In this paper I draw on survey data collected in a comparative study (Comparative National Election Project) to analyze the effects of structural and content based features of two types of social settings in which individuals are embedded in their everyday lives on their likelihood of participation in elections. Intimate and generic social settings roughly correspond to ‘strong’ and ‘weak’ ties, or ‘bonding’ and ‘bridging’ social networks. Results suggest that compared to people who encounter only political disagreement in conversations carried out in intimate settings, those who discuss politics with intimate others with whom they agree are more likely to vote. Exposure to a mix of political views appears to play a demobilizing role when it originates in conversations carried out in generic settings and to have a mobilizing effect when it emerges in conversations carried out in intimate settings, in countries with longer democratic experience. Drawing on empirical evidence from extant studies the paper speculates that compared to political talk that occurs in generic settings, exchanges of political views that take place in intimate social networks might stimulate real debate, leading to an increase in political engagement as a result of reflection rather than of social conformity. On the contrary, mixed political cues encountered in generic networks might be less often followed up in additional debates and, therefore, exposure to political diversity in generic settings might lead to increased ambivalence, delay in voting decision making, and finally abstention from voting.
