Research

We outline a macro-pandemic model where individuals within occupations can select into working from home or in the market. Market work increases the risk of infection. Occupations differ in the ease of substitution between market and home work, and in the risk of infection. We examine the evolution of a pandemic in the model as well as its macroeconomic and distributional consequences. The model is calibrated to British Columbian data to examine the implications of shutting down different industries by linking industries to occupations. We find that endogenous choice to self-isolate is key: it reduces the peak infection rate by 2 percentage points even without policy mandated lockdowns. Risk aversion generates a large drop and slow recovery. The model also produces widening consumption inequality, a fact that has characterized COVID-19.

Safe Implementation (Gavan and Penta 2025) combines standard implementation with the requirement that the implementing mechanism is such that, if up to k agents deviate from the relevant solution concept, the outcomes that are induced are still “acceptable” at every state of the world. In this paper, we study Safe Implementation of social choice correspondences in mixed Nash equilibrium. We identify a condition, Set-Comonotonicity, which is both necessary and (under mild domain restrictions) almost sufficient for this implementation notion.

This paper investigates how households smooth consumption against idiosyncratic wage shocks in recessions and expansions. Labour market uncertainty amplifies during recessions, captured through the cross-sectional dispersion of wages. I focus on the relative contribution of two insurance mechanisms to wage changes, namely, adjustments in labour supply and assets, during periods of high and low uncertainty. I exploit variation in expenditure, hours worked and wages over the business cycle to wage shocks, and apply it to US household panel data. I document a new empirical fact — the contribution of labour supply to consumption smoothing increases during labour market downturns. Households with low liquid wealth show the strongest asymmetric labour supply response between recessions and expansions. To jointly explain these empirical facts, I develop an incomplete market life-cycle model with multiple asset-types (liquid and illiquid) and an aggregate state that affects wage dispersion. The model shows that the key mechanism is the shift in portfolio composition towards liquid assets during high uncertainty periods.

This paper illustrates the role of low-skilled immigrants’ location choice as a channel through which local labour markets adjust to automation. We employ a shift-share instrumental variable approach to demonstrate that low-skilled immigrants are more mobile than low-skilled native born in response to robot exposure. Low-skilled immigrants are less likely to enter and more likely to exit from highly robot-exposed regions. Immigrants’ location decisions attenuate wage losses due to robot exposure for low-skilled natives. Low-skilled native workers experience a 0.07 percentage point smaller decline in wages comparing commuting zones at the 50th and 25th percentiles of low-skilled immigrant shares.

We study the macroeconomic implications of agricultural subsidies for the agricultural productivity gap and consumer welfare. We develop a dynamic general equilibrium model, where heterogeneous individuals endogenously sort between sectors and crop types, while facing mobility costs and incomplete markets. The benchmark economy, calibrated to India, features two tax-financed policies: intermediate input price subsidies and a support price program for staple crops. We find that eliminating either policy worsens agricultural productivity relative to non-agriculture, but increases welfare by reducing the tax burden, disproportionately benefiting asset-poor households. This highlights a trade-off between promoting productivity and improving welfare in industrial policy design.

Does improving access to financial institutions facilitate consumption smoothing at different stages of development? Using data for 135 countries, we show that greater financial access raises consumption volatility relative income volatility in developing countries but lowers it in developed economies. We develop a parsimonious small-open economy model with household heterogeneity in financial access to show that the asymmetric nature of technological shocks experienced by developing and developed economies rationalizes this finding.

This paper examines the source of gender and marital status differences in portfolio choices across U.S. households. Using the Panel Study of Income Dynamics (PSID) and the Survey of Consumer Finances (SCF), we find evidence that single female-headed households invest the least in risky assets, followed by single male-headed households. Further, married households invest the most in risky assets. Towards explaining these differences in portfolio allocations, we further document that women earn lower incomes and face higher individual income risk relative to men. To quantitatively investigate the importance of these gender differences in income profiles, we develop a two-asset incomplete market life-cycle model with heterogeneous households. Using the model, we show that the gender wage gap is important in explaining portfolio choice differences during the initial years of working life; however, higher income risk leads to lower risk-taking behavior by female-headed households in later working years. We also show that dual-earner households exhibit higher investment in risky assets compared to single-earning couples which is consistent with our empirical findings, indicating a role for spousal insurance.