This dissertation is an in-depth exploration of lame duck sessions of Congress. The old conventional wisdom, that lame duck sessions of Congress were insignificant periods where Congress conducted some housecleaning by passing minor bills left over from the regular session, ignores a key factor: elections. Elections do not just affect the composition of the next Congress; they also affect the legislative output of the current one. Specifically, when elections result in changes in partisan control, particularly from unified to divided government and vice versa, leaders and rank-and-file members of the political party on the way out have an incentive to pass more significant legislation before they relinquish the reins of power.

My research provides the theoretical basis for this expectation, weighing the different electoral permutations and discussing issues of representation, electoral mandates, and ideological polarization. Building on previous work, I create a statistical model that incorporates electoral results with measures of legislative significance and party polarization. Although this model is based on data from 1877 to 1995, it predicts with some accuracy the legislative outputs of subsequent lame duck sessions of Congress.

To provide a broader context, the dissertation includes a historical overview going back to the founding of the Republic, a review of relevant literature, and in-depth case studies of the three most recent lame duck sessions (2008, 2010, and 2012). The case studies go hand-in-hand with the statistical model, validating the conclusion that elections help determine the number and significance of laws enacted during subsequent lame duck sessions. Scrutiny of the output of lame duck sessions is a significant departure from the existing literature and is central to my contribution. Ultimately, this dissertation provides a theoretical and statistical basis for the hypothesis that changes in partisan control of one or more chambers of Congress - or the White House - affect the legislative outputs of lame duck sessions.