Embedded Files
Publications
Publications
Abstract: The accumulation of large genetic data is crucial for the scientific advancement of genetic research and precision medicine, but various participation biases threaten the validity of genetic research datasets. To better understand the decision to participate and its relationship with economic incentives and preferences, we studied the stated willingness to donate DNA for science by saliva sample in a representative panel of Dutch households. There were two randomized treatments, varying (i) the information material on benefits and risks and (ii) the intended financial incentive. The first treatment had no detectable effect, suggesting insensitivity to the information material. The higher incentive conditions had modest and diminishing effects, suggesting that offering higher incentives is not cost-effective. Stated reasons not to donate DNA concentrated on personal risks, e.g., privacy violations and data exploitation. Accordingly, stated risk willingness was found strongly associated, followed by trust and positive reciprocity. Revealed economic preferences were not associated. The results support previous findings for self-rated health, interpersonal trust and confidence in science or societal institutions; but not for certain demographic variables (e.g., age, education and religiosity). We conclude by proposing strategies to encourage participation, e.g., to reallocate resources to risk-minimizing or compensatory measures.
Working Papers
Working Papers
Abstract: India, like many other countries, is marked by persistent son preference and resulting gender inequalities. In response, policymakers instituted several Conditional Cash Transfer (CCT) programs to encourage the birth and education of girls in smaller families, along with sterilization. I evaluate two large CCTs that provide generous subsidies for girls, one that is restricted to families without boys and one without such a restriction. Using retrospective fertility data from the Demographic Health Survey and employing a difference-in-differences methodology, I find that the policy without the restriction to girls increases the probability of having a subsidy-eligible child composition by 8% relative to women in neighboring control states, driven by an increase in families with a composition that includes a boy. In contrast, all-girl compositions do not increase under either policy, suggesting that the demand elasticity for daughters is higher in the presence of a son. Both policies increase sterilization at subsidized compositions, but the girl-restricted policy does so without affecting fertility, suggesting that it inefficiently incentivizes women toundergo the risky procedure. However, the girl-restricted policy, by requiring subsidized girls to start school one year before the legally mandated age, increased early school enrollment.
Abstract: The US spends billions of dollars annually on policies to promote upward economic mobility. This paper studies the effectiveness of one such approach: targeted wage subsidies. We examine a large federal program, the Work Opportunity Tax Credit (WOTC), which covers up to 40% of wages for certain categories of disadvantaged workers and subsidizes over two million hires annually. Using linked administrative data, we apply four complementary quasi-experimentalapproaches to assess WOTC’s impact on labor market outcomes, including age-based eligibility discontinuities, eligibility expansions, and changes to program participation costs. We find precise null effects on hiring, employment, and earnings across all specifications, where we can rule out effects on hiring as small as 0.2 percentage points. Instead, employers hire the same workers as in the absence of subsidies, and subsidies operate as a pure transfer to the firms, where most of the benefits accrue to a small number of large firms. To understand these results, we collect original data on the hiring processes of WOTC-utilizing firms. These data suggest that firms often silo information on candidates’ subsidy eligibility from managers making hiring decisions, with some evidence that this is due to concerns about legal liability.
Abstract: I estimate the causal impact of an affirmative action policy, the implementation of job quotas for a socio-economically disadvantaged group in India, the Other Backward Castes (OBCs), in 1993, on their inter-generational education and occupation mobility. I use data drawn from the Indian Human Development Survey (2011-12), and employ a difference-in-differences strategy motivated by the fact that only OBCs of a school or college-going age in 1993 could potentially be impacted. Older OBCs would have already made human capital investment decisions that would be impossible or too costly to change in response to the policy. I find that the job quotas resulted in an increase in the absolute upward education mobility of OBC sons, likely reflecting heightened educational aspirations, as indicated by (1) an increased probability that their education is greater than their father’s and (2) an increase in the years of education of sons born to illiterate fathers. On the other hand, I find that the policy resulted in an absolute downward occupation mobility of OBCs, as indicated by a decrease in the probability that a son born to a father employed in an unskilled or farming occupation is employed in a professional or skilled occupation, even though public-sector employment among OBC youth expanded.
Selected Works in Progress
Selected Works in Progress
A Structural Model of Fertility Decisions in India
Abstract: In India, son preference arises due to a cost differential between raising boys and girls as well as cultural preferences that place a premium on the first son. I estimate a finite-horizon dynamic stochastic programming model of fertility decisions in India, which generates insights into the sex composition of children and adoption of sterilization. I use the estimated model to study the effect of counterfactual policies such as changing the subsidy amount and eligibility conditions of cash transfer programs aimed at encouraging the birth of girls.
Recreational Marijuana Legalization and Child Maltreatment (with Jessica Pac, Christine Piette Durrance, and Lauri Luosta)
Abstract: Since 2012, 24 U.S. states and Washington, D.C. have adopted recreational marijuana laws (RMLs), prompting concerns about potential effects on child welfare. In this paper, we examing this question using child protective services report data from the National Child Abuse and Neglect Data System and death certificate data from the National Vital Statistics System (NVSS). We implement a synthetic difference-in-differences design that exploits the staggered adoption of RMLs across states between 2010 and 2023. We find no evidence that RMLs affect overall child deaths, maltreatment deaths, or CPS reports of abuse or neglect. The opening of recreational dispensaries is associated with small, statistically insignificant reductions in child deaths. Results are robust to stratification by state poverty rates and opioid epidemic intensity. Overall, our findings suggest that recreational marijuana legalization has not meaningfully altered child maltreatment outcomes.
Privatizing Medicaid: Causal Machine Learning Evidence on Wait Times from Louisiana’s Managed Care Transition (with Anita Raj)
Old-Age Support in India: The Role of Child Composition and Pension Systems
Gendered Labor Market Response to Adverse Weather Events in Louisiana
Page updated
Report abuse