Vote switching in the U.S. house

Previous work on voting behavior in Congress leads to an expectation of stability in roll-call voting. Yet, there remains a degree of unexplained variability in roll-call voting. Utilizing a unique set of data from the 103rd Congress of vote switching on revotes, we seek to account for the random component of congressional voting. We offer four categories of variables to explain this instability: leeway effects, error correction, reduction in uncertainty, and bandwagon effects. While these variables explain some of the variation in vote switching, our findings show that the individual random component of roll-call voting may be more significant than previous theory has indicated.