On February 5, 2026, Natalie Letsa, Associate Professor in the Department of Political Science at the University of South Carolina and previously a 2018-19 CDDRL postdoctoral fellow, presented the argument and contributions made in her book, The Autocratic Voter: Partisanship and Political Socialization under Dictatorship. The book addresses the question: Why do some citizens choose to get involved in politics, while others do not? And, for those who get involved, why do some support the opposition, while others support the ruling party?  

In the context of electoral autocracies, the book argues, social identity theory provides a model of partisanship, rather than rationalist or materialist theories. This argument rests on three key parts:

• Part 1: “Ruling Party Partisan" and “Opposition Partisan” are unique social identities that have common meaning across electoral autocracies.
• Part 2: Processes of partisan socialization occur within social networks.
• Part 3: The political orientations of social networks are constrained by the structural homophily created by the political geography of electoral autocracies.
   

Within the idea that “ruling party partisan” and “opposition partisan” are social identities, ruling party partisans trust the regime and believe it is democratic, whereas opposition partisans do not. Amongst these identities, out-group animus occurs, but opposition partisans experience stronger animus. Letsa looked at two questions from the World Values Survey: "How democratic is your country today?" and “How satisfied are you with the political system?” Results found that partisans in electoral autocracies are divided on these questions, whereas partisans in democracies are not.

On social networks, Letsa argues that people raised by partisans are more likely to adopt partisan identities, and that people with party activists in their networks are more likely to do so. Lastly, Letsa asserts that people with more politically homogenous social networks are more likely to adopt partisan identities. When conducting an original survey in Cameroon, Letsa found that citizens with social networks that fully support the opposition are five times more likely to be opposition partisans themselves than citizens with no opposition partisans in their social networks.

On geography, Letsa argued that opposition strongholds are severely constrained in electoral autocracies, producing extreme structural homophily. Because of this structural homophily, the political orientation of a social network is largely determined by its physical location. However, political geography does not only affect the identities of partisans in party strongholds; it also predicts the political beliefs of nonpartisans and out-party partisans in these places. To test this, Letsa examined survey data from Cameroon and found that the average respondent living in an extreme ruling-party stronghold has a network composed of 29% opposition supporters. The average respondent living in an extreme opposition party stronghold has a network made up of 59% opposition supporters.

The Autocratic Voter offers a unified theory of partisanship that explains both support for the ruling party and for the opposition. It presents a new framework for understanding public opinion and political behavior in electoral autocracies. Finally, it bridges research on democracy and autocracy.