Replication data for: The Stability of Partisan Identification in the U.S. House of Representatives, 1789-1984

It is a basic premise of this study that politicians who switch parties are likely to do so when they are ideologically out of line with their current political party and more in agreement with the other party. Nor do we quarrel with the notion that sometimes the switch may also be related to electoral incentives. The problem lies not in these conclusions but in their utility. After all, there are many other members of congress who have never switched parties but are in as much or more ideological discomfort. Why haven't they all switched? As an ex post facto argument, these \"ideological dissonance\" and \"personal ambition\" hypotheses do correctly describe the switchers, but they do not adequately explain why these people were the ones to switch or why they did it at the time that they did. Many of these members have been in a state of ideological dissonance for years. In each case, why did the decision to change parties occur precisely when it did? It is our intention to advance the study of party switching by providing an explanation for why switching occurs more frequently at some times and not others. Unlike the cross-sectional data used in past work, we adopt a time series approach to analyze this questions. We discuss (1) political loyalty as an important but too rarely used explanation for certain political phenomena, (2) descriptions of House party switchers, (3) the specific models and variables we will use to explain switching, and (4) the estimates from several statistical analyses that help explain when and why switching occurs. The last three sections are meant as a specific exploratory study of one aspect of political loyalty.