Research

JOB MARKET PAPER -- Economic and Psychological Returns to Social Relationships: Alleviating Constraints to Network Formation in Malawi | Draft | AEA RCT Registry

I measure exogenous impacts of social linking among a sample that is socially isolated and economically vulnerable. I experimentally induce social interactions between female rural migrants in Malawi by facilitating poorer women in inviting other poor women, wealthier women, or a random mix of women for a shared meal. I cross-randomize a voucher for meat, a high-price good that improves the experience of sharing a meal. Within just one month of the intervention, participants are 25% more likely to earn income from self-employment relative to control. One year later, they experience a 0.13 SD increase in food security and a 22% reduction in depression. Women who invite wealthier guests drive the results on self-employment and food security, while women who invite other poor guests drive the results on depression reductions. Reducing the effort and uncertainty associated with initiating relationships alone leads 81% of women to send an invitation, while subsidizing the price of serving meat encourages cross-social-class linking. I draw three conclusions: (1) all social relationships yield large benefits, but effort costs and information asymmetries inhibit them from forming, (2) different types of relationships are more impactful across different domains, underscoring the value of economically diverse networks, and (3) social connection has a financial cost that reinforces income-based homophily.

Experiential and Social Learning with Agha Ali Akram, Reshmaan Hussam, and Akib Khan | Draft | AEA RCT Registry

We examine complementarities between experiential and social learning in health technology adoption. We conduct an 1800-household field experiment on water chlorination in peri-urban Pakistan. We implement a learning intervention, where participants record and visually track their children's diarrhea rate before and after we offer free chlorine tablets. While there are no differential effects on average between the learning arm and a group that receives only free chlorine tablets, learning arm households who also have a neighbor in the learning arm chlorinate their water at a significantly higher rate for almost one year after the end of the learning intervention. Households not in the learning arm exhibit no difference in behavior by whether they have a neighbor in the learning arm. We propose a model of learning where agents are sensitive to signal acquisition technologies and trust technologies they have personal experience with, giving rise to a complementarity between experiential and social learning. We rule out various alternative explanations, including changing beliefs about the returns to chlorine use. The welfare implications are significant: ITT (TOT) estimates suggest that learning households with learning neighbors exhibit a 0.16 SD (0.51 SD) increase in an index of child anthropometrics after one year.

Distorted Signals and Evaluator Bias: A Two-Stage Experiment on Gendered Hiring in Malawi with Mansa Saxena | Draft | AEA RCT Registry

We evaluate how well-meaning inclusion efforts can backfire if not implemented carefully. We partner with a firm in Malawi to conduct two sequential experiments to study supply-side and demand-side constraints to hiring women for formal jobs. The first experiment focuses on increasing the pool of female applications during a recruitment drive through female-directed advertising. Despite no differences in the objective skills of the female applicants across treated and control areas, the treatment has the perverse effect of leading to a reduction in female hiring. This surprising result informs our second experiment—a resume audit study—where we use real applications from stage one and manipulate application features, while holding qualifications constant, to isolate biases in hiring evaluations. We argue that the treatment backfires due to the combination of evaluator bias, where evaluators place greater weight on soft-skill signals for women, and the treatment leading women to change the way they fill out the application. The treatment weakens the correlation between soft-skill signals and objective technical skills, which are otherwise positively correlated. Combined with evaluators’ greater reliance on soft-skill signals for female applicants, this leads to the screening in of less-qualified women for interviews in treated areas.

Why do we Pay for Symbolism? Evidence from U.S. Geographic Landmark Name Changes with Emily McDonnell. | AEA RCT Registry | Data collection in progress

Do the processes that agents use to establish social and cultural inclusion matter for their ultimate success? We study symbolic policy changes by analyzing reactions to the renaming of U.S. geographic landmarks. Descriptively, people living in zip-codes that experience geographic landmark name changes are more likely to donate to to Republican political candidates and less likely to donate to Democratic political candidates after name removals are announced, but not when the replacement names themselves are announced. In a survey experiment, respondents express much more conservative policy preferences when we frame the 2015 renaming of Mt. McKinley to Mt. Denali as a process enacted by and benefiting groups of people, rather than a passive occurrence. Emphasizing the Indigenous origins of the name ``Denali'' does not generate the same response, and when combined with the active voice treatment actually mitigates the backlash. Jointly, these findings suggest that people lose utility through the processes that seek to enact, remove, or replace symbols, distinct from preferences over the symbols themselves. Crucially, tailored information can psychologically compensate people for these utility losses and reduce backlash.

General Equilibrium Impacts of Social Network Formation. | AEA RCT Registry | Data collection complete, analysis in progress.

Does exogenously generating social inclusion for one person have externalities for others' social inclusion? I leverage cluster-randomized treatment saturation of an experiment that creates exogenous opportunities for social linking, which effectively creates variation in the spatial proximity of Control participants to neighbors' network activity. The treatment design also randomly varies the socio-economic status of the exogenous interactions among low-income treated participants. This design allows me to answer fundamental questions about social network formation that we have limited causal evidence for: Does proximity to social network activity crowd in or crowd out one’s own social ties? Do these effects depend on socio-economic status of the agent who is proximate to others' social network activity, or the socio-economic status of the socially-active nodes to whom she is proximate?

Peer Effects in Adherence to HIV Treatment with Jessica Gallant. Analysis

This paper investigates the role of peer influence in health behavior by studying antiretroviral therapy (ART) adherence among adolescents living with HIV in Malawi, a group that is both developmentally sensitive to peer dynamics and disproportionately affected by poor health outcomes in the context of HIV. With high-frequency electronic medical records of almost 45,000 adolescents, across twelve years, 85 clinics, and 23 districts in Malawi, we track high-stakes health decision-making for as long as adolescents receive care. We utilize three empirical strategies to examine the causal impact of peer behavior on long-term health adherence: a staggered differences-in-differences analysis of the effects of Teen Club entry at the clinic on contemporaneous attendance on scheduled visits; a regression discontuinty design using age cut-offs in Teen Club eligibility and age at the time of initiation to measure long-term effects on outcomes such as mortality; and two instrumental variables specifications to measure how attendance at Teen Club is impacted directly by peer attendance in their prior Teen Club appointment, as instrumented by the peer's attendance record prior to the patient's initiation at Teen Club, or as instrumented by rainfall on the peer's appointment date.

Gendered Misallocation of Agrarian Labor with Nicholas Rahim. | AEA RCT Registry | Trial in progress

We test a novel explanation for disparities in agricultural output between men and women in Malawi: child-bearing as a unique shock to women's labor supply that inhibit efficient land utilization. Pregnancy and child-rearing are negative shocks to women’s available labor. In frictionless land, labor, and credit markets, or with substitutability between household member's labor, women should be able to make productive use of their land when their own labor supply is physiologically constrained. We implement a randomized controlled trial with female farmers who are pregnant or have a child under the age of one year old. We test the impact of subsidizing the cost and search frictions for hiring five days of agricultural labor during the farming season on agricultural and health outcomes.

Public Services and Private Behavior: Evidence from Water Infrastructure in Pakistan with Akib Khan. Data collection complete, analysis in progress

We study how private health behavior responds to new public water infrastructure. During an 18-month randomized controlled trial (RCT) in peri-urban Karachi, we distribute chlorine tablets to households and collect monthly, objective measures of chlorine use and water source choices. Concurrently, local NGOs independently implement water infrastructure projects in a staggered fashion. This natural variation in proximity to new standpipes and boreholes, combined with randomized access to water treatment, allows us to estimate the dynamic effects of infrastructure exposure on private behavior and health outcomes.