FIRST DRAFT. PLEASE DO NOT QUOTE WITHOUT PERMISSION.

Over the last 30 years, as legislative party polarization has increased, bipartisan compromise has become normatively more appealing and apparently less attainable. While an extensive literature is devoted to the antecedents and consequences of party polarization and gridlock, almost no attention has been devoted to the individual legislator level foundations of these aggregate level phenomena. Here I examine the factors that determine the extent to which individual legislators spend time on forming cross-party coalitions—bipartisanship. It is these individual actions shaped by institutions, elections, and citizen and elite preferences that determine collective outcomes, including party polarization and gridlock. The analysis will ultimately be based on two national surveys of state legislators conducted 7 years apart, but in this initial draft I examine only the 2002 survey. Legislators were asked how much time each spent on building coalitions to pass legislation. They were asked separately about building coalitions within their own party and across parties. The first section describes the data and presents the basic descriptive information on the time legislators spend on each type of coalition building. I also sho