Working Papers
Does Market Power in Local Agricultural Markets Hinder Farmer Climate Change Adaptation?
(with Ruozi Song and Jeffrey E. Sun)
Abstract | Draft | Mercatus Podcast
What role do government policies which distort market competition play in impeding farmers’ climate change adaptation? We study this question in the context of India, where longer-run adaptation to climate change has been inadequate — posing a considerable risk to its ∼250 million agricultural workers. We exploit spatial discontinuities in intermediary market power, created by state-level laws that restrict farmer-intermediary transactions to the same state, to determine how spatial competition affects farmers’ adaptation. We find that a farmer selling in the 75th percentile of the competition index compared to one that faces the 25th percentile of the competition index achieves a 4.9 percent higher output for each additional day of extreme heat. This effect is driven by increased input usage by farmers in anticipation of higher prices after climate shocks, an effect limited only to to high competition areas. We then propose and estimate a quantitative spatial trade model with intermediary market power to examine the welfare implications of higher competition for adaptation. Our structural estimates suggest that the farmer’s economic loss (i.e. their climate damage function) due to extreme weather could be mitigated by 13.8 percent if government regulation distorting market competition are dismantled. These results highlight the importance of understanding the political economy of reforming these competition-distorting laws to accelerate climate change adaptation.
Global Evidence on Urban Resilience and Adaptation to Flood Risk
(with Sahil Gandhi, Matthew Kahn, Somik Lall and Vaidehi Tandel)
Submitted
Abstract | Latest Draft | NBER Draft | VoxEU | WEF
We study how flood impacts, resilience, and adaptation differ across cities worldwide. Combining 3,931 major floods (2000–2023) with data on 9,468 cities in 175 countries, we show: (i) floods reduce urban economic activity by 4.5 percent on average, with losses 3.5 times larger in low-income coun- tries; (ii) cities are becoming more resilient, as sensitivity of economic activity and mortality to floods declines over time, driven by experience-based learning; and (iii) adaptation mechanisms vary by in- come — dams mitigate losses globally, while migration and critical infrastructure enhance resilience only in high-income cities. The disproportionate flood impacts and limited adaptation effectiveness in developing nations highlight an adaptation gap.
Do Water Audits Work?
(with Jesper Akesson, Robert Hahn and Robert Metcalfe)
Revise and Resubmit, Economic Journal
Abstract | Latest Draft | NBER Draft | The National News
Water suppliers are showing greater interest in using different mechanisms to promote conservation. One such mechanism is conducting home water audits, which involves assessing water use and providing tailored suggestions for conserving water for residential customers. Yet, very little is known about the economic impacts of these water audits. This paper helps fill this gap by implementing a natural field experiment in the United Kingdom. The experiment randomly allocates 45,000 water customers to a control group or to treatment groups that receive different behavioral encouragements to take-up an online water audit. Our analysis yields three main findings. First, encouraging subjects to participate in an audit with financial incentives reduces household consumption by about 17 percent over two months. Furthermore, we find that the size of the financial incentive used to encourage conservation matters for take-up, but not conservation. Second, although there are substantial improvements in water conservation for some interventions, they do not appear to yield net benefits of more than £1 per person in most cases. We also implement a marginal value of public funds approach that considers benefits and costs and reach a similar conclusion. Third, we find that targeting high users could double the effectiveness of the financial incentive interventions.
Maintaining Vehicle Emission Reductions Requires Policies to Remove Cars from Highways
(with Antonio M. Bento, Ruozi Song and Andrew R. Waxman)
Submitted (draft available on request)
Abstract | Draft
Cities have had some success to date in reducing pollution via tailpipe emission standards. We show these gains can be undone when cleaner vehicles are operated at inefficient speeds. We show how anti-congestion policies can address air pollution impacts by removing cars from the roads during the most congested hours. In principle, this could occur through public transit expansion or other transportation policies that disincentivize driving. This paper estimates the localized effect of automobile congestion on air pollution to better characterize the benefits of anti-congestion policies on human health. We leverage air pollution sensors on Google Street View cars, combined with fine grain speed and vehicle density observations on Los Angeles highways, to understand this effect at a level of granularity never previously considered using regression methods. Our findings reveal that higher pollution occurs at very low and very high speeds due to lowered engine efficiency.
Seasonal Climate Migration: Evidence from India
(with John Firth, Clément Imbert, Anant Sudarshan, and Tiffany Tee)
(draft available on request)
Abstract
This paper quantifies seasonal migration in response to weather shocks across India. We use station-to-station passenger data from the Indian Railways to construct district-to-district seasonal migration flows between 2011 and 2020. We find that seasonal migration increases when temperatures exceed optimal levels during the growing season, and decreases when temperatures remain in the optimal range. Seasonal migration responds not only to weather shocks in the current year, but also to cumulative shocks over the previous five years. Over a five-year interval, the magnitude of the seasonal migration response to climate shocks is comparable to that of permanent migration in the 2001 and 2011 population censuses. Our findings suggest that a substantial share of climate migration in India is seasonal rather than permanent.
Selected Work in Progress
How Can ML Improve Environmental Inspection Targeting: Evidence from the Clean Air Act and Clean Water Act
(with Michael Greenstone, Vishal Vincent Joseph, and Andrew Schallwig)
Can Prices Drive Conservation? An Experimental Approach to Mitigate Information Frictions in Utility Markets
(with Jesse Buchsbaum and Michael Greenstone)
(RCT underway)
Summary
Utilities rely on price-based policies to incentivize conservation by charging higher rates for higher consumption. However, such policies are often ineffective due to information frictions, as households tend to have limited understanding of price schedules and/or their consumption. The proposed study aims to estimate the type and magnitude of these frictions and to experimentally test the impact of interventions designed to mitigate them. We partner with a public water utility in Westminster, Colorado, to develop innovative information interventions that leverage real-time data from advanced metering infrastructure (AMI) to provide residential customers timely information on water price tiers and quantity consumed. We propose implementing a randomized controlled trial with 19,000 households to identify which information constraints are binding and to measure the effectiveness of real-time consumer notifications that increase the salience of i) prices; ii) quantity consumed, and; iii) both prices and quantity consumed. We will measure outcomes via administrative data (water consumption, bill magnitude) and surveys (customer perceptions and understanding of price schedules, water consumption, utilization of customer portal to access real-time household-specific data). Survey data will be collected in three waves: i) baseline (spring 2025, prior to the RCT); ii) midline (fall 2025,immediately post-interventions to measure short-term impact), and; iii) endline (summer 2026, to measure impact persistence). If proven successful, these real-time interventions will help consumers to respond optimally to prices, unlocking the potential of priced-based policies to achieve allocative efficiency. The study is highly policy-relevant as the low-cost interventions can then be tested and scaled in other jurisdictions and across utility types (e.g., electricity and natural gas).
Do Irrigation Alerts Change Behavior? A Field Experiment on Residential Water Use
(with Jesse Buchsbaum, Michael Greenstone, and Javier Alejandro Lopez-Aguilar, )
(RCT completed, analysis underway)
Summary
Water scarcity is becoming an increasingly pressing issue, prompting utilities and regulators to explore cost-effective strategies for reducing residential water consumption. While price-based policies are commonly used, empirical evidence suggests that they often have limited impact. In collaboration with a public water utility in Colorado, we designed and implemented a randomized controlled trial to assess the effectiveness of informational interventions in reducing inefficient lawn irrigation practices. Using real-time data from advanced metering infrastructure (AMI), we identified irrigation episodes among 7,000 households during the summer of 2024. Our intervention delivered targeted alerts to customers who watered their lawns during inefficient daytime hours or exceeded the recommended irrigation frequency. We find that the intervention led to a 5% to 10% reduction in daytime irrigation events and weekly irrigation frequency, depending on the treatment. Preliminary results suggest that total water consumption is significantly reduced in the short-run in response to our treatment.
Empowering Environmental Governance with Satellite Data: A Global Randomized Control Trial
(with A. Patrick Behrer, Michael Greenstone, Olga Rostapshova, Ruozi Song, Yixin Sun, and Shaoda Wang)
(Pilot underway)
Summary | UChicago Seed Fund Project
Climate change and air pollution are closely intertwined and share root causes — mainly fossil fuel combustion. Air pollution, with its immediate and visible health impacts can serve as a powerful entry point for broader environmental action. Yet much of the world lacks local pollution monitoring, limiting public awareness and pressure for reform. This project will run the first global randomized control trial to test whether delivering real-time, localized air quality data to stakeholders can raise awareness, prompt engagement, and reduce pollution. The project leverages novel, satellite-derived PM2.5 estimates to track daily air quality across 1,000+ cities and send alerts via social media to government officials, regulators, NGOs, journalists, and citizens in treated cities. This low-cost, scalable intervention offers the potential to bridge environmental data gaps and strengthen accountability around the world. Providing transparent information on pollution exposure can empower communities to demand and shape effective responses to air pollution and climate change.
Do Nudges Backfire? It Depends on the Experimental Design
(with Robert Hahn and Robert Metcalfe)
Summary
Water suppliers are showing greater interest in using non-price mechanisms that can help encourage conservation. One such mechanism is information campaigns, which provide households with actionable insights into ways to reduce water usage. Most of these campaigns are targeted towards the general population and aimed at reducing wastage during essential water usage. Little is known about the efficacy and efficiency of campaigns that target recreational water usage in high income households. This paper helps fill this research gap by implementing two natural field experiments in Texas, United States. We first implement a pilot, inviting 5,705 customers with swimming pools to participate in a draw that awarded a free pool checkup. Households which expressed interest are randomly allocated to three groups—a Checkup group which received a free pool checkup, a Mailer group which received an informational sheet on pool care, and a control group. Subsequently, we implement a second experiment, wherein the Mailer treatment is randomized to the entire population of pool owners under the purview of the utility. Our analysis yields three main findings. First, the information campaign backfires, with households in the treatment group consuming 270 liters per day (22 percent increase) more relative to the control group six months after treatment. Heterogeneity analysis based on observables shows that even the Checkup treatment backfires, with relatively richer households increasing their consumption by 430 liters per day (36 percent increase) half a year after treatment. Second, the scale up of the pilot did not yield, on average, a significant increase in consumption. However, even in this case, richer households in the treatment group have a significantly higher consumption of 128 liters per day (10 percent increase) six months after treatment. Third, reweighting our estimates from the pilot does not accurately predict the results from the scale up, which underscores the role of unobservable household characteristics in determining self-selection into the pilot. Our results have important implications for targeting of behavioral interventions for water conservation, and scale up of field experiments.