This research uses theory and experiments to explore questions related to a legislature in which representatives with heterogeneous preferences divide a fixed budget between a public good and pork projects (local public goods). The dissertation is organized into two chapters.

The first chapter examines the incentives of voters to vote strategically when selecting representatives for the legislature. It asks under what conditions strategic delegation, that is voting for a representative with different preferences from one's own, is optimal. Using the underlying bargaining model in Volden & Wiseman (2007), we show that voters have an incentive to strategically delegate to affect how the budget is divided at the legislative level. Contrary to the existing literature, the incentives for strategic delegation exist to direct more money to the public good and less to pork projects. In equilibrium there are at least as many representatives as districts that favor the public good. The paper also shows that the positive externalities associated with increased public goods contributions increase expected social welfare. An extension to the case of more than one representative per district is considered.

The second chapter reports the results of a laboratory experiment designed to investigate hypotheses about the way spending on pork facilitates public good provision in a legislative bargaining environment when preferences are heterogeneous (Volden & Wiseman, 2007). Contrary to the model's predictions, players favoring the public good often need no pork to pass heavily funded public goods proposals even when these players are in the minority. However, when they include pork in a proposal, proposers use amounts that closely coincide with the theory. Proposers preferring pork also deviate from the model when in the majority and offer predominantly minimum-winning coalitions that fund only pork and no public goods. Fairness concerns, reciprocal voting, and reference points are shown to partially explain these results. Reference points often induce players to hold out for higher payoffs than can be justified by their continuation value for the game and change depending on the preferences of the proposer. One implication is that in a handful of instances this can actually initiate more spending on pork and less on the public good than predicted by the theory as the votes of players preferring pork are less costly to attain.