This dissertation contains three chapters that examine various behavioral responses to statutory tax policies. In the first chapter, I develop a framework to estimate the impact of the marriage tax on the likelihood of marriage that incorporates into one analysis all four distinct household alternatives: single, cohabit, married, and separated. This is in contrast to previous works that consider only one of three separate choices. Using data from the March CPS from 1989-1999, I estimate a bivariate probit model and find that the marriage tax has a small, but significant, effect on the likelihood of marriage. Furthermore, my results indicate that studies that do not include all four possible alternatives can overstate by as much as 200% the effect of the marriage tax on the likelihood of marriage. The second chapter considers the net distributional impact of the federal tax deduction for charitable donations. If itemizers, who tend to have higher income than non-itemizers, give to charities that provide goods that they directly use or benefit from (egoism), the government is essentially subsidizing the activities of the high-income donors. Conversely, if itemizers donate to organizations that benefit the needy (altruism), the tax deduction aids in a form of income redistribution. I estimate this tax responsiveness of giving using the Center on Philanthropy Panel Study (COPPS) module of the PSID in 2001 and 2003 for 11 types of charities. Donations by high income individuals to charities that benefit the poor are more price elastic than donations to charities that benefit themselves. I find evidence that the current tax deduction induces itemizers to donate more to charities that benefit the poor than they would have without the deduction. The third chapter estimates the economic incidence of the excise tax on tobacco. Using historical price and tax data from 1954-2005, I estimate what portion of the tax is shifted to consumers. I experiment with controls for border crossing and indoor smoking bans. I find that a 10-cent tax increase causes price to increase by 8 cents immediately and by 13 cents in the long run.

The purpose of this dissertation is to examine an unexplored aspect of tax expenditures: the tax-price implications of tax exemptions, deductions and credits. Although this implication of tax expenditures has not been adequately examined, two separate lines of analysis have been suggested by the existing literature. Some authors have emphasized the welfare costs of tax expenditures. To the extent that tax expenditures narrow the tax base the introduction or extension of a tax expenditure undoubtedly makes the cost of raising revenue more than it would be otherwise. This kind of cost, denoted as a welfare cost, can be incorporated into a model of individual tax-price determination. On the other hand, other authors have emphasized another tax-price implication of tax expenditures: that the introduction or extension of a tax expenditure changes the cost-shares faced by each taxpayer, exclusive of any welfare cost. Since an individual's cost-share is nothing more than his personal tax base divided by the aggregate tax base, this result emerges because a tax expenditure usually changes the individual's tax base in a manner disproportionate to the change in the aggregate tax base. This dissertation will explore and combine each line of analysis, both theoretically and empirically. In the first portion of the dissertation a model of tax-price determination is developed that explicitly incorporates the welfare cost of taxation. Various tax expenditures are then introduced into the model and their effects on individual tax-price schedules discerned. In this way the influence a tax expenditure has on an individual's choice over public sector size can be surmised. The next portion develops within the confines of a simple median voter model some potential allocative implications of various tax expenditures. This portion traces out the expected change in the median voter's desired quantity of the collective good, given various tax expenditures, via an analysis of the cost-share impact of the various tax expenditures. Although in this section welfare costs are not explicitly considered or all possible political cases outlined, the analysis does look at a set of cases that are of general interest. The final portion of the dissertation considers the influence tax expenditures taken in toto have on both the cost-sharing arrangement among individual taxpayers and the welfare cost to individual taxpayers. The results are used to gauge both the distributive and allocative implications of tax expenditures. / Ph. D.