Advances in Distributional Analysis

John Barry

Advisor: Richard E Wagner, PhD, Department of Economics

Enterprise Hall, 318
April 26, 2004, 08:00 PM to 07:00 PM


This dissertation proposes several refinements of the conventional methodology followed to measure the distribution of federal individual income taxes to better account for important demographic characteristics when measuring the impact of changes in tax law. The conventional methodology followed by the primary government agencies responsible for measuring the distribution of federal taxes all suffer from generalization; generally all tax returns are treated exactly the same within these agencies' analyses no matter whether the tax return is filed by a single individual living in a small town or a large family living in a metropolitan area. As long as the total income reported on these two returns is the same, the tax returns are considered the same. This generalization prevents lawmakers from properly measuring the horizontal and vertical equity of changes in tax law and thereby leading them to make decisions that they may not otherwise make with better information. Lawmakers can be presented with more accurate information by accounting for important demographic differences such as filing status, geographic location, the number of individuals represented on each tax return, and primary source of income when determining the distribution of taxes. This dissertation demonstrates that the current methodology is indeed inadequate in presenting lawmakers with an accurate picture of the distribution of taxes and that the proposed refinements do improve the clarity of this picture. Moreover, the improvements recommended in this dissertation may have a meaningful effect on federal tax policy and therefore a meaningful economic impact.