Search

This paper introduces a new approach to modeling the relationships between PACs and House members. I combine information about the entire dynamics history of contributions between PAC-candidate pairs, with cross-sectional information about the portfolio of candidates/incumbents supported by each PAC. Prior approaches focus either on single election cycles, or a small set of elections to analyze relationship between candidates and PACs. No prior work has drawn on information about the portfolio of candidates supported by a PAC to consider the dynamics of replacing a previously supported candidate with someone new..x000d..x000d.Key results include showing that most money to incumbents comes from PACs who supported the member prior to their initial election, and offering new scaling results for PACs. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

I provide methods for analyzing learning and repeated observations ofvoter behavior within and across election campaigns. My method isuniquely useful for new panel studies where the time betweeninterviews is randomly distributed. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

To test and distinguish between formal models of contributor behavior, I develop quantitative tests for the qualitative features of these models. Contributions from PAC in open seat races (1984--2004) are considered. [ABSTRACT FROM AUTHOR]

I provide new results for the study of repeated choice behavior over a finite set of discrete alternatives using a probabilistic theory which is consistent with random utility maximization. In the main theorem, transition probabilities between choices are derived without imposing restrictions on how the systematic component of utilities may change over time. I show how existing methods of estimating transition probabilities relate to these new results. [ABSTRACT FROM AUTHOR]