The Politics of Voting: Using Network Analysis to Explore the Role of Status and Reciprocity in Politicians' Voting Behavior

Introduction Previous work in the analysis of politicians' voting behavior has proposed that voting is often influenced by a combination of party, and one's relative status within a particular governing body (Cho and Fowler 2010, Clark, Pelika, and Rigby 2010, Porter et al. 2005). These papers used a variety of techniques to develop networks of politicians, with the most notable being those that used bill co-sponsorship, name occurrence in major newspaper articles and serving on committees together as the criteria for establishing “friendship” between two politicians. These studies found that the level of seniority and one's political affiliation have differing impacts on voting behavior depending on placement within the House or Senate. Furthermore, these papers suggested that the extent to which rank matters in the two governing bodies differs because of fundamental aspects of the two committees (namely the size and composition), and that network effects are more important than previously discovered social factors. In order to explore the role of status in voting behavior, we will leverage an approach advanced by Leskovec, Huttenlocher and Kleinberg (2010) that suggests that relationships among triads of individuals (in directed, dynamic graphs from social media sources) more closely reflect the theory of status than the theory of structural balance. However, this analysis of status has not been used in the political arena, nor has it looked at the emergence of status over time, as conferred by repeated voting between the same pairs. Accordingly we will compare several time delimited iterations of the House and Senate relational maps in order to observe how status and reciprocity are manifested. The research questions are as follows: