CAREER: A Legislative Theory of Public Spending, Debt and Taxation

Cornell University, Ithaca NY

Investigators

Abstract

It has long been argued that legislatures in which representatives are elected by geographically defined districts will make inefficient decisions. According to conventional wisdom, the attempts of legislators to benefit their constituents at the expense of the general community through pork barrel spending lead to both excessive spending and a misallocation of government revenues between distributive policies and important national public goods. Despite this widely held view, economics and formal political theory tells us little about the dynamics of legislative policy choices. Will legislative policy-making result in a long run size of government that is too large? What will be the time path of public debt and investment in national public goods? What features of the environment determine the magnitude of the distortions arising from legislative policy-making? This research project addresses these questions in a novel infinite horizon model of legislative policy-making. The basic model is that of an economy in which policy choices are made by a legislature comprised of representatives elected by single-member, geographically defined districts. In each period, the legislature chooses the level of a distortionary income tax, the level of public debt, and decides how to allocate these resources between investment in a public good that benefits all citizens (infrastructure, national defense, etc.) and district-specific unproductive transfers (pork barrel spending). Legislative policy-making in each period is modeled using the legislative bargaining approach of Baron and Ferejohn (1989). The dynamic linkage across periods is created by the level of public capital and public debt, which are state variables in the model. The analysis of this dynamic model is conducted through four interrelated projects. In the first two projects we characterize the growth path of this economy and the nature of the inefficiencies that may arise in legislative policy making under alternative assumptions on the ability of legislatures to finance expenditures. The remaining two projects integrate theoretical and experimental research by testing in the laboratory a model of dynamic bargaining and alternative voting mechanisms. Intellectual merit. This project provides a rigorous formal underpinning for the conventional wisdom mentioned above by showing conditions under which the steady state size of government is too large and the level of public goods is too low. However, it shows that this conventional view needs important qualifications. When the economy's taxable capacity is small relative to its public good needs, legislative decisions will actually be efficient in the long run, despite the fact that legislators can benefit their districts via distributive policies. Even when efficiency is not achieved, the nature of the inefficiency could be quite different from that which the conventional wisdom assumes: the overall size of government could be below optimal and revenues could be allocated to their most productive uses. The project obtains sharp comparative statics and novel positive predictions on the size and dynamics of legislative coalitions, and the relationship between economic variables as the elasticity of labor supply and the productivity of the private and public sectors, and the political equilibrium. It provides a positive theory of public debt useful to predict how political institutions react to shocks. With the experimental work this project tests the robustness and empirical validity of the theoretical results. Broader impact. Understanding the link between economic variables and political equilibria is of crucial importance to design better institutions, and to enhance policy makers' ability to predict how existing institutions respond to shocks and other environmental changes. All the proposed projects have direct policy implications, especially on constitutional design. The laboratory experiments provide an effective way to test the practicality and robustness of these policies. The educational agenda of the proposal builds on these research components. The project develops a semester length graduate class that explores dynamic models in political economy; and generates a set of tools useful to many other researchers in the field (datasets, software programs). The project will also organize a two-day conference titled "Dynamic Models in Political Economy" which will draw together researchers and students working in this area.