Essays on the marginal cost of public funds

This dissertation investigates theoretical issues concerning the marginal cost of public funds (MCF). The first essay addresses alternative measures of MCF within two different frameworks. I prove that the two main candidates for MCF can be equally useful for "the local cost-benefit analysis" if the benefit side of a public project is appropriately measured. I also show that the marginal welfare cost (MWC)--the basis of one of these MCF measures--is a more critical concept for "the global second-best level analysis." The second essay develops a new approach to the "level problem" based on MWC. I prove that, if the MWC of a distortionary tax is everywhere positive, then the second-best level cannot exceed the first-best level. Moreover, I argue with both theory and examples that a uniformly positive MWC schedule is also necessary in a sense for the second-best level to be below the first-best level. The third essay proposes an approach to multiperiod project evaluation that is consistent with welfare maximation. Based on this approach, I revisit three important concepts for public project evaluation: MCF, the social discount rate, and the shadow price of capital. The fourth essay uses a two-period model to incorporate the savings effects of a labor income tax. This model is comparable with the widely used static labor supply model in the MCF literature but accounts for the individuals' savings decisions and the revenue from capital income taxes. Hence, I succeed in isolating the impact of savings effects on the MCF of a labor income tax. My main finding is that, with a tax-distorted capital market like the one that exists, accounting for savings effects (in addition to labor supply effects) significantly increases MCF magnitudes.