The Statistics of Swing: The Demographics and Political Properties of U.S. Presidential Swing States
Presidential swing states, particularly the demographic and political features that predict them, have been under-researched in the field of political science. This thesis examines what features might predict swing states in U.S. presidential elections. After conducting a literature review and determining what the best measure of a swing state is, ten hypotheses are proposed and tested using regression models. The statistical results indicate that states with a high number of middle class households, a lack of religious adherents, and a low electoral vote per capita count are more likely to be swing states. This is substantial because it can help guide political scientists, pollsters, and political campaigns in future elections when deciding what states should deserve the most attention.