Fall 2026 Departmental Seminar Series
Seminar Schedule
September 25, 2026, 10:30 a.m.–12:00 p.m. (noon) | Stockbridge 303
Shresth Garg, University of Pennsylvania
Title: Adaptation For Hire
Abstract: Rainfall shocks are an important source of risk in agriculture. Groundwater irrigation can substitute for missing rainfall, but it requires investment in irrigation infrastructure. In the rice- growing districts of Uttar Pradesh and Bihar, India, we find that rice yields fall substantially more during rainfall shortfalls in districts with limited well access. However, owning this infrastructure is difficult for small farmers to justify because the investment is indivisible. In our data, ownership rises from 35 percent among marginal farmers to 67 percent among large farmers. Small farmers instead rely on rental markets for water. Rental expands access but raises the cost of irrigation. In a dry year, renters’ irrigation spending rises by Rs 1,030 per hectare, twice the increase for owners. We develop a model of ground water markets and use it to study the impact of a subsidy for well ownership. We find that the subsidy is poorly targeted because most recipients already access groundwater through rental markets. For these farmers, the subsidy replaces rented irrigation services with owned infrastructure rather than expanding access.
October 2, 2026, 10:30 a.m.–12:00 p.m. (noon) | Stockbridge 303
Mariyana Zapryanova, Smith College
Title: Liquid Assets: Plasma Donation Income and Crime
Abstract: Paid plasma donation functions as an income-earning opportunity in the US, with access conditioned on passing drug-use screenings, so the arrival of blood plasma center represents a contingent but meaningful income shock for would-be donors. Using detailed data on the location of blood plasma centers and three complementary difference-in-differences research designs, we study the impact of these centers on crime outcomes. Our findings indicate that the opening of a plasma center in a city leads to a 12% drop in the crime rate, an effect driven primarily by property and drug-related offenses. A within-city design confirms these findings, highlighting large crime drops in neighborhoods close to a newly opened plasma center. We also document declines in recidivism among individuals released into areas with newly opened plasma centers, showing that the effects extend across multiple margins of criminal behavior. To probe the underlying mechanisms, we conduct additional analyses that examine heterogeneity across sociodemographic groups, offender age profiles, and incapacitation using center-level opening times. We find strong support in favor of a financial channel, which operates most strongly in less affluent areas and among younger age groups. We find no support for the idea that incapacitation drives our results.
October 30, 2026, 10:30 a.m.–12:00 p.m. (noon) | Stockbridge 303
Jeffrey McCullough, University of Michigan
Title: TBA
Abstract: TBA
December 11, 2026, 10:30 a.m.–12:00 p.m. (noon) | Stockbridge 303
Jayash Paudel, University of Oklahoma
Title: Adaptation to Environmental Stress: Evidence from Himalayan Mountaineering
Abstract: Estimates of the economic damage from environmental change hinge on how much agents can adapt, and on whether that adaptation is endowed or purchased. I study 41,835 climber-expedition records on 103 Himalayan peaks above 7,000 meters from 1979 to 2024, linked to summit-day air pressure reconstructed from reanalysis at each peak's own coordinates and summit elevation. I exploit within-peak deviations in summit-day pressure from its day-of-year climatology, net of peak-year means. A one-standard-deviation decline in summit-day oxygen availability lowers the probability of summiting by 5.9 percentage points on a base of 60 percent and raises mortality by 21 percent of its mean. Wind at the summit has a comparable effect, reducing success by 5.7 percentage points per standard deviation. Climbers raised above 3,700 meters bear 37 percent less of this penalty than those raised at sea level, but this advantage is compositional: it disappears when I compare climbers on the same expedition or Nepali high-altitude workers to one another. Other attributes, including expedition support, prior experience on the peak, and predicted ability, raise summit probability without reducing sensitivity to conditions. Only supplemental oxygen both raises performance and attenuates the response to environmental conditions, although its use is endogenous, and its buffering effect cannot be separated from selection. The findings caution against inferring adaptation from cross-sectional differences in exposure and suggest that policies that facilitate access to protective inputs may be more effective than relying on presumed endogenous adaptation.