Questions in Public Administration

Abstract

My dissertation looks at three different public administration challenges and explores the effect that bias, tradeoff of efficiency and accuracy, and unintended consequences have on decision making. The first public administration challenge is bias in administering public services. Using disability appeals at the Social Security Administration, I find that traits such as political ideology, gender, and age of administrative law judges robustly predict the outcomes of disability appeals at the Social Security Administration. To my knowledge, this is the first research to estimate partisan effects on disability decisions and the first to incorporate the third common political self-identification of Independent in an analysis of quasi-judicial decision making. Democratic judges are most lenient, Independents are least lenient; male judges are more lenient than female judges. My model predicts that a female Independent judge awards disability benefits at 10.1 percentage points less often than a male Democratic judge. The second chapter analyzes the tradeoff between accuracy and efficiency in public administration. How do time constraints shape approval rates of disability appeals at Social Security Administration? The random assignment of Social Security disability cases to U.S. administrative law judges—and the requirement that they adjudicate 500 cases per year to remain in good standing—provide an opportunity to offer some answers to this important question. The quota on the number of cases an administrative law judge must adjudicate in a year robustly predicts outcomes for a small subset of administrative law judges, 6% of my sample, who appear to be time constrained. These judges are 2.1% more likely to approve appeals cases throughout the entire year and 3.7% more likely to approve appeals cases in the final month of their fiscal year. The third chapter tackles the final challenge: unintended consequences when awarding government contracts. Federal procurement has a goal of awarding 23% of all federal contract dollars to small businesses. I look at very small firms participating in the federal procurement economy and find that they differ from small firms in the regular economy in what types of industries they participate in. Additionally, small firms’ success by contract dollars obligated is strongly positively correlated with how many industries they participate in: Winning a contract in an additional 4-digit NAICS code predicts a 48% increase in contract dollars. I present evidence that participating in multiple NAICS codes is a proxy for understanding the procurement market’s regulation, processes, and networks rather than authentic diversification. This suggests that success in the procurement market for small firms has less to do with expertise in an industry and more to do with specialization in federal procurement.