Published online by Cambridge University Press: 01 August 2014
Most models of agenda control examine dyadic relations—for example, those between a committee and the floor of a legislature. Such relations, however, are always embedded in a larger context, namely, a political environment composed of voters and interest groups. In this paper we model agenda setters (a legislative committee) as decision makers with limited cognitive abilities who adjust over time to their larger political environment. The legislators' policy positions are endogenous, reflecting the relative strengths of voters wielding the district-specific resource of votes and of interest groups wielding the transferable resource of money. The resulting outcomes indicate that neoclassical models of voting and pluralist models of group influence have each told part of the story. When only votes matter, our boundedly rational agents grope toward equilibria close to those of neoclassical models; however, when mobile resources matter as well, the outcomes depart systematically from those of previous models. In particular, interest groups can make themselves worse off by capturing the committee. The results suggest that agenda control is less powerful than conventionally believed and point toward conditions shaping its effectiveness—conditions highlighting the distinctive contributions of pluralist and neoclassical thinking to a broader theory of political institutions.