Social capital, social heterogeneity, and electoral turnout

Abstract Among the many studied determinants of voting, we predict that i) increased social capital will be positively associated with turnout, while increased heterogeneity will be negatively associated; ii) both factors will work through their influence on the costs of information gathering and on the social norms of voting; and iii) that heterogeneity will interact with social capital in its association with turnout. We test these predictions at the extremely fine “meshblock” level by regressing New Zealand voter turnout in its 2017 national election on its 2013 census characteristics. We use roughly 40,000 meshblock volunteering rates to measure social capital, and heterogeneity based primarily on ethnic fragmentation. We find social capital is positively associated with voter turnout, while heterogeneity is negatively associated. We find robust evidence consistent with ethnic heterogeneity working through information costs and social norms, but less so social capital. We also find a robust interaction between social capital and heterogeneity in their association with turnout, consistent with ethnic heterogeneity raising bridging social capital that has a stronger association with turnout than in‐group bonding social capital.