We find that in biracial elections (i.e., a Black candidate and a White candidate compete in an election) in the United States, in-group processes among Whites significantly decrease votes for African-American candidates and approval of the first Black president. As expected by Social Identity Theory and the existing literature, we also find that out-group processes (specifically racial resentment) are more influential when a Black candidate and a White candidate compete in an election. Importantly, the in-group processes significantly affect vote choice above and beyond the out-group processes that have garnered so much scholarly attention. We also find that white racial identity reduced President Obama’s approval, in turn reducing the vote shares of Democrats in Congressional elections across the country.
Using the full Social Identity Theory (SIT) framework, beyond the out-group negativity previously studied, provides new insights. Social Identity Theory focuses on both in-group processes and out-group processes. However, most political science research on group conflict focuses on out-group processes, the processes that result when people see “them” less positively than “us”; factors such as racial resentment and negative stereotyping of African-Americans. In contrast, we test whether in-group processes (specifically strength of White racial identity among self-reported Whites) influence vote choice in seven electoral settings.
Our findings relate to previous research. Considering such in-group processes helps political science move beyond yes or no, and gets to when, regarding debates about race of candidate effects and whether the major parties have become ‘racialized’ (i.e. perceived as the pro/anti-Black and anti/pro-White). In particular scholars have debated whether the Democratic Party was ‘racialized.’ We find that the degree to which in-group processes lead Whites to vote against Democrats varies depending on the race of the candidates. While it remains to be demonstrated if advertising and/or campaign messages can make White in-group identity influential, we show that a Black candidate can activate this in-group process. Furthermore, future research about race of candidate should consider both in-group and out-group processes.