The thesis explores the relationship between an incumbent governor's popularity and state legislative election outcomes in the contemporary era. The study employs data collected from 49 states over a 16-year period (1988 - 2003). Existing archival election and economic data were compiled, compared and analyzed using regression analysis to provide additional insights into the state legislative elections literature. The paper offers a better understanding of the relative effects that certain variables may have on state legislative election outcomes. The study provides evidence that voters are apt to "split their ticket" between state and national elections: that state elections are not mirrors of national elections. This could be a consequence of the trend towards devolution and state government's increasing jurisdiction over a broadening array of public policy. The findings suggest that the coattails of presidential and senatorial candidates are not extending to state legislative candidates in any consistent, significant way. However, the positive relationship between a governor's popularity and legislative seats gained by his/her party in the legislature is the one pervasive theme that emerged consistently throughout this project. The result was evident even when variables representing national level candidates, elections and/or officeholders are included. The predicted/expected value for the dependent variable increases during ?on? gubernatorial election years, when an incumbent governor is seeking reelection. The consistent statistical significance associated with both variables representing the governor's popularity indicates that partisan "cues" may be pertinent factors that help determine voting behavior in state legislative elections. / Master of Arts

I reassess the influence of presidential approval on state legislative election outcomes, incorporating the period from the 1940s to the 1970s in my analysis. Previous research finds that presidential approval has a significant effect, but such findings may be biased-they focus on elections after the 1970s, when the president was more visible to the public. Using an original state partisan balance dataset, I measure the effects of presidential approval and find that it has as much influence on state legislative elections from the 1940s to the early 1970s. These findings may engender concerns of state legislative accountability-if state legislators’ electoral prospects become increasingly reliant upon assessments of the president than themselves, they are less likely to feel beholden to voters and uphold their interests.

This dissertation consists of three projects related to reputational effects in legislative elections. Building on the candidate emergence, repeat candidates and congressional donor literatures, these articles use novel datasets to further our understanding of repeat candidacy and the impact of interest group endorsements on candidate contributions. The first project examines the conditions under which losing state legislative candidates will appear on the successive general election ballot. Broadly speaking, I find a good deal of support for the notion that candidates respond rationally to changes in their political environment when determining whether to run again. The second project aims to measure the impact of repeat candidacy on state legislative election outcomes. Ultimately I find a reward/penalty structure through which losing candidates for lower chamber seats that perform well in their first election have a slight advantage over first-time candidates in their repeat elections. The final chapter of this dissertation examines the relationship between interest group endorsements and individual contributions for 2010 U.S. Senate candidates. The results of this chapter suggest that some interest group endorsements lead to increased campaign contributions, as compared to unendorsed candidates, but that others do not. This research points to a number of opportunities for future research as the relationship between endorsements and campaign resources is vastly understudied. / Political Science