Masters Theses

Date of Award

12-2003

Degree Type

Thesis

Degree Name

Master of Arts

Major

Economics

Major Professor

Matthew Murray

Committee Members

Robert Bohm, Jon Shefner

Abstract

We focus this paper on a political economic question, What determines the U.S. states’ sales tax treatment of equity-related goods? We consider a good to be “equity related” when its consumption is essential to survival. States sometimes choose to tax these goods uniquely, often exempting them from the sales tax base. The equity-related goods for this paper are food, clothing, water, electricity, household fuel, and natural gas. All these goods are for home consumption. This question is relevant to the larger issue of how tax structure evolves or is determined. We test the theoretical explanation of tax structure determination that Hettich and Winer developed in a series of papers. Hettich and Winer hypothesize that politicians’ efforts to win election shape the tax structure. We test the Hettich-Winer hypothesis with two regressions. First, we adapt the Hettich-Winer model of income-tax share determination to the sales tax and test that model with random-effects regression on panel data for 1990-1999. Next we generalize the sales-tax share model to a theoretical explanation of what determines the probability a state taxes an equity-related good. We test this second model with a Poisson regression on the count of equity-related elements taxed. The regression on sales tax shares implies that the Hettich and Winer hypothesis is a tenable explanation of tax structure determination. However, the Poisson regression does not support the Hettich-Winer hypothesis. In other words, the Hettich-Winer model works well at the macro level of sales tax share, but poorly at the micro level of element-by-element analysis. We offer two explanations for this apparent contradiction. First, the micro model of equity-related goods taxation might be poorly specified. The factors influencing political cost included in the model are poor proxies and not inclusive of all relevant forces. Second, the Hettich-Winer hypothesis might be a poor description of tax structure determination. Perhaps the Hettich-Winer premise imparts excessive rationality to political agents, or perhaps tax structure equilibriums are path dependent, which would require Hettich and Winer to include the influence of history in their model. The good fit of the macro models of sales tax and income tax shares would not necessarily contradict the assertion that the Hettich-Winer hypothesis is flawed. For example, the Keynesian consumption function is a statistically good fit at the macro model but is generally recognized as a poor explanation of micro-level consumption decisions.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Economics Commons

Share

COinS