Job Market Paper
This paper introduces and quantifies a new mechanism, precautionary electrification, whereby households adopt electricity instead of gas to insure themselves against volatile gas prices. The analysis is motivated by two facts: natural gas prices in the United States have been far more volatile than electricity prices over the past two decades; and lower-income households are much more likely to rely on electricity for heating than higher-income households. Using state-level data from 1999 to 2023, I show that greater gas price volatility leads to higher electrification, particularly among low-income households. I develop a structural model of household energy choice with non-homothetic preferences and costly fuel switching. The model matches the empirical relationship between income, fuel choice, and energy expenditure, and implies that gas price volatility generates regressive welfare losses and a trade-off between short-term price insurance and long-term electrification incentives.
Working papers
The Financial Instability Contribution Scores: Theory and Application (with Joaquim Guilhoto and Gregory Legoff)
Carbon taxes can increase default rates in greenhouse gas-emitting sectors, potentially destabilizing the banking system. This paper introduces Financial Instability Contribution Scores (FICS), a metric designed to assess how different industries contribute to banking instability and enable cross-sector comparisons. Using a simple model incorporating heterogeneous sectors, default risks, and an input-output framework, FICS measure each sector’s impact on a representative bank’s z-score, both with and without a carbon tax. The metric accounts for three key factors: sectoral default risk, debt-to-asset ratio, and demand effects. Applying FICS to U.S. data highlights utilities, agriculture, and air transport as the sectors most likely to threaten banking stability. This approach provides a straightforward, quantitative tool for evaluating the financial risks associated with climate policies.
The Energy-Efficiency Elasticity of Durable Demand (Draft available upon request)
Energy-efficiency adoption by households is a micro margin with macro consequences for aggregate energy demand, expenditure shares, and welfare. I quantify the energy-efficiency elasticity of durable demand in a heterogeneous-agent Bewley–Aiyagari model with lumpy replacement, fixed adjustment costs, and endogenous utilization. Two durable types (baseline vs. efficient) are priced along an empirically disciplined efficiency–price locus. Calibrated to U.S. room air-conditioner data, the model matches ownership and spending moments and delivers aggregate counterfactuals. A 10% improvement in efficiency raises adoption of efficient durables by 9% even when purchase prices rise accordingly, reduces residential energy use by 2%, and increases welfare. This framework supplies policy-relevant elasticities and sufficient statistics that can be embedded in DSGE environments to evaluate efficiency standards, rebates, and shocks to energy prices.
Work in progress
Equitable Green Transition (with Glen Kwende)
Innovation and Natural Disasters (with Meha Sadasivam)
Macroeconomics and Energy-Efficiency Heterogeneity
Other writing
Greeniums and Sovereign Debt (Short policy note prepared for COP30 Economic Council Chair, 2025)
This report explores the presence and significance of the greenium—the yield differential between green and conventional (“brown”) bonds—in sovereign debt markets. The evidence suggests that sovereign greeniums are modest, averaging around 3 basis points, and can vary over time. While some emerging markets appear to exhibit higher greeniums, the robustness of this finding remains uncertain. Case studies from countries like Germany and Denmark, which have issued near-identical twin bonds, confirm the existence of greeniums, albeit small in magnitude.
Pre-PhD
We report results from an online economics experiment that examines the effect of nudging retirement savings decisions. In the experiments, participants make decisions in a finitely repeated retirement savings game, in which income during working years is uncertain, and retirement age is known. Participants, who are household financial decision-makers, are nudged with automatic savings in each period of the game. We find that that the nudge simply replaced natural decision-making observed in the absence of a nudge in this experiment, even to the extent that it resulted in nearly identical inferred decision rules. This surprising result highlights the unpredictability of the effect of nudging human behavior.
We report results from an economics experiment that examines the role of financial literacy in retirement savings. In the experiments, participants make decisions in a retirement savings game, in which income during working years is uncertain. Participants are nudged to varying degrees with automatic savings in each period of the game. Some participants receive financial literacy training in the form of training to compute the expected savings needed at retirement to smooth consumption over the entire life cycle. We find evidence that literacy increases savings and improves efficiency. Our finding has implications for choice architecture for retirement savings.