Thinking Temporally about Social Influence: Examining Networks over Time.

A sizable literature has demonstrated the importance of social networks for both electoral decision-making (e.g., Huckfeldt and Sprague 1995) and political participation (e.g., Mutz 2002). However, the question of how networks operate over time has been largely unexamined. Do voters discuss politics with the same individuals during a general election as during a primary? Moreover, how do the characteristics of an individual's network - in terms of density, composition, and the content of political discussion - change over time? And what are the potential behavioral consequences of such changes (or non-changes)?I address these questions using an original longitudinal study following a group of Franklin County, Ohio Republican primary voters. The sample of voters was first surveyed in the summer of 2006 after the gubernatorial primary; the follow-up survey was conducted in late 2006 after the general election. Given that both the original survey and the general election follow-up were specially designed to tap the frequency, content, structure, and context of interpersonal political discussion, the data are particularly appropriate for answering the questions posed. This paper makes several contributions to the study of voting behavior and public opinion. The county-wide network panel study is among the first to examine voters' social networks over time and across multiple electoral contexts, and therefore provides a first look at how people's social circles react to political developments. By thinking temporally about social influence, this paper provides a better understanding of how social channels enable (or inhibit) information flow and mobilization. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]