For shame! The effect of community cooperative context on the probability of voting

The question of why some people vote in American national elections and others do not has been the focus of a vast literature in social science. Numerous empirical regularities have been established, such that we now know "who votes" and who doesn't, in the sense that various demographic characteristics -- most notably education -- are strongly correlated with turnout (Wolfinger and Rosenstone 1980; Teixeira 1987, 1992). A consensus on "why", in the form of theories and evidence on the motives of individuals, has been slower to emerge. This study builds on previous work emphasizing the political relevance of civic norms prescribing social cooperation. In this analysis, we use a county-level variable -- mail-in census response rates -- to measure the strength of civic norms in counties represented in the 1992 American National Election Study (NES), finding that the likelihood of one’s voting increases with the county’s census response rate, controlling for the usual set of factors associated with turnout. We explore one information source by which people may learn about community expectations, the newspaper.