SYMBOLIC RACISM AND OPPOSITION TO BLACK LIVES MATTER

            In the summer of 2014, several police shootings of unarmed black men sparked a wave of protests throughout the United States. Several of the demonstrations were organized by the group Black Lives Matter (BLM), which had formed only two years earlier in response to the acquittal of George Zimmerman for the murder of a black teen named Trevon Martin.[1] Many of the police shootings in 2014 had been caught on camera and circulated quickly on social media, igniting a contentious political debate in American society. Supporters of BLM have claimed that the political system—including the police and the judiciary—are biased against African Americans. These claims are not without merit. Numerous academic studies indicate the existence of prejudice against blacks in the judicial system, the market place, and many other areas of political life (Bobo, Kluegel, and Smith, 1997). However, opponents of BLM have claimed that African Americans are involved with more altercations with police because of the culture of the black community itself, which they claim is conducive towards more criminality and anti-police attitudes. Given the importance of BLM in American political life and the intense opposition to it, it is critical that we gain a better understanding of who supports and opposes the group, and the reasons why they do so.

            There are two theories that can potentially explain support and opposition to BLM within the American population. First, attitudes towards policing is one explanatory variable. Based on this theory, conservative Republicans are more inclined to oppose BLM because of their support for aggressive policing, the war on drugs, and long prison sentences whereas liberal Democrats are more likely to support the group because of their opposition to the criminalization of narcotics, the militarization of police departments, and the rapid expansion of America’s prison system. While conservatives may view recent police shootings as officers defending themselves in a violent environment riddled with crime, liberals may see it as the police abusing their power and using excessive force where it is unwarranted. In other words, the political divide over BLM might be an ideological one—specifically, over law enforcement policies in America.

            Secondly, attitudes towards BLM might be based on the politics of race, which are heavily influenced in America by a phenomena known in academia as symbolic racism. This form of racism is very different from the kind of prejudice that existed towards African Americans during the Jim Crow era over a half century ago. Previously in American history, racism was biological in nature, and the majority of white Americans viewed blacks as being naturally inferior (Tarman and Sears, 2005). These beliefs were used to first justify a system of segregation in the South. However, following the migration of hundreds of thousands of African Americans to urban areas in the early twentieth century and the rise of the Civil Rights movement, the Jim Crow system collapsed as did the theory of biological racism that legitimized the system. In 1964, Lyndon Johnson signed the Civil Rights Act into law, and formal segregation came to an end. Nevertheless, sharp socioeconomic and political divides remained between America’s white and black communities. To bridge the racial gap, the majority of African Americans and liberal whites have supported policies favoring the redistribution of wealth to the black community and affirmative action as a redress for several centuries of slavery, segregation, and political oppression. In contrast, conservative whites have viewed slavery and segregation as part of a distant past and problems in the African American community as being a result of cultural deficiencies. From this perspective, individual blacks should embrace the Protestant ethic, pull themselves up by their own bootstraps, and work harder. By the 1980’s, the biological racism that once predominated in the white community had been replaced by a cultural form of prejudice scholars now known as symbolic racism. Conservative whites have used this ideology to justify their opposition to policies that would redistribute wealth to disadvantaged black communities in order to protect their own economic interests and privileged position in society. Several scholars also argue that white liberals, conscious of this history of inequality and inflicted with guilt, overcompensate for their own innate prejudices and show bias towards African Americans (Nail, Harton, and Decker, 2003). This form of prejudice is known as adverse racism. Consequently, white conservatives might oppose BLM because they view the black community as being responsible for their own problems; meanwhile, blacks and liberal whites might support BLM because they believe that the political and economic system is rigged against the African American community, so problems within that community are a consequence of political and economic structures created by whites and not the responsibility of individual blacks. This may explain why conservatives and liberals have been so sharply divided over every salient incident involving a police shooting of an unarmed black man regardless of the particular nature of each individual shooting.

            To determine which of these theories explains the variance in support for BLM, I will conduct a multi-regression analysis using data from the 2016 American National Election Survey (ANES), which contains a stratified sample of the American population. After eliminating missing responses, the data contains 2,045 respondents. The dependent variable is level of support for BLM, and this will be measured with the ANES feeling thermometer that measures support of BLM on a scale of 0 to 100. Below is a histogram showing the variation on the dependent variable:

With a mean of 48.4 and a standard deviation of 32.1, there is a significant degree of variation in terms of the level of support for the movement.

            The independent variables are attitudes towards policing in America and symbolic racism. To operationalize the first independent variable, I will use the ANES feeling thermometer for the police, which askes respondents to state their level of support for the police on a scale of 0 to 100. A negative coefficient will indicate support for the theory that ideological stances on policing affect one’s position on BLM. In addition, dummy variables for political ideology (conservative and liberal) and party identification (Democrat, Republican, and Independent) were also added.  The correlation of these variables with opinions on BLM will strengthen the case for the first theory.

            The second independent variable, symbolic racism, will be measured using five 2016 ANES survey questions, which have been frequently used by other academics to measure the concept (Tarman and Sears, 2005). The questions are as follows:

1.       ‘Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.’

2.       Are you for or against preferential hiring and promotion of blacks?

3.       On a scale of one to seven, how much should the government help blacks?

4.       Do you favor, oppose, or neither favor nor oppose allowing universities to increase the number of black students studying at their schools by considering race along with other factors when choosing students?

5.       ‘It’s 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.’ Agree/disagree.[2]

Some of these questions ask respondents to either agree or disagree while others ask respondents to give their opinions on an ordinal scale of either one to five or one to seven. With the exception of question three, negative coefficients for these variables would support the symbolic racism theory. In addition, I tested for Jim Crow racism or old fashioned racism by using the ANES feeling thermometer for blacks. It should be noted that only a small minority of citizens hold strong anti-black feelings towards African Americans so this variable is less consequential in this analysis. This is indicated in the histogram below:

Symbolic racism—not old fashioned racism—is the main independent variable of interest. The feeling thermometer for whites was added as a placebo as it is not expected for attitudes towards whites to be correlated with support or opposition to BLM. Finally, a dummy variable on race was also added with black being coded as one and white being coded as zero. It is expected that racial identity will be correlated with attitudes towards BLM.

            Numerous control variables were also added to ensure that other factors are not explaining the variation on the dependent variable. Demographic variables such as age, education (labeled as BA, which indicates whether or not the respondent has a college education or not) and gender (female is coded as one and male as zero) will be included in the multivariate analysis. In addition, I controlled for the variable authoritarianism by using the following question from the 2016 ANES survey:

When protestors get ‘roughed up’ for disrupting political events, how much do they generally deserve what happens to them?

It is possible that attitudes towards BLM might be affected by attitudes towards democracy and specifically people’s right to protest. A negative coefficient would indicate support for this alternative theory. Finally, a variable testing the interaction between gender and party identification was also added on the assumption that male Republicans would have a more negative attitude towards BLM than female Republicans, and female Democrats would have a more positive attitude towards BLM than male Democrats.

            The multivariate analysis is shown below and standardized variable coefficients are on the right side of the chart. It is my hypothesis that the independent variables measuring symbolic racism will explain a greater degree of the variation in support for BLM than the variable measuring attitudes towards the police. The results of the multivariate analysis are as follows:

Regression Model for Black Lives Matter
	Coefficients	Standard Error	T Scores
Republican	-2.56	(1.88)	-1.36
Democrat	8.94	(2.14)	4.19	***
Ideology	-2.24	(0.47)	-4.75	***
Income	-0.10	(0.07)	-1.45
Female	6.93	(1.28)	5.40	***
Female*Democrat	-1.30	(1.93)	-0.67
Age	0.04	(0.03)	1.31
Education	-0.84	(1.05)	-0.81
Race	9.34	(2.02)	4.63	***
Slavery Question	-2.75	(0.45)	-6.17	***
Black Work Ethic Question	1.62	(0.49)	3.34	***
Police Thermometer	-0.05	(0.03)	-1.84
Authoritarianism	-1.93	(0.43)	-4.49	***
Affirmative Action at Uni	-1.96	(0.33)	-6.00	***
Black Welfare Question	-2.36	(0.38)	-6.27	***
Affirmative Action	-4.61	(1.50)	-3.08	***
White Thermometer	-0.05	(0.03)	-1.56
Black Thermometer	0.23	(0.03)	7.52	***
_cons	70.99	(4.56)	15.57
Observations	1999.00
R-Squared	0.58
*p<.05
**P<.01
***p<.001

Overall, these variables explain approximately 58 percent of the variation in attitudes towards BLM according to the adjusted r-squared value. The results also indicate strong support for the symbolic racism thesis. All five questions are statistically significant (P<.001) and indicate that respondents who hold a symbolically racist ideology also tend to oppose BLM. Furthermore, racial identity is also heavily correlated with support for BLM as blacks are far more likely than whites to support BLM. Not surprisingly, old fashioned racism is also correlated with opposition to BLM. As predicted, the variable measuring attitudes towards police provided very weak evidence to support the first theory. The coefficient for the police thermometer is correlated in the right direction but barely missed statistical significance with a p-value of .066. Furthermore, while being a democrat is correlated with support for BLM with statistical significance, being a Republican is surprisingly not strongly correlated with opposition to BLM and was not statistically significant. However, being both conservative and liberal were strongly correlated with the dependent variable. This indicates that conservative and liberal differences over race and not policing had the largest effect on the dependent variable. As for the control variables, age and education also have no effect on the dependent variable, although females are far more likely than males to support BLM. The interaction variable for gender and political party identification was also statistically insignificant. Finally, attitudes towards authoritarianism are correlated with opposition to BLM, and this has some potential implications that are outside the bounds of this study.

            Robustness checks were also conducted for heteroscedasticity and multicollinearity. To check for heteroscedasticity, the Breuch-Pagan Test was conducted. The test resulted in a p-value of .0057, so the regression analyses was conducted again using the Huber Sandwich estimator to ensure statistical robustness. The results did not differ with statistical significance from the original regression model. To check for multicollinearity, the variance inflation factor was analyzed, but none of the coefficients were above 10.

            Overall, the results show that blacks, liberals, and women are statistically more likely to support BLM whereas whites, conservatives, and men are more likely to oppose BLM. The reasons why have less to do with the current political debates over policing in America and have far more to do with attitudes towards race.

            Nevertheless, there are several problems with the internal validity of this study that might have created biased results. Previous studies have indicated that feeling thermometers on police may not measure how respondents feel about “local” policing but the “global” idea of police (Scaglion and Condon, 1980). While respondents may support liberal criminal justice reforms, they may not have an overall negative feeling on the police as an institution. This may have created biased results. More specific questions on attitudes towards policing and criminal justice are needed to improve upon this study. Furthermore, other control variables should have been used as well. It is possible that conservatives oppose affirmative action and the redistribution of wealth to the black community due to their stance on the welfare state.  However, based on the results of previous studies, I have reason to doubt that including these variables will affect the results (DeSante, 2013), but they should have been included anyway as a robustness check. Overall, based on the strong correlation between all five symbolic racism questions and the dependent variable, I can confidently conclude that attitudes towards race had a very real effect on support and opposition to BLM.

Work Cited

Christopher D. DeSante. 2013. “Working Twice as Hard to Get Half as Far: Race, Work Ethic, and America's Deserving Poor.” American Journal of Political Science, 57(2): 342-356.

Kinder, Donald and David Sears. 1981. “Prejudice and Politics: Symbolic Racism versus Racial Threats to the Good Life," Journal of Personality and Social Psychology, 40(3): 414-431.

Lowery, Wesley. January 17, 2017. “Black Lives Matter: Birth of a Movement,” https://www.theguardian.com/us-news/2017/jan/17/black-lives-matter-birth-of-a-movement.

Nail, Paul R., Helen C. Harton, and Brian P. Decker. 2003. "Political Orientation and Modern versus Aversive Racism: Tests of Dovidio and Gaertner's (1998) Integrated Model." Journal of Personality and Social Psychology, 84(4): 754.

Scaglion, Richard and Richard Condon. 1980. “The Structure of Black and White Attitudes towards Police,” Human Organization, 39(3): 280-283.

Sniderman, Paul M., Thomas Piazza, Philip E. Tetlock, and Ann Kendrick. 1991. “The New Racism.” American Journal of Political Science, 35(2): 423-447.

Tarman, Christopher and David Sears. 2005. “The Conceptualization and Measurement of Symbolic Racism.” The Journal of Politics, 67(3): 731–761.

Terkildsen, Nayda. 1993. “When White Voters Evaluate Black Candidates: The Processing Implications of Candidate Skin Color, Prejudice, and Self-Monitoring.” American Journal of Political Science, 37(4): 1032-1053.

“User’s Guide and Codebook for the ANES 2016 Time Series Study,” http://www.electionstudies.org/studypages/anes_timeseries_2016/anes_timeseries_2016_userguidecodebook.pdf, 2016.

Weizer, Ronald. 2004. “Race and Perceptions of Police Misconduct,” Social Problems, 51(3): 305-25.

---

[1] Lowery, Wesley. January 17, 2017. “Black Lives Matter: Birth of a Movement,” https://www.theguardian.com/us-news/2017/jan/17/black-lives-matter-birth-of-a-movement.

[2] “User’s Guide and Codebook for the ANES 2016 Time Series Study,” http://www.electionstudies.org/studypages/anes_timeseries_2016/anes_timeseries_2016_userguidecodebook.pdf, 2016.