies on race) with empirical ndings in public opinion research (symbolic racism scale; and McDonald, 2001), and the model is especially fruitful given the subtlety of racial attitudes (Mandelberg, 1999) in modern, multiracial American society (Junn & Masuoka, 2008; Kim, 1999, 2000). Reective of Hochschilds propositions (2000), this study recognizes that race can only be understood most efciently when scholars adopt a multiperspective and multimethod approach. Due to the limitation of secondary analysis, we understand that our study is bound by several limitations, and should thus be read with caution. First, survey questions of afrmative action and immigration policies may not be perfectly comparable. As mentioned, the afrmative action questionnaire contains explicit racial cues (i.e., fair jobs for blacks), while immigration policy questionnaire entails relatively implicit racial cues (i.e., new immigrants from foreign countries). Even though this study succeeds in nding clearly contrasted opposition patterns between black-positive and nonblack-positive groups (e.g., in Figs. 4 and 5), validity of these ndings should be tested using comparable questionnaires in future studies. Second, this study does not provide contextual effect mechanisms explaining why white citizens develop differentiated stereotypes of racial minorities. One possibility is the effect of socio-economic status (SES) on white citizens construction of multiracial evaluation. For example, high SES whites are more likely to be in contact with

136

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

Asians because Asians in general achieve relatively higher SES than blacks (Kim, 2000), whereas low SES whites are more likely to be in contact with blacks because of residential patterns (for a classic study, see Du Boise, 1973). Recent scholars interested in racial environments of white communities have reported that white racial attitudes and policy opinions can be substantially explained by racial heterogeneity, or SES status of a county (Branton & Jones, 2005; Campbell, 2008; Oliver & Mendelberg, 2000). Further studies are encouraged to clarify contextual inuences of multiracial evaluations. Nevertheless, despite its limitations, we believe this study succeeds in producing novel and interesting ndings that contribute to the understanding of subtle racial dynamics and public opinion on racialized policies in modern, multiracial American society, and provides a solid theoretical explanation for contemporary nuances in white Americans multiracial evaluations and delicate sentiments toward both black-oriented and nonblack-oriented policies. Appendix A. All analyses are based on American National Election Studies 2008, and we selected (1) self-identied white Americans (V083251A). Effects of interviewers race Race of interviewer: V082254 Social demographics Gender: V081111A Age: V081104 Household income: V083248 Education achievement: V083217

Political orientations Party afliation: V083097 Political Ideology: V083069 Equality scale Equality scale is the composite of six items (Kinder & Sanders, 1996): V085162 to V085167 Our society should do whatever is necessary to make sure that everyone has an equal opportunity to succeed. We have gone too far in pushing equal rights in this country (recoded). One of the big problems in this country is that we do not give everyone an equal chance. It is not really that big a problem if some people have more of a chance in life than others (recoded). The country would be better off if we worried less about how equal people are. If people were treated more equally in this country we would have many fewer problems. Limited government scale: V083105 Some people think the government should provide fewer services, even in areas such as health and education, in order to reduce spending. Other people feel that it is important for the government to provide many more services even if it means an increase in spending. Where would you place yourself on this scale, or have not you thought much about this? Symbolic racism scale: V085143V085146 Symbolic racism scale is the composite of the following ve items:

Table B1 Results testing symbolic racism, black-oriented policy, and nonblack-oriented policy hypotheses using Hispanicblack typology (i.e., all-positive, allnegative, Hispanic-positive, and black-positive groups). Source d.f. F-Statistics Symbolic racism scale Hispanicblack typology [cate] Socio-demographics Gender [cate] Age (years) [cont] Educational achievement [cont] Household income [cont] Political orientations Party afliation [cate] Political ideology [cate] Policy-related attitudes Equality scale [cont] Limited role of government [cont] Ingroup stereotype [cont] Race of interviewers [cate] Residual 3 1 1 1 1 2 2 1 1 1 1 1,037 3.33* 3.04 1.67 20.50*** 0.35 3.26* 25.09*** 112.78*** 1.71 9.51*** 4.45* (.62) Black-oriented policy 2.72* 3.29 0.06 0.07*** 0.56 3.07* 3.75*** 33.98*** 0.38 2.35*** 2.87* (.76) Nonblack-oriented policy 6.32*** 5.18* 2.17 21.57*** 1.65 0.11 12.02*** 18.00*** 2.54 10.53*** 1.18 (1.00)

Source: ANES 2008 (N = 1,186, self-identied white respondents only). Note: Mean square of errors entered in parentheses. [cate] denotes the entered variables are treated as categorical variables; and [cont] denotes continuous variables. * p < .05. *** p < .001.

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138

137

Table B2 Results testing symbolic racism, black-oriented policy, and nonblack-oriented policy hypotheses using Asianblack typology (i.e., all-positive, all-negative, Asian-positive, and black-positive groups). Source d.f. F-Statistics Symbolic racism scale Asianblack typology [cate] Socio-demographics Gender [cate] Age (years) [cont] Educational achievement [cont] Household income [cont] Political orientations Party afliation [cate] Political ideology [cate] Policy-related attitudes Equality scale [cont] Limited role of government [cont] Ingroup stereotype [cont] Race of interviewers [cate] Residual 3 1 1 1 1 2 2 1 1 1 1 1,034 5.75*** 2.70 1.98 18.07*** 0.20 3.59* 24.82*** 109.94*** 1.40 6.95* 4.98* (.62) Black-oriented policy 3.20* 4.02* 0.12 0.12 0.41 3.04* 3.53* 35.88*** 0.27 0.65 3.21+ (.76) Nonblack-oriented policy 5.52*** 5.37* 1.27 22.17*** 1.47 0.03 13.69*** 17.33*** 2.28 7.01** 1.49 (.99)

Source: ANES 2008 (N = 1,186, self-identied white respondents only). Note: Mean square of errors entered in parentheses. [cate] denotes the entered variables are treated as categorical variables; and variables with [cont] are considered as continuous variables. * p < .05. ** p < .01. *** p < .001.

Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should to the same without any special favors. Its really a matter of some people not trying hard enough; if blacks would only try harder they could be just as well off as whites. Do you think that civil rights leaders are trying to push too fast, are going too slowly, or are they moving about the right speed? Generations of slavery and discrimination have created conditions that make it difcult for blacks to work their way out of the lower class. Over the past few years blacks have gotten less than they deserve. Opposition to afrmative action policy: V085079 Opposition to the increase of immigrants: V085082 Appendix B. See Tables B1 and B2. References
Allport, G. W. (1954). The nature of prejudice. Reading, MA: AddisonWesley. Anderson, M., & Fienberg, S. (2001). Census 2000 and the politics of census making. Society, 39(1), 1725. Branton, R. P., & Jones, B. S. (2005). Reexamining Racial Attitudes: The Conditional Relationship between Diversity and Socioeconomic Environment. American Journal of Political Science, 49(2), 359372. Browning, R. P., Marshall, D. R., & Tabb, D. H. (1984). Protest is not enough: The struggle of blacks and Hispanics for equality in urban politics. Berkeley, CA: University of California Press. Campbell, D. E. (2008). Why we vote: How schools and communities shape our civic life. Princeton, NJ: Princeton University Press. Davis, D. W., & Silver, B. D. (2003). Stereotype threat and race of interviewer effects in a survey on political knowledge. American Journal of Political Science, 47(1), 3345.

Du Boise, W. E. B. (1973). The Philadelphia Negro: A social study. Millwood, NY: Kraus-Thomson Organization Ltd. Entman, R. M., & Rojecki, A. (2000). The black image in the white mind: Media and race in America. Chicago, IL: University of Chicago Press. Fiske, S. T. (1998). Streotyping, prejudice, and discrimination. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. II, pp. 357414). New York: The McGraw-Hill Companies, Inc. Guinier, L., & Torres, G. (2002). The miners canary: Enlisting race, resisting power, transforming democracy. Cambridge, MA: Harvard University Press. Henry, P. J., & Sears, D. O. (2002). The symbolic racism 2000 scale. Political Psychology, 23(2), 253283. Hochschild, J. L. (2000). Lumpers and splitters, individuals and structures: Comments on racialized politics. In D. O. Sears, J. Sidanius, & L. Bobo (Eds.), Racialized politics: The debate about racism in America (pp. 324343). Chicago, IL: University of Chicago Press. Junn, J., & Masuoka, N. (2008). Asian American identity: Shared racial status and political context. Perspectives on Politics, 6(04), 729740. Kim, C. J. (1999). The racial triangulation of Asian Americans. Politics & Society, 27(1), 105138. Kim, C. J. (2000). Bitter fruit: The politics of BlackKorean conict in New York City. New Haven, CT: Yale University Press. Kinder, D. R., & Sanders, L. (1996). Divided by color: Racial politics and democratic ideals. Chicago, IL: University of Chicago Press. Kinder, D. R., & Sears, D. O. (1981). Prejudice and politics: Symbolic racism versus racial threats to the good life. Journal of Personality and Social Psychology, 40(3), 414431. Lesser, J. (1999). Negotiating national identity: Immigrants, minorities, and the struggle for ethnicity in Brazil. Durham, NC: Duke University Press. Maldonado, M. M. (2006). Racial triangulation of Latino/a workers by agricultural employers. Human Organization, 65(4), 353361. Mandelberg, T. (1999). The race card: Campaign strategy, implicit messages, and the norm of equality. Princeton, NJ: Princeton University Press. McDonald, S. J. (2001). How Whites explain Black and Hispanic inequality. The Public Opinion Quarterly, 65(4), 562573. Moss, P., & Tilly, C. (2001). Stories employers tell: Race, skill, and hiring in America. New York: Russell Sage Foundation. Oliver, J. E., & Mendelberg, T. (2000). Reconsidering the Environmental Determinants of White Racial Attitudes. American Journal of Political Science, 44(3), 574589. Pefey, M., & Hurwitz, J. (2002). The racial components of race-neutral crime policy attitudes. Political Psychology, 23(1), 5975. Petersen, W. (1966, January 9). Success story, JapaneseAmerican style. The New York Times Magazine.

138

Y.M. Baek, A.M. Lee / The Social Science Journal 49 (2012) 127138 Sonenshein, R. J. (1997). The prospects for multiracial coalitions: Lessons from Americas three largest cities. In R. P. Browning, D. R. Marshall, & D. H. Tabb (Eds.), Racial politics in American cities (2nd ed., pp. 261276). New York: Longman. Sonenshein, R. J., & Pinkus, S. H. (2005). Latino incorporation reaches the urban summit: How Antonio Villaraigosa won the 2005 Los Angeles mayors race. Political Science & Politics, 38(4), 713721. Tuan, M. (1998). Forever foreigners or honorary whites? The Asian ethnic experience today. Brunswick, NJ: Rutgers University Press.

Sidanius, J., Devereux, E., & Pratto, F. (1992). A Comparison of Symbolic Racism Theory and Social Dominance Theory as Explanations for Racial Policy Attitude. Journal of Social Psychology, 132(3), 377395. Skidmore, T. E. (1974). Black into white: Race and nationality in Brazilian thought. New York: Oxford University Press. Sniderman, P., & Piazza, T. (1993). The scar of race. London, UK: Belknap Press. Sniderman, P. M., & Carmines, E. G. (1997). Reaching beyond race. Cambridge, MA: Harvard University Press.

You might also like