White candidate support continues amidst explicit and implicit white identity cues

Abstract

ABSTRACTTo what extent are appeals to pro-white sentiment effective in American politics? Does it matter if such appeals are implicit or explicit? We investigate the degree to which white Americans are less likely to vote for candidates that use explicit appeals to white identity than they are to those who use more implicit appeals to white identity or to out-group animus. In doing so, we provide a crucial supplement to recent work on this topic, which finds that the implicit/explicit model is becoming less relevant for whites, as well as to research on the increased relevance of white in-group attitudes for white voter behavior. To examine the effect of such white identity appeals, we conduct a survey experiment with a national sample of 2746 white Americans. We find that respondents accept both explicit and implicit white identity appeals at the same rate as out-group appeal. Additionally, we see evidence that those who identify strongly as white are more likely to support candidates who engage in racial messaging. Our findings have broad implications for our evolving understanding of racial appeals in American politics, as well as the role of white identity in contemporary political discourse.KEYWORDS: Racial politicscommunicationsocial identitywhite identity Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 https://www.cnn.com/2016/01/22/politics/donald-trump-retweet-white-genocide/index.html2 https://www.npr.org/2018/08/23/641181345/heres-the-story-behind-that-trump-tweet-on-south-africa-and-why-it-sparked-outra3 https://www.npr.org/2020/09/30/918483794/from-debate-stage-trump-declines-to-denounce-white-supremacy4 https://www.washingtonpost.com/news/powerpost/wp/2017/12/27/ryans-pro-white-primary-foe-denounced-by-breitbart-after-his-anti-semitic-tweets/5 https://www.washingtonpost.com/magazine/2022/01/04/jd-vance-hillbilly-elegy-radicalization/6 Specifically, we dropped from the survey all those respondents who completed it in less than one third of the median time. This was in line with Lucid's recommendations about how to deal with speeders, satisficers, and inattentive respondents. However, results are substantively the same regardless of whether speeders are included as shown in Appendix 5.7 To check the robustness of our results, we also collapsed this variable into a binary indicator that coded replies as 0 if respondents indicated “Not at all likely” and 1 otherwise. Our results are robust to this alternative measurement strategy.