Developing organizational theory to explain voter behavior, we explore the relationship between the population structure of local newspapers and electoral turnout. We argue that diversity in newspaper identities increases turnout in ways suggested by certain inter-organizational processes engendering public debate and political activation. We also argue that populations containing many newspapers with opposing political identities will be especially likely to enhance turnout. We analyze a century-long panel of bi-annual observations on 48 local communities in the US to test hypotheses about the consequences of identity-based organizational diversity in the press on turnout in congressional elections. We examine both political identities and the specialized general contents of newspapers. From fixed effects estimates of linear panel regression models, we find that overall identity-based diversity of the press shows either no relationship to turnout (political diversity) or an unexpected negative relationship (content diversity). However, we also find that counts (densities) of papers with specific political identities often do affect turnout. Moreover, we find interaction effects of densities of papers with political identities at opposing ends of the political spectrum, consistent with an "opponent activation" process of newspaper interaction. Finally, analysis shows that this process interacts with the expected vote margin and with the coincidence of presidential elections.
