Book Description

The linkages between partisanship, split-ticket voting, and divided government have long been treated as trivial, even deterministic. Independents, it has been consistently shown, are more likely to split their tickets than are partisans, to the point that this behavior has sometimes been considered the truest indicator of independence. Similarly, the relationship between split-ticket voting and divided government has gone unquestioned, with several scholars asserting that universal straight-ticket voting inevitably leads to unified government. This dissertation argues that more nuance is needed, and that both relationships can, under certain circumstances, be inverted. The first two chapters focus on the relationship between partisanship and ticket splitting. Using a simple formal model, I predict that when the same party runs stronger candidates in two races, partisans of the disadvantaged party will be more likely than independents to split their tickets. In the first and second chapters, respectively, the theory is confirmed using ANES data from 1952 to 2016 and an experiment using Amazon's Mechanical Turk. In the third chapter, I show that, rather than promoting divided government, ticket-splitting is sometimes a necessary condition for unified government. In 2012, Barack Obama won the presidency while losing the majority of congressional districts, meaning that government would have been divided even if no voters split their tickets. As a way of examining whether or not this was a unique occurrence, I re-examine the five presidential elections from 1876 to 1892. Previous scholars have claimed that the rarity of divided government during this period was due to the lack of split-ticket voting. However, using a novel dataset of presidential election results in congressional districts, I demonstrate that in two of these five elections, divided government would have resulted had voters split their tickets. The chapter argues for a more sophisticated understanding of the ways in which electoral bias in presidential and House elections, as well as candidate quality, shaped political outcomes in the late nineteenth century.