Collaborative Research: Transformations in Political Party Organizations and the Rise of Candidate-Centered Elections in the U.S., 1878-2008

Candidate-centered politics and personal voting in the United States are much stronger today than they were in the past, and party organizations are much weaker. This project asks two main questions: i) Why did this happen?; ii) What are the implications of this development for the future of American democracy?

Conventional wisdom in the American Politics literature is that the era of candidate-centered elections took hold sometime in the 1950s or 1960s. At this point the traditional party organizations finally succumbed to a variety of pressures, including changes in campaign advertising technologies such as television, the replacement of patronage with civil service employment, and an increase in the resources available to individual elected officials for constituency service. Voters saw parties as increasingly irrelevant, and their partisan loyalties grew weaker. However, several prominent scholars argue that this transition began earlier and was more gradual. They emphasize the importance of primary elections in weakening party organizations and cultivating personalistic politics.

This research will pursue several empirical projects that will shed some light on the relative importance of various factors that might have weakened party organizations and fostered candidate-centered elections. How important was the direct primary? the expansion of civil service and decline of patronage? the introduction of new communications technology? By how much did the campaign efforts of candidates increase relative to those of parties, and when did these changes occur? Did they occur earlier in states with competitive primaries?

The research will also investigate the effects of candidate-centered voting in primary elections. How often do incumbents involved in scandals lose in the primary? How often does this happen in safe districts or states, where these incumbents would be likely to win in the general election despite the scandal? Do primaries foster or hinder inter-party competition? Do primaries prevent the election of geographically or ideologically balanced tickets? Do they cause polarization?

Finally, the research will investigate the relationship between the weakening of party organizations and the rise of intra-party factions over time and across states. Do voting data and newspaper accounts provide evidence of regular divisions within party electorates? Are the factions ideological, sectional, sssue-based, urban vs. rural, or personal? Did personalistic factions grow in importance after the adoption of primaries? The work can also use the study of intra-party factions to investigate voter behavior and information. How strongly do ideological or issue-based factional divisions affect voter behavior in primaries? Do these divisions also spill over into general elections?

The intellectual merit of this project reflects not only its capacity to shed light on US politics. The research addresses a general puzzle in the literature on elections: why are some elections candidate-centered and others party-centered? In addition, even today there is considerable debate about the benefits and costs of candidate-centered politics compared to party-oriented politics (e.g., arguments about the decline of collective responsibility countered by arguments about better representation), and some of this stems from disagreement about the main forces driving candidate-centered elections. This work will help to assess these costs and benefits.

The project will yield broad educational and social benefits, including (i) three new digital databases that will be made publicly available; (ii) publications in a variety of different media; (iii) datasets of use to journalists and political commentators; (iv) a research collaboration between two major educational institutions; (iv) mentoring and employing research assistants who are women and under-represented minorities; and (v) professionalization of undergraduates and graduate students who co-author with the PIs.