Rethinking Religion and Political Participation: The Case of Voting Among Religiously Unaffiliated Americans

Is civic disengagement correlated across institutions? One case of this question is a long-observed "secular voting gap" where religiously unaffiliated Americans are less likely to vote than their affiliated counterparts. This work often uses self-reports or exit polls that cannot measure variation within the unaffiliated. Using an improved measure of validated voter turnout in four presidential election years (2008, 2012, 2016, and 2020), I find estimates of the secular voting gap are attenuated by demographic controls. More importantly, the mechanism that explains this finding is that more frequent church attendance associates with a lower probability of turnout among respondents who are unaffiliated, and results vary by voting method. These results support a theory of civic disengagement as a domain-specific process and demonstrate the substantive value of revisiting classic findings about religion and political behavior amid social change.