Location-based tax policies are redistributive in nature, evident by their placement in distressed areas. However, the previous literature has focused on mean effects, which can mask important effects that the program has on the distribution of households. To investigate the importance of studying distributional changes in the context of location-based tax policies we study the federal Empowerment Zone (EZ) program, which offers a set of tax incentives to firms if they locate their business in specified distressed areas and hire workers who reside in the area. Specifically, the EZ program offers wage and capital tax incentives as well as service block grants to the community in order to induce businesses to locate in specified distressed areas.

We utilize information on the counts of households and individuals within specified ranges of household income. These counts allow us to construct estimates of the density of individuals and households across the income distribution, as well as other distributions, in each year of our data. Thus, we can estimate how the densities of individuals and households changes at various points in these distributions, thereby providing a more complete picture of the effects of the program compared to average estimates.  For identification, we utilize the selection process of the program that chose distressed areas from a set of qualified applicants. Specifically, we compare changes in the distribution in areas that received EZ designation with areas that qualified and applied for the EZ program but were rejected. Our procedure produces a control group with nearly identical empirical distributions in both 1980 and 1990, and therefore changes in the pre-designation distributions, compared to the EZ areas.

We do not find evidence that the impoverished residents benefited from the program. Our findings are consistent with the areas becoming more attractive to high-income households. The improvements in the areas were concentrated in those portions of each zone that were relatively better off prior to EZ designation. The results confirm the prior literature findings that the areas, on average, became more attractive but also suggest that the benefits of the program likely did not accrue to the lower-income residents of the EZ areas. If the goal of the program is simply to make these economically distressed areas better or more attractive, then the program appears to have been successful. If instead, the program was supposed to aid the impoverished existing residents of the EZ areas then our results suggest that the program may have had little impact.