Enduring Disappointment: How Voters' Affective Attachments Sustain Partisan Support in Young Democracies

Abstract

Voters in the Global South often parse political decisions guided by a sense of economic utility over alternatives. However, not only is there abundant evidence from other democracies that voters can be mostly motivated by non-economic incentives, but it appears increasingly likely that the same types of incentives exist alongside pecuniary motives in the Global South. This dissertation focuses on the nature and role of psychological or expressive incentives that extend from a voter’s internalized affective connection to a politicized social group, or their social identity. Throughout, it draws primarily on original survey and interview data collected in Ghana between 2018 and 2022. The first empirical chapter explores the nature of these identities relative to prevailing concepts and measures of identity by introducing a multi-item measure of group identification. Observational results show that not only do social identities appear to differ in their correlates but explain variation in key political attitudes not captured by conventional (binary) measures. The second leverages a lab-in-the-field experiment to understand the impacts of activating a social identity. My results suggest that priming identities increases individuals’ group orientation and may be an important avenue for parties to sustain support. The third and final empirical chapter reanalyzes the same experiment to consider whether specific identities provide the same behavioral incentives in politics. Comparing respondents primed to feel either more ethnic or partisan, I find that these identities differ significantly in their ability to further group-oriented behaviors and biases. I conclude that scholars should avoid collapsing these two identities and afford greater attention to the underlying social construction and context of an identity when studying their impacts on behavior. Together this points to a broader need of comparative political scientists to identify the limits of instrumental motives in voting, what lies beyond those limits, and how (and for whom) economic and non-economic motives interact. Toward policy actors working to improve governance, it suggests a need to adopt richer models of behavior that afford voters psychological and affective biases as they evaluate candidates, make electoral decisions, and process political information.