Racial Sympathy and its Limits Appendix

Figures 0 .2 Racial Sympathy .4 .6 .8 1 Figure 1: The Distribution of Racial Sympathy and Racial Resentment 0 .2 .4 .6 Racial Resentment Source: 2013 CCES .8 1 .8 .6 .4 .2 Community Service Punishment 1 Figure 2: Racial Sympathy and Criminal Punishment, by Race of Culprit 0 .2 .4 .6 .8 Racial Sympathy Index Black Culprit White Culprit Source: 2016 YouGov Study 1 Tables Table 1: Racial Sympathy Index The Racial Sympathy Index Mrs. Lewis, a white woman with young children, posts advertisements for a nanny on community bulletin boards. She receives many inquiries and decides to interview all applicants over the phone. Mrs. Lewis is most impressed with a woman named Laurette, who has relevant experience, is an excellent cook, and comes enthusiastically recommended. Mrs. Lewis invites Laurette over for what she expects will be the final step of the hiring process. When Laurette arrives, Mrs. Lewis is surprised to see that Laurette is black. After Laurette's visit, which goes very well, Mrs. Lewis thanks her for her time but says that she will not be offered the job. When Laurette asks why, Mrs. Lewis says that she doesn't think that her children would feel comfortable around her. Laurette is upset about Mrs. Lewis' actions. Tim is a white man who owns a hair salon. His business is growing rapidly and so he decides to place an advertisement to hire new stylists. In the advertisement, he writes that interested applicants should come for an interview first thing next Monday. When he arrives at the salon on Monday, he sees a line of seven or eight people waiting outside the door, all of whom appear to be black. He approaches the line and tells the applicants that he's sorry, but the positions have been filled. The applicants are upset; they feel they have been turned away because of their race. Milford is a mid-sized city in the Northeast. The main bus depot for the city is located in the Whittier section of Milford, a primarily black neighborhood. Whittier community leaders argue that the concentration of buses produces serious heath risks for residents; they point to the high asthma rates in Whittier as evidence of the bus depot's harmful effects. The Milford Department of Transportation officials, who are mostly white, state that Whittier is the best location for the depot because it is centrally located and many Whittier residents take the bus. Furthermore, it would be expensive to relocate the bus depot to a new location. Whittier community leaders are very upset by the Department's inaction. Michael is a young black man who lives in a midwestern city. One day Michael is crossing the street and jaywalks in front of cars. Some local police officers see Michael jaywalk and stop and question him. Michael argues that he was just jaywalking and is otherwise a law-abiding citizen. The police officers feel that Michael is being uncooperative and so they give him a pat down to see if he is carrying any concealed weapons. Michael is very upset by this treatment. Table 2a: Principal Factor Analyses of Whites’ Responses to Racial Sympathy Index Variable PFA Results Factor 1 2.3 Eigenvalues (58) (% of variance explained) Vignette 1: Laurette - hiring Vignette 2: Hair salon applicants Vignette 3: Bus depot Vignette 4: Michael - police 0.81 0.82 0.67 0.72 Source: 2013 CCES Table 2b: Principal Factor Analyses of Whites’ Responses to Racial Sympathy Index and Racial Resentment Index Variable PFA Results Factor 1 3.43 Factor 2 2.80 Eigenvalues (43) (35) (% of variance explained) Vignette 1: Laurette - hiring Vignette 2: Hair salon applicants Vignette 3: Bus depot Vignette 4: Michael - police Racial Resentment – Irish Racial Resentment – Generations Racial Resentment – Try harder Racial Resentment - Deserve 0.1 0.05 -0.06 -0.23 0.93 0.88 0.79 0.84 0.88 0.86 0.63 0.58 0.11 0.01 -0.07 -0.02 Source: 2013 CCES Note: See Web Appendix for full text of the vignettes and question wording of the racial resentment index. Table 3: Racial Sympathy and Support for Racialized Public Policies Racial Sympathy Racial Resentment Party ID Limited Govt. Constant Observations Govt Aid to Blacks 0.10** (0.046) -0.61*** (0.035) -0.02 (0.029) -0.06*** (0.022) 0.08* (0.042) 750 Black Welfare Businesses 0.16** 0.23*** (0.065) (0.073) -0.29*** -0.46*** (0.054) (0.064) -0.12** -0.09* (0.047) (0.050) -0.18*** 0.05 (0.035) (0.041) 0.53*** 0.16** (0.062) (0.071) 570 288 Source: 2013 CCES Black Schools 0.17* (0.085) -0.37*** (0.076) -0.18*** (0.065) -0.07 (0.050) 0.46*** (0.094) 289 Black Scholarships 0.26*** (0.085) -0.42*** (0.072) -0.11* (0.058) -0.06 (0.049) 0.24** (0.091) 289 Aff. Action 0.03 (0.071) -0.58*** (0.072) -0.08 (0.056) -0.15*** (0.039) 0.27*** (0.079) 570 *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The column headings display the dependent variables, which are questions about policy opinion. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were also included in the models: income, age, education, gender, region (South) and church attendance. Table 4: Racial Sympathy and Support for Women’s Policies Abortion Model Racial Sympathy Racial Sympathy -0.01 (0.082) Gender Sympathy Racial Resentment Party (1=Republican) Limited Govt Constant Observations R-squared Gender Sympathy 0.08 (0.079) -0.17*** (0.062) -0.08* (0.048) 0.77*** (0.087) 569 0.303 Women’s Leave Racial Sympathy Gender Sympathy 0.16 (0.102) 0.11 (0.074) 0.06 (0.075) -0.16** (0.062) -0.08 (0.048) 0.69*** (0.089) -0.29*** (0.104) -0.28*** (0.079) -0.17*** (0.061) 0.52*** (0.106) 569 566 0.308 0.289 Source: 2013 CCES Women’s Affirmative Action Racial Sympathy Gender Sympathy 0.05 (0.060) 0.16** (0.078) -0.32*** (0.099) -0.28*** (0.079) -0.16*** (0.062) 0.49*** (0.106) -0.35*** (0.059) -0.07 (0.055) -0.08** (0.031) 0.32*** (0.071) 0.18*** (0.052) -0.34*** (0.055) -0.05 (0.053) -0.07** (0.031) 0.23*** (0.068) 566 0.291 571 0.281 571 0.305 *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The column headings display the dependent variables, which are questions about policy opinion. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were included in the models: income, age, education, gender, region (South) and church attendance. 6 Table 5: Racial Sympathy and Criminal Punishment, by Race of Culprit Black Culprit =1 Racial Sympathy Black Culprit x Racial Sympathy Racial Resentment (r coded) Black Culprit x Racial Resentment Constant Observations Model 1 Community Service Punishment (1=more hours of community service) 0.23** (0.11) 0.41*** (.13) -0.63*** (.17) 0.13*** (.06) 210 Model 2 Community Service Punishment (1= more hours of community service) 0.26** (0.11) 0.24* (.14) -0.48** (.20) .29* (.17) -0.26 (0.22) 0.09 (.07) 210 Source: YouGov 2016 Study *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. Unlike preceding tables, the dependent variable is coded such that 0=the pro-Black outcome (less punishment). These results are robust to models that include control variables, including: party identification, limited government, education, income, gender, region and church attendance. However, since not all scholars agree this is an optimal approach, I present the bivariate results here (see Mutz 2015 and Morton & Williams 2010). 7 Appendix Appendix Figure 1: Racial Sympathy and Discriminant Validity Permitting gays and lesbians to adopt children Endorsing laws that protect homosexuals against job discrimination Permitting gays/lesbians to serve in the army Permitting gays and lesbians to marry 0.036 (.053) 0.028 (.05) 0.036 (.05) 0.071 (.04) 0.092 (.06) Invade Iran Bomb Iran’s nuclear development sites Spending on education Belief that global warming is bad Environment/Job Tradeoff 0.068 (.05) -0.058 (.046) -0.047 (.029) 0.042 (.048) -0.015 (.031) Federal spending on the environment Investing social security in the stock market Federal govt. should make it more difficult to buy a gun Abortion -0.00 (.039) -0.026 (.042) 0.021 (.032) -0.01 (.041) Favor direct diplomatic talks with Iran Support for nuclear power plants -0.39 (.044) Source: 2012 ANES Whites only; analyses are weighted for national representativeness. Cell entries are the OLS coefficients for racial sympathy on policy opinion, with controls for party, income, age, education, gender, region, church attendance, limited government, and racial resentment included. Standard errors are in parentheses underneath. The questions about gay issues restrict the sample to self-identified straight whites. Bold coefficients indicate those issue areas that are significantly predicted by racial resentment (p<0.05). A similar analysis, using egalitarianism in place of resentment, finds that all displayed policy areas, with the exception of abortion, are also significantly predicted by egalitarianism. 8 Appendix Table 1: Distribution of the Racial Sympathy Vignettes* Response Categories “A great deal of sympathy” “A lot of sympathy” “Some sympathy” “A little sympathy” “I do not feel any sympathy” 43.37 29.26 16.65 4.53 5.79 36.96 27.25 22.45 8.34 5.01 11.02 18.82 37.16 18.20 14.33 19.82 19.97 24.46 15.04 20.64 Items Black woman, “Laurette”, turned away for nanny job after interview Group of Blacks turned away for salon jobs Black neighborhood ignored by local government Black man, “Michael”, pat down by police after jaywalking Source: 2013 CCES *See Table 1 for vignette text. 9 Appendix Table 2: Correlates of Racial Sympathy and Racial Resentment Party (1=Republican) Income Age Education Gender (1=Female) Church Attendance Region (1=South) Limited Government Constant Observations Racial Sympathy -0.16*** (0.037) -0.03 (0.055) -0.04 (0.044) 0.07* (0.039) 0.02 (0.022) -0.01 (0.031) -0.01 (0.023) -0.07* (0.034) 0.73*** (0.050) 571 Racial Resentment 0.27*** (0.039) 0.01 (0.054) 0.13** (0.056) -0.11** (0.043) 0.04* (0.023) 0.00 (0.033) -0.03 (0.026) 0.17*** (0.033) 0.61*** (0.059) 571 Source: 2013 CCES *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The column headings indicate the dependent variables, the racial sympathy and racial resentment indices. 10 Appendix Table 3: Racial Sympathy and Support for Racialized Public Policies Govt Aid to Blacks Welfare Affirmative Action VARIABLES 2012 ANES 2008 ANES 1994 GSS 2012 ANES 2008 ANES 1994 GSS 2012 ANES 2008 ANES 1994 GSS Racial Sympathy 0.12*** 0.17*** 0.22*** 0.05** 0.05* 0.06 0.15*** 0.15*** 0.09** (0.027) (0.037) (0.060) (0.026) (0.032) (0.057) (0.034) (0.039) (0.040) Racial Resentment -0.53*** -0.49*** -0.98*** -0.27*** -0.25*** -0.27*** -0.49*** -0.40*** -0.31*** (0.024) (0.038) (0.088) (0.025) (0.039) (0.081) (0.033) (0.044) (0.056) Party ID -0.09*** -0.09*** -0.04 -0.10*** -0.07** -0.19*** -0.01 -0.06** -0.05 (0.017) (0.029) (0.049) (0.018) (0.027) (0.049) (0.023) (0.030) (0.033) Limited Govt -0.08*** -0.06*** 0.05 -0.14*** -0.10*** -0.06 -0.07*** -0.04* -0.00 (0.014) (0.022) (0.072) (0.015) (0.022) (0.072) (0.019) (0.022) (0.047) Constant 0.74*** 0.69*** 0.23** 0.72*** 0.85*** 0.44*** 0.62*** 0.56*** 0.21*** (0.027) (0.046) (0.092) (0.025) (0.040) (0.094) (0.036) (0.048) (0.061) Observations 5,142 2,013 414 5,403 2,169 508 5,360 2,109 628 R-squared 0.454 0.381 0.337 0.302 0.219 0.119 0.291 0.234 0.114 1 Sources: 2012 American National Election Study (ANES), 2008 ANES, 1994 General Social Survey (GSS) *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The top column headings display the dependent variables, which are questions about policy opinion. Below these headings, the results for each survey are presented. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were also included in the models: income, age, education, gender, region (South) and church attendance. 1 For all GSS analyses, see note in appendix about variable construction. 11 Appendix Table 4: Racial Sympathy and Support for Government Aid to Blacks – Robustness Checks 2012 ANES Racial Sympathy Party ID Limited Govt Racial Resentment Stereotypes Stereotypes Egalitarianism 0.11*** (0.027) 0.11*** (0.027) -0.09*** -0.08*** (0.017) (0.017) -0.08*** -0.06*** (0.014) (0.015) -0.51*** -0.50*** (0.026) -0.11*** (0.039) (0.026) 2008 ANES Close to blacks 0.11*** (0.027) 0.09*** (0.017) 0.08*** (0.014) 0.53*** (0.025) Stereotypes Egalitarianism IAT^ Stereotypes Egalitarianism Contact with blacks 0.17*** (0.038) 0.16*** (0.038) 0.19*** (0.059) 0.23*** (0.061) 0.19*** (0.071) -0.09*** -0.08*** -0.04 -0.01 -0.02 (0.030) (0.030) (0.048) (0.051) (0.061) -0.06*** -0.05** 0.07 0.07 0.04 (0.022) (0.023) (0.070) (0.073) (0.086) -0.49*** -0.47*** -0.92*** -0.96*** -0.95*** (0.039) -0.06 (0.069) (0.040) 0.19*** (0.038) 0.09*** (0.031) 0.06*** (0.023) 0.47*** (0.040) (0.089) -0.46*** (0.118) (0.091) (0.105) 0.12*** (0.030) Egalitarianism 1994 GSS 0.10* (0.050) 0.09 (0.059) 0.03 (0.024) Close to Blacks -0.04 (0.052) Implicit Racial Attitudes Contact with Blacks Constant Observations R-squared 0.79*** (0.033) 5,132 0.456 0.63*** (0.039) 5,142 0.459 0.72*** (0.031) 5,128 0.454 0.71*** (0.055) 1,988 0.382 0.60*** (0.068) 2,013 0.384 0.69*** (0.058) 1,288 0.385 0.51*** (0.117) 409 0.370 0.43*** (0.101) 455 0.142 -0.04 (0.038) 1.23*** (0.131) 283 0.325 *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The table displays the results from three independent samples – as indicated in the top row—the 2012 ANES, the 2008 ANES, and the 1994 GSS. Additionally, each individual column displays a different control variable that was used in the analysis. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were also included in the models: income, age, education, 12 gender, region (South) and church attendance. The stereotypes variable represents the extent to which the respondent rates blacks as lazy relative to whites with a score of 1= blacks are lazier than whites and 0 =whites are lazier than blacks. ^ The IAT appeared on the 2008-9 ANES Panel. 13 Appendix Table 5: Correlates of Gender Sympathy & Racial Sympathy Gender Racial Sympathy Sympathy Party (1=Republican) -0.04 -0.16*** (0.037) (0.037) Income -0.06 -0.03 (0.052) (0.055) Age 0.06 -0.04 (0.045) (0.044) Education -0.06* 0.07* (0.035) (0.039) Gender (1=Female) 0.04** 0.02 (0.020) (0.022) Church Attendance -0.01 -0.01 (0.030) (0.031) Region (1=South) -0.03 -0.01 (0.022) (0.023) Limited Government 0.01 -0.07* (0.032) (0.034) Constant 0.71*** 0.73*** (0.041) (0.050) Observations 571 571 Source: 2013 CCES *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The column headings indicate the dependent variables, the racial sympathy and the gender sympathy indices. Appendix Table 6: Racial Sympathy and Discriminant Validity Racial Sympathy Racial Resentment Party ID (1=Republican) Limited Government Constant Observations Environment/ Job Trade Off (1=Protect Environment) Keystone Pipeline (1=Support) Increase Border Patrols (1=Increase) Deny automatic citizenship to U.S. born children (1=Deny) 0.08 (0.068) -0.13** (0.061) -0.13** -0.08 (0.111) 0.30*** (0.107) -0.03 -0.05 (0.116) 0.41*** (0.116) 0.18* -0.10 (0.115) 0.62*** (0.104) 0.01 0.03 (0.116) -0.13 (0.114) -0.27*** (0.052) -0.12*** (0.040) 0.48*** (0.073) 571 (0.081) 0.35*** (0.065) 0.72*** (0.145) 554 (0.094) 0.14* (0.071) 0.70*** (0.140) 571 (0.091) 0.23*** (0.070) 0.56*** (0.145) 571 (0.090) -0.26*** (0.068) 0.56*** (0.142) 564 Ban assault rifles (1=For ban) Source: 2013 CCES *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. All variables are coded from 0 to 1. The top column headings display the dependent variables, which are questions about policy opinion. Below these headings, the results for each survey are presented. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were also included in the models: income, age, education, gender, region (South) and church attendance. Appendix Table 7: Racial Sympathy and Support for Public Policies, Contingent on Race Black Beneficiary=1 Racial Sympathy Black Beneficiary x Racial Sympathy Racial Resentment Business Subsidy Business Subsidy Scholarships Scholarships Schools Schools -0.37*** (0.07) 0.11 (0.08) -0.40*** (0.07) -0.04 (0.09) -0.37*** (0.08) 0.29*** (0.09) -0.40*** (0.07) 0.02 (0.08) -0.14* (0.08) 0.45*** (0.03) -0.11 (0.08) 0.21** (0.09) 0.30*** 0.35*** 0.25** 0.29*** 0.06 -0.01 (0.11) (0.11) -0.21*** (0.06) -0.09** (0.04) -0.02 (0.04) 0.63*** (0.09) 818 (0.11) (0.11) -0.32*** (0.06) -0.07 (0.04) -0.08** (0.03) 0.66*** (0.08) 819 (0.12) (0.11) -0.29*** (0.06) -0.15*** (0.05) -0.12*** (0.04) 0.58*** (0.08) 820 Party ID (1=Republican) Limited Government Constant Observations 0.53*** (0.06) 818 0.50*** (0.06) 819 0.33*** (0.06) 820 Source: 2013 CCES *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. The column headings display the dependent variables, which are questions about policy opinion. Coefficients on additional variables included in the models are not shown here for space considerations – the following variables were included in the models: income, age, education, gender, region (South) and church attendance Web Appendix Web Appendix Table 1: Supplemental studies referenced in the manuscript (in chronological order) Date Platform Summer 2012 MTurk Sample Single wave. 323 white respondents. Single wave. 301 white respondents 2 waves. 280 white respondents. Spring 2013 MTurk Winter 2015 MTurk Spring 2015 MTurk 2 waves. 241 white respondents. Fall 2015 MTurk 2 waves. 365 white respondents. Fall 2015 SSI 2 waves. 155 white respondents. Summer 2016 MTurk 2 waves. 103 white respondents. Summer 2017 MTurk 2 waves. 176 white respondents. Main variable(s) explored Racial sympathy and empathy Racial sympathy and egalitarianism The relationship between racial sympathy index and ANES Question, humanitarianism, and egalitarianism. Racial sympathy and other racial attitudes. Racial sympathy and financial donations (to NPAP (Nationl Police Accountability Project)) The relationship between racial sympathy and intensity of black suffering. The relationship between racial sympathy index and ANES Question. Racial sympathy and MIE. The relationship between racial sympathy and counterstereotypes. Web Appendix Web Appendix Table 2: Racial Sympathy and Support for Affirmative Action Affirmative Action 2012 ANES Racial Sympathy Party ID Limited Govt Racial Resentment Stereotypes 2008 ANES Stereotypes Egalitarianism Close to blacks 0.15*** (0.034) -0.01 (0.023) -0.07*** (0.019) -0.49*** (0.035) -0.01 (0.046) 0.15*** (0.034) -0.01 (0.024) -0.06*** (0.019) -0.47*** (0.035) 0.08** (0.032) -0.06*** (0.021) -0.11*** (0.017) -0.41*** (0.031) Egalitarianism Stereotypes Egalitarianism IAT Stereotypes Egalitarianism Contact with blacks 0.15*** (0.040) -0.07** (0.030) -0.04* (0.023) -0.42*** (0.044) 0.09 (0.070) 0.15*** (0.04) -0.07** (0.03) -0.05* (0.023) -0.41*** (0.040) 0.15*** (0.038) -0.04 (0.031) -0.05** (0.023) -0.38*** (0.044) 0.08** (0.041) -0.05 (0.034) -0.01 (0.047) -0.29*** (0.060) -0.10 (0.080) 0.13*** (0.042) -0.04 (0.035) 0.00 (0.049) -0.29*** (0.060) 0.15*** (0.054) -0.07 (0.046) -0.02 (0.063) -0.33*** (0.074) 0.07** (0.034) Close to Blacks 1994 GSS -0.01 (0.050) 0.13*** (0.039) 0.07** (0.029) Implicit Racial Attitudes -0.07 (0.064) Contact with Blacks Constant Observations R-squared 0.74*** (0.031) 5,393 0.303 0.56*** (0.052) 5,360 0.293 0.71*** (0.031) 5,384 0.306 0.74*** (0.031) 5,393 0.303 0.57*** (0.059) 2,169 0.234 0.57*** (0.060) 2,023 0.231 0.39*** (0.126) 494 0.124 0.25*** (0.065) 567 0.144 Sources: 2012 ANES, 2008 ANES, 1994 GSS *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. See note under Table 5. -0.00 (0.029) 0.54*** (0.097) 319 0.191 Web Appendix Web Appendix Table 3: Racial Sympathy and Support for Welfare Welfare 2012 ANES Racial Sympathy Party ID Limited Govt Racial Resentment Stereotypes 2008 ANES Stereotypes Egalitarianism Close to blacks 0.05** (0.026) -0.10*** (0.018) -0.14*** (0.015) -0.26*** (0.027) -0.04 (0.046) 0.04 (0.026) -0.08*** (0.018) -0.11*** (0.016) -0.22*** (0.028) 0.05* (0.026) -0.10*** (0.018) -0.14*** (0.015) -0.27*** (0.026) 0.20*** (0.030) Egalitarianism 1994 GSS Stereotypes Egalitarianism IAT Stereotypes Egalitarianism Contact with blacks 0.05 (0.032) -0.10*** (0.018) -0.14*** (0.015) -0.26*** (0.027) -0.09 (0.062) 0.04 (0.033) -0.05* (0.027) -0.08*** (0.022) -0.21*** (0.040) 0.05 (0.033) -0.08*** (0.029) -0.09*** (0.023) -0.27*** (0.042) 0.06 (0.057) -0.20*** (0.049) -0.06 (0.073) -0.28*** (0.084) 0.09 (0.120) 0.09 (0.060) -0.16*** (0.052) -0.03 (0.076) -0.29*** (0.087) 0.05 (0.067) -0.21*** (0.058) 0.03 (0.087) -0.34*** (0.096) 0.20*** (0.050) 0.12** (0.054) 0.02 (0.027) Close to Blacks 0.10* (0.058) Implicit Racial Attitudes Contact with Blacks Constant Observations R-squared 0.74*** (0.031) 5,393 0.303 0.54*** 0.71*** 0.74*** 0.69*** 0.80*** (0.038) (0.031) (0.031) (0.055) (0.051) 5,403 5,384 5,393 2,169 2,074 0.320 0.306 0.303 0.234 0.216 Source: 2012 ANES, 2008 ANES & 1994 GSS 0.39*** (0.126) 494 0.124 0.43*** (0.101) 455 0.142 *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. See note under Table 5. -0.02 (0.037) 0.64*** (0.134) 349 0.128 Appendix Web Appendix Table 4: Racial Sympathy, Personality and Support for Racialized Public Policies 2012 ANES Racial Sympathy Agreeableness Openness Government Aid to Blacks 0.12*** (0.03) -0.01 (0.03) 0.02 (0.02) Welfare 0.06** (0.03) 0.02 (0.03) -0.00 (0.02) Affirmative Action in Hiring 0.15*** (0.04) -0.06* (0.03) 0.05* (0.03) 2015 MTurk Government Welfare Aid to Blacks 0.16*** (0.06) -0.03 (0.07) 0.02 0.16** (0.06) (0.07) -0.59*** -0.52 Racial Resentment -0.53*** -0.26*** -0.50*** (0.05) (0.06) (0.03) (0.03) (0.03) -0.04 -0.17 Party ID -0.08*** -0.10*** -0.01 (0.04) (0.05) (0.02) (0.02) (0.02) -0.04 -0.13 Limited Govt -0.08*** -0.14*** -0.08*** (0.03) (0.04) (0.02) (0.02) (0.02) 0.21*** 0.49*** Constant 0.72*** 0.70*** 0.63*** (0.06) (0.08) (0.03) (0.04) (0.04) 283 283 Observations 4,943 5,182 5,147 0.626 0.58 R-squared 0.448 0.298 0.299 Sources: 2012 American National Election Study (ANES) & 2015 MTurk Study Humanitarianism *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Cell entries are ordinary least squares regression coefficients (standard errors in parentheses). All variables are coded from 0 to 1. The column headings display the dependent variables, which are questions about policy opinion. Coefficients on additional control variables included in the models here are not shown for space considerations – the following variables were also included in the models: income, age, education, gender, region (South) and church attendance. Web Appendix Web Appendix Table 5: Racial Sympathy and Support for Barack Obama Obama VoteÇ Racial Sympathy Party (1=GOP) Income Age Education Gender (1=Female) Region (1=South) Church Attendance Limited Government Racial Resentment Constant Observations Strong preference for Obama 0.66 0.09* (0.510) (0.048) -5.37*** -0.57*** (0.468) (0.038) -0.60 -0.04 (0.604) (0.051) -1.49*** -0.14*** (0.526) (0.047) -0.68 -0.09* (0.529) (0.046) -0.11 -0.00 (0.245) (0.022) -0.31 -0.01 (0.256) (0.023) -0.38 -0.01 (0.249) (0.023) -0.98*** -0.10*** (0.348) (0.033) -3.48*** -0.26*** (0.643) (0.064) 6.65*** 1.18*** (0.778) (0.067) 1,015 1,015 Source: 2008 ANES Liking Obama Obama Thermometer 0.07** (0.030) -0.40*** (0.027) -0.01 (0.035) -0.04 (0.035) -0.06* (0.032) 0.01 (0.015) 0.00 (0.016) -0.03* (0.016) -0.12*** (0.022) -0.33*** (0.039) 1.09*** (0.037) 1,298 0.06** (0.028) -0.34*** (0.024) -0.02 (0.033) 0.01 (0.027) -0.02 (0.030) 0.02 (0.014) -0.01 (0.014) -0.02 (0.014) -0.10*** (0.020) -0.33*** (0.037) 1.04*** (0.034) 1,303 *** p < 0.01; **p<0.05 ; * p < 0.10. White respondents only; analyses are weighted for national representativeness. Ç The vote choice model uses logistic regression – all other models use OLS. Web Appendix Figure 1: Race Experiment Results, 2013 CCES Figure 2a: Business Subsidies Figure 2b: Scholarships Scholarships .8 .6 .4 .2 -.2 0 Support for Scholarship Policy .8 .6 .4 .2 0 -.2 Support for Business Subsidy Policy 1 1 Business Subsidies 0 .25 .5 Racial Sympathy Poor Condition .75 0 1 .25 .5 .75 Racial Sympathy Poor Condition Black Condition Figure 2c: Schools .8 .6 .4 .2 0 -.2 Support for School Policy 1 Schools 0 .25 .5 .75 Racial Sympathy Poor Condition For full model specification, see Appendix Table 4. Black Condition 1 Black Condition 1