Date of Award

Fall 10-1-2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Political Science

First Advisor

Stokes, Susan


Partisan identification is intimately related to many central aspects of political behavior. It influences vote choice, voters’ policy positions, and the manner in which voters process new political information. Partisanship has the power to undermine politicians’ accountability and it can even trump voters’ democratic values in polarized societies. Yet large gaps remain in our understanding of the relationship between partisanship and another key subject of political behavior: voter participation. We know that partisans are more likely to turn out to vote than their non-partisan counterparts; but why are partisans more likely to turn out? Are partisans simply people who are more interested in politics (and, therefore, more likely to participate)? Or is there a direct, causal relationship between partisanship and turnout? In this dissertation, I examine competing theories of why partisans participate more than non-partisans. With a series of novel empirical tests aimed at causal identification, I clarify the mechanisms underlying the relationship between partisanship and turnout. Partisanship mobilizes voters by introducing additional incentives to vote, separate from the intensity of their policy preferences. Partisanship generates expressive incentives to vote and engenders a sense of partisan duty — an obligation to do one’s part to contribute to the party’s success. When voters adopt partisan identities, they conceive of themselves as part of a larger group. These group identities fundamentally alter the calculus of voting, facilitating a cooperative logic of turnout. In contrast to oft-made claims in the political behavior literature, I find that the causal chain does not run in both directions: voting does not foster partisan identities. Prior work posited that the act of voting makes people more likely to adopt partisan identities, either through a process of political learning or a desire to resolve cognitive dissonance. The hypothesis that voting fosters partisanship underlies the dominant theory of how compulsory voting laws shape partisanship. Thus, my finding that voting does not foster partisanship calls for a new approach to understanding partisan dynamics under compulsory voting. In the second part of this dissertation, I present a new theory of party-voter linkages under compulsory voting. In both compulsory and voluntary voting systems, parties face a breadth-versus-depth tradeoff in their outreach to voters. When voting is voluntary, parties prioritize depth: they work to build stronger partisan identification among a smaller subset of the population. These stronger partisan identities are necessary to ensure that a party’s supporters are motivated to show up on Election Day. In compulsory voting systems, where parties need not concern themselves with extensive mobilization efforts, breadth becomes more important than depth. Parties competing in compulsory voting systems must win over a larger share of the population, since would-be abstainers are compelled by law to participate. But they don’t face the same burden of mobilizing their supporters, so they don’t need to invest in building ties as strong as those in voluntary systems. The result is that more voters identify with parties in compulsory voting systems, but the strength of their identification is, on average, weaker.