Distributive Politics and the Law of 1/n

Distributive politics models often predict that legislators will demand inefficiently large projects, with inefficiency increasing in the number of districts, and that this will translate into larger projects and higher spending. The relationship between efficiency and legislature size is often referred to as the ‘‘law of 1/n’’(Weingast, Shepsle, and Johnsen 1981). We demonstrate that the ‘‘law of 1/n’’ result with respect to project sizes and total spending is dependent on several factors, including the type of good being provided, the costs of raising revenue, and whether the local government has to share in the project’s cost with the central government. In general, the ‘‘law of 1/n’’ need not hold for total government spending, and in fact a ‘‘reverse law of 1/n’’ often holds. In light of our theoretical findings, we reassess the empirical literature on this topic. The results have implications for a wide variety of applications in American and comparative politics.