A vast literature has documented a growing ideological divide between the parties in the contemporary U.S. Congress. This research almost universally measures this polarization based on estimates from roll-call voting behavior. However, a recent, burgeoning literature has cast doubt on the over-time comparability of such roll-call based measures due to changes in the congressional agenda. We leverage data from candidate surveys that allow us to hold the policy agenda constant across the time period under study, 1996-2006 (and eventually 2008). Our results suggest that polarization is not solely an artifact of an evolving agenda. In contrast to past results, we do not find evidence that members are adopting more extreme positions throughout their tenure in the House when we hold the agenda fixed. Instead, the increasing polarization within the fixed policy agenda is entirely due to replacement: the new members elected to the House during this time period are more extreme than those leaving the House during this period. Our findings suggest that the within-member movement to more extreme positions observed in roll-call voting behavior is the result of a changing agenda, increasing party pressure, or some combination of the two.
Professor Snyder's primary research and teaching interests are in American politics, with a focus on political representation. He has written on a variety of topics, including elections, campaign finance, legislative behavior and institutions, interest groups, direct democracy, the media, and corruption. He is a Research Associate at the National Bureau of Economic Research, and a Fellow of the American Academy of Arts and Sciences.