Early twenty-first-century American politics is characterized above all by partisan polarization, so understanding polarization's impact on Congress's capacity to make public policy is imperative. To do so, I argue, we need to recognize that the impact has both a direct and an indirect component. If the political parties become more internally ideologically homogeneous and more distant from each other, the character of the feasible winning coalitions is affected. Polarization as a concept implies a spatial understanding of the distribution of policy preferences and, within that framework, a winning coalition can be formed only if a policy change that is preferred by all its members to the reversion point can be found.

Thus the best-known models attempting to explain policy gridlock and policy change assume a spatial distribution of stable, exogenously determined policy preferences (Krehbiel 1998; Brady and Volden 1998). A policy is said to fall within the “gridlock interval” if there is no policy preferred to it by a coalition able to effect change; that is, there is no policy that can – based on preferences alone – muster a majority in both houses, overcome a Senate filibuster, and not fall to a presidential veto (Brady and Volden 1998, 14–20). Keith Krehbiel, David Brady, and Craig Volden show that such a single-dimensional model predicts that any policy within the “gridlock interval” will not be changed while any outside the interval will quickly be changed.

I contend that, in addition to this direct effect, polarization has had an indirect effect through the changes in internal chamber organization and distribution of power it has spawned. In particular, polarization has made possible stronger party leadership, especially in the House of Representatives (Bach and Smith 1988; Rohde 1991; Sinclair 1983; 1995; 2006). Stronger party leadership and greater internal rewards for acting as team members amplify the behavioral effects of externally based polarization and thereby amplify the impact of partisan polarization on policy. I argue, in essence, that in the preferences versus party debate, Cox and McCubbins (1993) (and Rohde 1991; Sinclair 1995; Smith 2007; and a large number of other scholars) have it right; policy outcomes cannot be explained solely by stable, exogenously determined member preferences as Krehbiel contends. As I demonstrate in the body of the chapter, the indirect effect of polarization is a party effect and is crucial to understanding contemporary politics and policy making.