Search

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2018. "Chapter two was jointly written with Gregorio Caetano and Josh Kinsler. Chapter three was jointly written with Richard DiSalvo and Josh Kinsler"--Page xi., This dissertation develops and applies new techniques to measure the causal effects in observational data. Chapter one investigates the effect of online social networks on consumer behavior. A key challenge is that social network formation is endogenous. Thus, it is hard to distinguish the impact of the social network itself from the impact of the factors leading to the formation of the network. I attack this identification problem using a strategy that exploits social network friendship variation within groups of individuals with similar characteristics. Specifically, I use data from the Yelp Open Dataset to estimate the causal effect of social network friendship on restaurant choices. I apply a machine learning technique to classify individuals into groups based on a high-dimensional vector of characteristics. Using the variation in social network friendship within these groups, I find that compared to non-friends, social network friends are 67 percent more likely to visit the same restaurant within a year. Additional exploration of the heterogeneous effects suggests that this social network effect is stronger for newer users and much weaker for chain restaurants or when a friend’s rating is negative (i.e. less than median). The second chapter studies the effect of children’s time allocation on their skill development. The work is jointly written with Gregorio Caetano and Josh Kinsler. While there is broad agreement that a significant amount of skill acquisition and development occurs early in life, the precise activities and investments that drive this process are not well understood. In this paper we examine how children’s time allocation affects their accumulation of skill. Children’s time allocation is endogenous in a model of skill production since it is chosen by parents and children. We apply a recently developed test of exogeneity to search for specifications that yield causal estimates of the impact time inputs have on child skills. We show that the test

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2015. The third chapter is jointly written with Sulagna Mookerjee., Chapter one of this dissertation presents the first comprehensive attempt to empirically investigate the effect of illegal firearm prevalence on crime rates in the United States. Previous studies in the empirical literature have provided evidence for the existence of both positive and negative effects of firearm prevalence on crime but have focused exclusively on proxies for legal gun ownership. Most gun crime, however, is committed through illegally acquired firearms making it crucial to study their impact on crime rates but we lack any measure of for their prevalence. Using incident level crime data from the NIBRS, I create a novel proxy for illegal firearm prevalence: the number of firearms reported stolen by victims of property crimes in each police jurisdiction. I employ standard panel data methods using jurisdiction and time fixed effects, along with controlling for lagged crime aggregates. Results show that a 1% increase in stolen firearms over the last two quarters leads to a 0.047% increase in homicides by firearms, a 0.104% increase in aggravated assault involving firearms, and a 0.052% increase in firearm robberies in the current quarter. Given the lack of an exclusion restriction, I perform a battery of falsification tests but find no evidence of a spurious relationship between illegal firearms and crime rates that could potentially bias the results. Back-of-the-envelope calculations show welfare cost savings at the order of $230,000 per recovered illegal firearm for high crime cities, like Detroit. The second chapter is in the field of health economics and provides new evidence from birth certificate data (2010 - 2012), to study the impact of the Supplemental Program for Women Infants and Children (WIC) on birth outcomes. Due to the lack of an exogenous variation, the literature largely uses a selection-on-observables approach, with varying success given the limitations of existing datasets. However, the immense coverage of the birth certificate data, a

Thesis (Ph. D.)--University of Rochester. Dept. of Economics, 2014., This dissertation contributes to two literatures: Urban Economics and Labor Economics. In the first chapter I estimate the effect of home ownership on individual workers' unemployment and wage growth, as well as other labor market outcomes. Because of higher moving costs, home owners will be less willing than renters to relocate for work and could therefore face longer unemployment spells. To elaborate on this hypothesis, credited to Oswald (1996), I build a simple search model and obtain a set of labor market predictions to test. The current microeconomic literature has reached mixed results regarding home ownership's impact, with most studies concluding that home ownership reduces unemployment. I show that the instruments used are likely to be invalid because of, among other reasons, Tiebout (1956) type sorting into housing markets. I use an instrumental variable free of the endogeneity present in other work: the county level home ownership rate when and where the worker grew up. This IV affects workers' preferences for housing but not, conditional on my covariates, their labor market ability. My results indicate that home ownership is a significant hindrance to mobility, and homeowners suffer longer unemployment spells and slower wage growth because of it. In the second chapter I use a dynamic model of neighborhood choice to estimate household preferences over the demographic characteristics of a neighborhood. I focus on the racial mix, average income and housing price level of a neighborhood, and whether households prefer neighbors that are similar to themselves. Identification of these preferences is complicated by the social aspect of neighborhood amenities. A household's valuation of a particular choice (neighborhood) is a function of the choices other households in the market have made and will make in the future. I show that demographic characteristics of a neighborhood are therefore endogenous to neighborhood quality. Standard estimates of preferences over

Thesis (Ph. D.)--University of Rochester. Dept. of Economics, 2013., Although numerous reforms for improving public education have been proposed in the United States, the effects of implemented programs remain controversial. Focusing on recent reform policies, my dissertation examines the impact of school choice, educational accountability, and teacher performance-pay programs on student achievement. Chapter One analyzes the effects of introducing charter schools, fast-growing school-choice programs, on students at neighboring traditional public schools. Unlike prior work, which estimates the effects at the school level, this study examines the impact at the grade level by exploiting the fact that charter schools expand their grade ranges over time. I define direct impact as the impact on traditional-school students in overlapping grades and indirect impact as the impact on those in non-overlapping grades. Using student-level panel data from North Carolina, this study shows that the entry of charter schools generates a positive and significant direct impact on student achievement. Moreover, I demonstrate that one-quarter of the positive direct impact is driven by student sorting while three-quarters result from competition. Chapter Two presents evidence from a regression-discontinuity analysis of North Carolina's accountability program, in which teachers are awarded an additional cash bonus for improving their students' achievement. Results show that teachers who failed to reach an expected benchmark for their students' achievement, resulting in no bonuses, performed significantly better in the subsequent year than those who reached this benchmark and thus received a bonus. Moreover, the results demonstrate that such impact disappeared once the state government repealed the pay scheme| another indication that teachers actively respond to monetary bonuses. Chapter Three examines the performance-pay program from a different viewpoint. To date, no studies in this literature have examined the potentially negative effects of repealing inc

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2016., This dissertation develops and applies new techniques for the measurement of causal effects in observational data. The first chapter studies the employment effects of the minimum wage using a novel empirical strategy that can identify treatment effects when more than one control group is available, even if all such control groups are imperfect. Expanding on previous research that analyzes labor market outcomes across neighboring regions that experience differential changes to the minimum wage, I compare employment outcomes in border counties where the minimum wage increases to a set of neighboring counties, a set of neighbor-of- neighboring counties, and so on. The key innovation is to model the ratio of the biases present in each of these comparisons. The model I choose uses the relative similarity of control groups to the treated group on observables as a guide to their relative similarity on unobservables. Crucially, models of this type have a testable implication when there are enough control groups. Using data from the United States, I find that recent minimum wage increases have produced modest or zero disemployment effects for teenagers. The second chapter extends the method developed in the first chapter to the domain of regression discontinuity designs. In particular, I study the effects of party incumbency on margin of victory in United States House of Representatives elections. There are two key findings. First, the selection ratio method developed in the first chapter produces estimates consistent with regression discontinuity design estimates for districts where elections are narrowly decided. Since regression discontinuity designs are widely considered to be credible, I interpret this as evidence in favor of the selection ratio method in a naturalistic setting. Sec- ond, I show that the same method can also be used to recover treatment effects for populations away from the discontinuity in a regression discontinuity design. Incumbency effects appear to

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2018., In chapter one, I study how payments from pharmaceutical companies affect physician behavior. I focus on the most ubiquitous type of payments, those associated with the practice of detailing (marketing to physicians). Causal payment effects are confounded, as pharmaceutical companies assign physicians in and out of payments dynamically. I use a new identification strategy to overcome this endogeneity. I exploit variations to payment schedules following pharmaceutical acquisitions, and provide evidence suggesting that these are exogenous to physician behavior and characteristics. Focusing on the mainstay drug class used to treat inflammatory bowel disease, my analysis reveals that payments have a class-expanding effect (rather than a brand-substituting effect). I find that stopping brand-related payments to physicians reduces the probability that the physicians would prescribe any brand from the class by around five percentage points. I use these estimates to decompose payment effects to permanent and transient components; I find that at most 50 percent of brand-specific effects and 80 percent of class-specific effects are persistent. Chapter two provides a model of strategic interaction between public and private insurers in setting prices for physician services, where physicians are allowed to choose a desired insurer-specific quality of service to patients. We test our model using a unique dataset of private insurer procedure prices from 2011-2013, which we combine with Medicare price data and physician utilization data. We find that private payers act as if they consider public prices as strategic complements, by using their rates as a means to compete over the physicians' supply of services with the public payer. Moreover, we find that physicians respond to these price differences: Measuring variations of private to public price ratios within geographic regions (3 digit zip codes), we estimate that a 10 percent increase in this ratio leads to a 1.2-3.2 percent d

Thesis (Ph. D.)--University of Rochester. Dept. of Economics, 2013., Chapter 1 estimates the compensating differential with moral hazard. I show that, in an environment of asymmetric information, the wage premium of an unpleasant job attribute not only includes the compensating differential but also the economic rent that stimulates worker effort. The wage premium is equal to the compensating differential if and only if workers are perfectly monitored. Estimates of a structural model using data from the National Longitudinal Survey of Youth find that the wage premium of risky jobs is 6.6%, which is decomposed into a compensating differential that accounts for 2.8% and an efficiency wage premium that accounts for 3.8%. Since the wage premium is interpreted as the compensating differential in a hedonic wage regression, the results show that the reduced-form estimate of the compensating differential has upward bias. Chapters 2 and 3 examine the problem of racial segregation in friendships. Chapter 2 estimates the effect of friendship segregation on educational achievement. Estimating models with data from the National Longitudinal Study of Adolescent Health, I find that a higher exposure to minority friends lowers the standardized test scores of minorities, but do not have significant effects on whites. Specifically, a 10% increase in the proportion of minority friends lowers 4.8% of a standard deviation in the standardized test scores of minorities. The effect is large enough to conclude that friendship segregation increases the minority-white achievement gap. Completely integrating friendships can close 30% of the gap. Chapter 3 evaluates the effectiveness of busing policy on friendship integration. Policymakers can affect students' friendship opportunities by relocating students across schools, but they cannot manipulate students' preferences to achieve desired friendship outcomes. Simulation results show that, by relocating students across schools, the percentage changes in the proportion of minority friends are much smaller than th