The Rise of Racial Diversity and Voting in U. S. Presidential Elections

• In Event: Nationalization and Other Long-Term Trends in Voting

Abstract

Title The Rise of Racial Diversity and Voting in U. S. Presidential Elections

Summary

Using census data and ANES data from the 1956-2012 presidential elections, we show how the rise of racial-ethnic voting in recent presidential elections can be directly linked to the growing levels of racial-ethnic diversity that are transforming most U. S. counties. .

Abstract

We explore how America’s growing racial-ethnic diversity is reshaping the racial-ethnic identities and candidate preferences of voters in U. S. presidential elections. Using census data and ANES data from the 1956-2012 presidential elections, we show how racial-ethnic diversity has always influenced the presidential vote and how its effects compare with other predictors in the so-called funnel of causality such as racial-ethnic, religious, and social class identities; party identification; economic issues and ideology; and candidate images. In particular, we test the hypothesis that the rise of racial-ethnic voting in recent presidential elections can be directly linked to the growing levels of racial-ethnic diversity in U. S. counties. We also show how the racial-ethnic identities of voters are activated by their immediate racial-ethnic environments and how their votes are influenced by the national political context and political mobilization of voters. Our results show how racial-ethnic diversity increases the probability of voting Democratic and decreases the probability of voting Republican. At the same time, we demonstrate how whites are more likely to vote Republican in racially and ethnically diverse counties. Between 1956 and 1976, white racial identities increased the Republican presidential vote by 4.1 percent and their interaction with the nation’s increasingly diverse voting environments increased it by another 15 percent. Between 1980 and 2012, racial identities increased the Republican vote by 8.9 percent and the interaction of racial identity and the racial-ethnic environment increased it by an additional 1.9 percent. These estimated effects are independent of all the other variables that are generally subsumed in the funnel of causality including dummy variables for the level of economic dissatisfaction in each election year. Among other things, we conclude that the standard voting model is deficient because it soes not take into account the nation’s growing racial-ethnic diversity and the interaction of voters’ racial-ethnic identities with their voting environments.

Author

• Joel A. Lieske