Optimal dynamic fiscal policy

This thesis contributes to theoretical and quantitative understanding of fiscal policy in dynamic environments. We analyze how to structure indirect and capital taxes, how to optimally design a disability insurance system, and how capital gains taxation affects decisions of entrepreneurs.; The first essay considers an environment in which agents' skills are private information and follow arbitrary stochastic processes. We prove that it is typically Pareto optimal for an individual's marginal benefit of investing in capital to exceed his marginal cost of doing so. This wedge is consistent with a positive tax on capital income. We also prove that it is Pareto optimal for the marginal rate of substitution between any two consumption goods to equal the marginal rate of transformation. This lack of a wedge is consistent with uniform taxation of consumption goods within a period.; The second essay describes how to optimally design a disability insurance system. The key friction in the model is imperfectly observable disability. We solve a dynamic mechanism design problem and provide a theoretical and numerical characterization of the social optimum. We then propose a simple tax system that implements an optimal allocation as a competitive equilibrium and includes only taxes and transfers that are similar to those already present in the U.S. tax code: a savings tax and an asset-tested transfer program. Using a numerical simulation, we compare our optimal disability system to the current disability system.; The third essay develops a model of entrepreneurial choice in which some individuals have a comparative advantage in starting new business enterprises. Efficiency dictates that entrepreneurs should specialize in start-ups and sell successful start-ups to professional managers. We consider the role of capital gains taxation. With taxation of capital gains, some of the entrepreneurs get inefficiently locked into running their own enterprises. We quantify the role of this effect and argue that it is large.