ABSTRACT
American politics has entered, what Michael Tesler calls, a “most racial period,” causing racial attitudes to be increasingly salient. This shift provides an opportunity to examine the consequences to white Americans’ political behavior when their racial attitudes are easily activated. In this paper, we examine the influence of racial attitudes on (1) white Americans’ presidential vote choice and (2) how racial attitudes, in this period, may undercut the influence of partisanship on vote choice. We use data from the American National Election Studies Time Series Survey conducted from 2004 to 2020 to explore these questions. First, we find that anti-Black attitudes increasingly predict white vote choice. Next, we find that during this period, racially resentful, white Democrats became less likely to vote for the Democrat. This was not the same for white Republicans who, regardless of their racial attitudes, did not shift to the Democratic party. Finally, we find that white Republicans appear to hunker down and become significantly less likely to split their vote as they express more anti-Black attitudes. White Democrats become significantly more likely to jump ship as they endorse more anti-Black attitudes, revealing that racial attitudes were influential enough to overcome party loyalties during this period.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
2 The 2016 ANES removed the Intelligence assessment and instead asked respondents to place Black people and white people on a scale of Peaceful (1) to Violent (7). So, in the 2016 and 2020 Stereotype Index this measure was used in place of the Intelligence assessment.
3 In 2016 and 2020, this is work ethic and tendency towards violence.
4 The 2020 ANES has not yet included respondents’ state of residence; so, we were unable to control for Southern residency in our 2020 models.
5 The coding of all of these controls are described in greater detail in Table B in the Appendix.
6 For models with fewer controls, see Tables C through F in the Appendix.
7 These results can be found in Tables H and I in the Appendix.
8 The difference in the size of the racial resentment coefficients between all other years, was not statistically significant. See Table H in the Appendix.
9 The change in the effect of the stereotype index is not statistically significant between no other years. See Table I in the Appendix.
10 The full results of this analysis can be found in the Appendix, Tables J and K.
11 The racial resentment coefficient for white Democrats was positive and statistically significant (P < 0.00).
12 The stereotype index coefficient for white Democrats was positive and statistically significant (P < 0.01).
13 The results of the logistic regression for all five years can be found in the Appendix, Tables M and N.
14 The finding that the effect of racial resentment on white, Republican split-ticket voting is larger than the effect of the stereotype index makes sense, given the concerns of social desirability bias. If white, Republican respondents are not expressing their true stereotypes of Black people relative to white people, we would anticipate that the effect of the stereotype index on voting behavior would be underestimated.