Essays on partisan politics in electoral and legislative arenas

The dissertation consists of three essays that examine the role of political parties in American politics. The first essay develops a formal model of electoral competition where two parties present platforms (consisting of a policy position and a level of endorsement) for their candidates to voters. Voters' utilities are defined over a candidate's policy location and the level of endorsement that he receives. Results identify cases in which parties can successfully pull candidates away from the favored policies of the median members of their constituencies, and towards the parties' interests. The results provide one reason for non-centrist electoral outcomes and the results also have indirect implications for the debate over party power in legislatures. The second essay subjects the analytical results of the first essay to large-sample empirical tests. Theoretical parameters are operationalized using data from Illinois General Assembly elections across an eight-year period, and the empirical findings support the theory. Candidate positions and party campaign contributions are shown to be responsive to district partisan predisposition, even when controlling for conventional determinants of candidate positions and party contributions. Uncontested elections occur more often in politically lopsided districts than in districts where there are more even levels of partisan competition even when controlling for incumbency strength. Finally, an inverse relationship between candidate extremity and campaign contributions, consistent with the analytical results of the theory, is observed. The third essay is coauthored with Keith Krehbiel and challenges conventional notions of party leadership in congressional politics by considering the speakership of Joseph G. Cannon during the 58--61st Congresses. Specifically, this paper assesses the conventional wisdom on Cannonism by systematically analyzing the committee transfer patterns. Employing the Groseclose-Stewart method for estimating values of committee seats, we study variation in member-specific committee portfolio values. Of particular interest is whether members who voted against Cannon's policy agenda on key roll calls were systematically punished in that they were assigned to less-desirable committees than Cannon loyalists. The findings fail to support the widely accepted Cannon---as tyrant thesis, and (if anything) support the notion of Cannon as a majoritarian.