We study whether banks internalize social deposit risk when arranging syndicated loans to carbon-intensive (‘brown’) borrowers. We construct Democratic Deposit Risk (DDR), a bank-year index that scales the size of a bank’s branch network by the partisan tilt of the counties it serves. Merging FDIC branch data, county-level election results, and U.S. syndicated loans from Dealscan, we document two patterns. Banks with higher DDR are less likely to participate in brown loans and, when they do, structure them with more intensive syndication. These results link banks’ geographic deposit bases to their asset allocation and risk-sharing choices.

India is the world’s largest groundwater user, with 90% used for agriculture. Groundwater, however, is a common pool resource, generating a tragedy of the commons that threatens agricultural sustainability. We develop a parsimonious model to show how a popular policy intervention — subsidizing efficient irrigation technology — can exacerbate distortions away from socially optimal groundwater extraction. We test the model’s predictions by leveraging geophysical variation in extraction externalities and a $1.35 billion program subsidizing efficient irrigation. Consistent with the model’s predictions, the policy’s impact depends on the severity of extraction externalities: extraction falls 9.2% in low-externality areas but rises 11.0% in high-externality areas. Low-externality farmers maintain cultivation using less groundwater, while high-externality farmers cultivate more intensively. Finally, the program causes climate-adaptive responses in low-externality areas – reducing extraction during normal rainfall and increasing it during droughts – but the opposite pattern in high-externality areas, consistent with climate maladaptation. Our findings illustrate that the same common pool conditions that typically justify an intervention may also determine its welfare implications.

We study how physical access to ChileAtiende—Chile’s nationwide one-stop service platform—shapes take-up of social programs, with a focus on housing subsidies. We assemble a national, multi-year panel that links (i) the timing and location of ChileAtiende office openings/closures and office-level visit logs, (ii) administrative records on applications and awards for major benefits (with DS01 housing subsidies as a core case), (iii) beneficiary origin and destination addresses to measure distance to the nearest office, and (iv) dated policy announcements that plausibly shift information demand. Our central question is whether improved proximity to an office increases applications and awards—and for whom—versus simply reallocating demand across channels. We test whether effects are stronger for households living nearer (vs. farther) to offices, around office openings, and in response to salient national announcements. The design leverages staggered office rollouts and announcement timing in a difference-in-differences framework with granular location and time fixed effects.

*Currently raising funding. This study examines how financial intermediary incentives and financial literacy interventions shape financial product uptake and consumer protection among unbanked smallholder farmers in rural Colombia. Through a randomized controlled trial, we vary agent compensation structures (fixed versus commission-based pay) and client-focused messaging (financial literacy workshops versus standard promotional materials) to assess impacts on product comprehension, responsible use, and susceptibility to risky financial behaviors. By disentangling the roles of persuasion and information, the study contributes new insights into how agents can serve as effective educators rather than exploitative salespeople. Our findings will directly inform policy and practice, providing actionable evidence for regulators and financial institutions aiming to balance market expansion with consumer welfare.

*Paused due to USAID funding cuts. In the face of increasing climate risk, Zambia has suffered from substantial droughts in the past 20 years, with severe consequences for smallholder farmers. We consider two broad approaches to help farmers adapt and build resilience to this new reality: the provision of crop insurance, to insure farmers in the face of increased risk; and the encouragement of diversification and climate-resilient crop choice. While these two approaches are often considered in isolation, we argue that programs like Zambia’s Food Security Pack program should consider them in unison. Farmers face initial risk and uncertainty when changing crops, making them reluctant to do so, even if it would reduce risk in the long run. Insurance, in overcoming some of this risk, has been shown to increase ex-ante investments, including by changing to more productive crops and varieties (Karlan et al. 2014). Moreover, switching to crops more suited to local agroeconomic conditions is argued to increase productivity substantially (Adamopoulos and Restuccia 2021). Switching to riskier but higher average return crops may thus be another benefit of insurance provision. We study the logistical feasibility of a large randomized controlled trial with Food Security Pack beneficiaries, farmers who meet the income and wealth requirements delineated by the government. Participants are cross randomized into free insurance and crop choice incentive treatments, in order to determine the added value of insurance in experimentation with new crops and higher productivity.