The Importance of Ideological Domains in American Presidential Elections

• In Event: Executive Politics and Public Opinion

Abstract

Previous studies have found positive results for spatial voting models for one ideological dimension. However, there is great evidence that American ideology is multi-dimensional. Particularly, I find evidence for five dimensions: economics (the government's role in the economy), race, morality, military strength, and immigration. I investigate if these ideological dimensions are important for a person's vote choice in American presidential elections.

But one should only expect an ideological dimension to be significant in a logistic regression of presidential vote choice if the candidates are significantly differentiated on that dimension. Thus, I do a content analysis of presidential candidate convention nomination speeches to estimate candidates' positions on each of these dimensions. Using rigorous criteria, I do this content analysis myself along with two other research assistants (our results have a high degree of correlation). I also estimate citizens' latent opinions on these dimensions using data from the ANES.

I find that a dimension is influential to a person's presidential vote choice when there is a significant difference between the candidates on that dimension. However, the dimensions differ in their effect. The more emotive dimensions of race, immigration, and morality have had a much more significant effect on a person’s presidential vote choice (corresponding to a large increase in the difference between the candidate positions on these dimensions in recent elections).

Author

• Marcos Menchaca, University of Texas at El Paso