Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Business and Economics



First Advisor

Dr. David R. Agrawal


This dissertation looks at the role that geographically limited taxation and the behavioral response of economic agents to different forms of regional taxation. While geographic taxation is typically thought to be confined to state and local taxes, federal tax policy can also create designated areas that receive preferential tax treatment. In this dissertation, both types of regional taxation are examined and are unified under common themes such as tax mobility and the definition of the tax base.

In chapter 2, I examine the important effect of how the location of high income earners employment is influenced by taxes in the state of employment. Previous literature has focused on the residency decision of high income earners, but that is not the only margin that matters, as in both the United States and the European Union taxes are instead due where income is earned. Using the universe of PGA tour participants from 1970-2018, I estimate the participation response to golfers with respect to a change in their effective tax rate. In the baseline specification, I find a tax participation elasticity of 0.32 and for those in the top 25 percent of prior year earnings, consistent with the superstar effect, find a larger elasticity of .82. I also find that federal taxes play a minimal role in the tax response of golfers and the primary effect is driven by changes in state taxes, suggesting that golfers reallocate their income away from low tax states towards higher tax states instead of a true labor supply response. In Chapter 3, I look at an important distinction in tax policy, whether tax base changes are different than an equivalent change in the tax rate. This is in the context of a gradually narrowing consumption tax base that has largely left both physical and digital services out of the tax base. Using variation in food tax inclusion and the food tax rate combined with data from the Nielsen Corporation on the retail sale of food, I find that a one percent increase in the gross price of food decreases sales in border areas by less than one percent using a border pair identification strategy in both a static and dynamic setting. In addition I estimate the effect of removing food from the tax base using removals from West Virginia and South Carolina and find no additional effects beyond an equivalent rate reduction, suggesting among this very specific population, there is no meaningful difference between the two. This is largely suggestive that broadening the tax base would provide large gains in revenue with little loss in efficiency. Chapter 4 explores whether place-based policies, which are geographically defined programs that provide special benefits to particularly poor neighborhoods lead to persistent effects once the programs expire. This question is important as the United States as well as other countries continue to invest in place-based policies without knowing the long run impacts. Using restricted access American Community Survey and the expiration of Renewal Communities relative to Empowerment Zones, I find little evidence for a permanent improvement within Renewal Communities, with RC employment, income, and rent decreasing relative to the continuing Empowerment Zones. However, I do find that for those individuals who live and work in the zone, there is no subsequent decline in employment or income, suggesting that there is heterogeneity in persistence. Given that the newest iteration of place-based policies in the United States, Opportunity Zones, largely shares the same benefit structures as Renewal Communities it is unlikely that we will see lasting benefits from it's implementation.

Digital Object Identifier (DOI)