Browsing by Author "Garlick, Robert J"

This dissertation includes three chapters in behavioral labor economics. In the first chapter I study with Victoria Lee the effect of wage transparency policies on workplace interactions and workplace choice. We run an online experiment in which we manipulate whether wages are known or secret, and also control the wage allocations that participants faced in the teams. The results show that wage transparency can affect the level of hostility as well as the target of the hostility in the teams -- as measured by a punishment game. These treatment effects are moderated by the level of wage inequality within the teams. We also find that wage transparency can alter which job offers participants accept. This result suggests that wage transparency policies can have general equilibrium effects due to workers changed sorting behavior. We argue that these two sets of results can be explained by the presence of social preferences and participants' inaccurate beliefs about the wage allocations that are corrected when wages are transparent.

In the second chapter Nayoung Rim, Roman Rivera, Bocar Ba and myself are studying racial bias within the police. Although there is substantial evidence showing racial bias in firms' hiring decisions, less is known about bias in career recognition. We construct a novel dataset of police award nominations to measure bias against minority employees. Exploiting quasi-random variation in supervisor assignment and randomized timing of annual evaluations, we find that white supervisors are less likely to nominate black officers than white and Hispanic officers leading up to and during the evaluation period. Further, the black-white nomination gap widens with the number of arrests. These patterns suggest that the disparity is not due to in-group favoritism towards white officers but rather bias against black officers. We conduct an online experiment to examine evaluator engagement and find that evaluators are less likely to engage with black officers vs. white officers. Our findings suggest bias in career recognition may have important implications for the black-white promotion gap, the lack of diversity in upper-management positions, and, ultimately, the racial wage gap.

In the last chapter I estimate the effects of behavioral interventions on sleep and cognitive performance among undergraduate students. Three interventions are tested during a three-week long framed field experiment: bedtime text reminders, sleep hygiene tips and an educational meditation video. Throughout the experiment participants' sleep patterns are measured using fitness trackers. The results show that sleep quantity and quality may not be easy to move with mild nudges -- although the estimates are noisy due to the small sample size. Estimates with panel data methods support the positive link between sleep and cognitive outcomes.

This dissertation considers the role gender plays in labor markets, household decision-making, and health in sub-Saharan Africa.

The first chapter considers the impact of fast Internet access on employment outcomes and household dynamics. I find the introduction of fast Internet to sub-Saharan Africa significantly increased employment for males, but had little impact on female employment. In addition, it significantly increased perceived acceptability, among both genders, of domestic violence against women.

The second chapter considers the differential impact, by gender, of an experimental labor market intervention in South Africa, which measured skills of workseekers and provided a mechanism for workseekers to communicate their results to potential employers. I find that men experienced a larger effect of the intervention on employment outcomes than did women. This difference is largely explained by pre-existing differences between genders, rather than differential responses to treatment.

The third chapter considers the factors that contribute to female genital cutting (FGC) in Mali and tests various hypotheses to explain the persistence of the tradition. I find that maternal preference is pivotal in the decision to cut daughters. I also find that the marriage market hypothesis and the identity hypothesis of FGC decision-making are alone insufficient to explain persistence.

This dissertation explores questions in political economy and in development

economics. I ask and answer two research questions.

First, I look at whether peaceful or violent protests are more effective at

steering policy change. I study this question in the context of the US Civil

Rights Era, and evaluate the effects of protests on legislator votes in the

US House. I use a fixed-effects specification, and find that peaceful protests

caused a liberal shift and therefore were effective from the point of view of

the Civil Rights Movement but violent protests caused a conservative shift

and therefore backfired.

Second, I look at whether the structure of social networks in rural West-

ern Kenya is affected by a large development intervention. In joint work with

Robert Garlick and Kate Orkin, we evaluate the effects of a large unconditional

cash transfer and a psychological intervention. We cross-randomize

villages into these two interventions, and measure household interactions in

four types of networks: talking about goals, talking about challenges, giving

money or goods, and receiving money or goods. We estimate effects on total

link counts, measures of homophily, and measures of link intensity.

This dissertation presents three essays on topics that lie at the intersection of development and labor economics. The essays relate to poverty and inequality in the United States and internationally, focusing on labor markets, housing markets, and human capital. The first essay sheds light on how land use regulation affects local economic development, operating through local housing and labor markets. I study this topic in the context of federally recognized American Indian reservations, where land tenure categories include fee simple land, which is free from transaction restrictions, and trust land, which is held in trust by the US federal government and associated with restrictions on transactions. Drawing on a range of hard-to-access microdata products, my analysis indicates that land use regulation lowers local wages and employment but that it is possible to foster economic development without changing the status of the land. In fact, land use regulation increases the likelihood that the local population benefits from demand-driven, place-based policies. Using casino adoption on reservations as local labor demand shocks, I find that reservations with a larger share of land in trust experience larger increases in real wages following economic shocks. The second essay paints a picture of poverty in rural America with a particular focus on relating the Black-White wealth differential to the Black-White human capital differential. Using a low-asset subgroup of the rural population, I show that when the wealth gap narrows, we are less likely to observe racial heterogeneity in rates of high school completion and teenage motherhood. This essay establishes two important insights that help explain the observed race gaps in cities: (i) Black-White differences in wealth are much larger in urban areas than rural areas, and (ii) household income may not accurately reflect poverty. The third essay turns to poverty in low-income countries, focusing on labor market frictions and employment outcomes in South Africa. In this essay, I describe the design and implementation of a large-scale field experiment that tests whether an online professional networking platform changes the labor market engagement of youth. Specifically, my co-authors and I test whether training young work-seekers to use LinkedIn improves their employment outcomes. We find that the LinkedIn treatment rapidly increases employment at the extensive margin.