Theories of party power in Congress differ on the circumstances under which majority parties have the ability to shift policy outcomes away from the preferences of pivotal voters and toward the majority's preferred position. The theory of Pivotal Politics states that it is unlikely parties have such power. The theory of Conditional Party Government states that parties can influence policy outcomes when they are ideologically unified, while the Cartel theory suggests that parties can influence outcomes all of the time by controlling the agenda.

In this dissertation, I propose and test three hypotheses addressing the extent of party power using an original dataset of the legislative history of federal appropriations bills and case studies of two time periods in Congress. Appropriations bills are an effective way to study trends in Congress because they must be passed every year. In the last three decades, Congress has shifted from its traditional method of passing the 13 bills that fund the federal government individually to packaging them together in massive "omnibus" bills. I show that the decision of party leaders to create omnibus bills is a form of agenda control that allows party leaders to meet a variety of goals ranging from protecting the majority party's reputation to adopting partisan policy. Omnibus bills help party leaders meet their goals because they are multidimensional, "must pass" bills that members are reluctant to oppose. They are particularly useful in the Senate, where they provide an effective counter to the ever present threat of a filibuster.

I make three major arguments. First, I contend that the ability of a majority party to control the agenda with omnibus spending bills is independent of its degree of ideological diversity. In the last 30 years, omnibus bills have been used both when the majority party is ideologically diverse and when it is unified. Second, I contend that the likelihood a majority party will seek to control the agenda with omnibus bills depends on the ideological distance from the majority's median voter to other pivotal voters on the floor. These distances have varied over time with the ideological diversity and margin of control of the majority party. Large ideological gaps between pivotal voters are an indication that the floor is a challenging arena for the majority party and create an incentive to control the agenda. Third, I contend that the policy consequences of omnibus bills vary with the majority party's ideological diversity. Diverse parties are likely to use omnibus bills to "keep the trains running" by passing the budget, while unified parties are likely to use omnibus bills to pursue partisan policy goals.

My findings expand our understanding of the motivations of members of Congress. Theories of Congress rooted in the reelection motive state that individual behavior, and by extension, the behavior of parties, is motivated primarily by the desire to improve prospects for reelection. Evidence from the history of appropriations bills over the last 30 years suggests that ideologically unified parties will use omnibus bills to pursue policy goals even if those goals create some additional risk of not being reelected.