Document Type


First Faculty Advisor

Richard Holtzman

Second Faculty Advisor

Alicia Lamere


voter behavior; identity; rational choice; behavioralism; 2016 election

Rights Management



Studying voter behavior through the lens of identity is central to making sense of the 2016 presidential election. The traditional models for explaining voter behavior are rational choice and behavioralism. The former is grounded in instrumental partisanship and a voter’s issue positions, with the latter grounded in an expressive, psychological attachment to partisanship. More recent, social identity theory related models discuss voter behavior through group belonging and the partisan mega-identity (Mason 2018). My analysis used the ANES 2016 Time Series Study. To measure a voter’s issue positions, I created a new Identity Index alongside the expansion of an established Issue Index. To measure the more expressive aspects of voter behavior, the demographic identities of race, age, class, and gender were used, as well as feeling thermometer measures such as a respondent’s feelings towards police, Black Lives Matter supporters, scientists, and other groups. My analysis showed that the Identity Index was a better proxy for determining voter behavior in 2016 than the Issue Index. The demographic variables, except for gender, were also significant in both models tested. Through the feeling thermometers, certain groups could be identified as having more influence in voter behavior than others. MANOVA analysis demonstrated that the feeling thermometer scores also vary based on different demographic interactions. The findings of this research demonstrate that the consideration of identity in the rational choice and behavioralism models is crucial to explain voter behavior in 2016.