This dissertation consists of three empirical studies in Economics. The first two chapters mainly focus on the effects of recent technological developments and its implications, by exploiting variation in exposure to industrial robot adoption across commuting zones. In chapter one, I empirically investigate whether low-income families are still supported by Earned Income Tax Credit (EITC) under automation. The analysis is inspired by the fact that welfare programs for low-income families in the U.S. have increasingly emphasized in-work aid over out-of-work aid since the 1990s, whereas recent technological progress such as robots and Artificial Intelligence (AI) is replacing human labor, not just routine-intensive jobs, which consequently implies that the U.S. safety net provides only modest benefits to non-workers. Based on the instrumental variable (IV) estimation, I find that there is no statistically significant difference in EITC usage across commuting zones and that EITC helps single parents out of poverty. Overall, there is no evidence that current EITC is ineffective in local labor markets experiencing growth in automation. In chapter two, I investigate the distributional effects of automation on income in the U.S. in the period 1990-2015. Applying the quantile regression framework for a group-level treatment in Chetverikov et. al. (2016), I examine the effect of each region's industrial robot adoption on the within-region income distribution. The results show that robots are much worse for low-income people as they experience a decrease in both hourly and annual earnings, in which the latter implies reduced hours worked due to robots. Also, automation makes the adverse effect of females on income stronger at the bottom of the distribution, and there is some evidence that it increases the positive effect of educational attainment on income above the middle of the distribution. Overall, the results suggest that industrial robots are more likely to harm less-skilled and low-wage workers. The last chapter empirically examines what determines the distribution of federal economic development funding in the U.S. at the county level. Since there are 425 counties in persistent poverty, I also explore whether there is any difference in the funding distribution between persistently poor counties and non-poor counties. The result presents that neither economic need nor political concerns affect the funding distribution on persistently poverty counties. Also, I find that there are no structural differences in the funding distribution.

Degree Date

Summer 8-4-2020

Document Type


Degree Name





Daniel Millimet

Subject Area


Number of Pages




Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Available for download on Friday, July 28, 2023