The evidence and anecdotes in the previous chapters have, we think, established the plausibility of our model's key assumption – that the majority party pays lower consideration costs than the minority party (and other rogue senators) in trying to advance its agenda items – and grounded our theory in the procedure and agenda-setting stages of the Senate. They do not, however, test the model's predictive power. In this chapter, we turn to that task.

We focus on our model's predictions about the effects of changes in party control and changes in majority consideration costs on the direction of policy movement and the size of winning coalitions. In the first empirical section, we test our model's predictions against predictions from models that take no account of parties. We focus on the aggregate (i.e., Congress-level) direction of policy movement on final passage votes over a time series spanning 1949–2008, finding support for the prediction that Senate decisions predominantly move policy toward the majority party's side of the political spectrum.

In the second empirical section, we test predictions made by our model that are not made by other party models. We use vote-level data to examine variation in the direction of policy movement and coalition sizes on final passage, across sets of bills that have different majority consideration costs. We find that policy is more likely to move toward the majority party and coalition sizes are smaller on votes involving lower majority consideration costs than on votes involving higher majority consideration costs, as predicted by our model.