Accepted at Review of Economic Studies

I study linear panel data models with predetermined regressors (such as lagged dependent variables) where coefficients are individual-specific, allowing for heterogeneity in the effects of the regressors on the dependent variable. I show that the model is not point-identified in a short panel context but rather partially identified, and I characterize the identified sets for the mean, variance, and CDF of the coefficient distribution. This characterization is general, accommodating discrete, continuous, and unbounded data, and it leads to computationally tractable estimation and inference procedures. I apply the method to study lifecycle earnings dynamics among U.S. households using the Panel Study of Income Dynamics (PSID) dataset. The results suggest the presence of unobserved heterogeneity in earnings persistence, implying that households face varying levels of earnings risk which, in turn, contribute to heterogeneity in their consumption and savings behaviors.

(with Mario Fiorini and Gregor Pfeifer)

Revision requested at Journal of Applied Econometrics

We establish identifying assumptions and estimation procedures for the ATT in a Difference-in-Differences model with staggered treatment adoption in the presence of spillovers. We show that the ATT can be estimated by a simple TWFE regression that extends the approach of Wooldridge (2022)'s fully interacted regression. Moreover, we broaden our framework to nonlinear cases, offering estimation of the ATT by Poisson, Probit, and Logit regressions. We apply our method to revisit a corresponding application from the crime literature. Monte Carlo simulations suggest that our estimator performs competitively.

Many policy interventions are adopted in response to unexpected outcome shocks, such as mask-wearing mandates following surges in COVID cases, job training following earnings dips, and political reforms following declines in GDP growth. Such surge-triggered treatment adoption invalidates the standard parallel trends assumption, making standard difference-in-differences (DID) methods inapplicable. I propose a potential-outcomes framework that generalizes the standard DID framework to accommodate surge-triggered treatment adoption. I show that the average treatment effect on the treated (ATT) is not point-identified but is partially identified, characterize the identified set for the ATT in closed form, and propose an easy-to-implement estimation and inference procedure based on Stoye (2020). As an empirical application, I study the effect of mask-wearing mandates on reported COVID cases in the United States.

Seoul Journal of Economics, 2025, 38(1):69-84.

This review provides an introduction to identification and estimation methods for fixed-effects models with heterogeneous coefficients, which require identification strategies that are notably different from those for standard fixed-effects models. The strategies imply consistent estimation methods for the parameters of interest, which are also different from those used in standard fixed-effects models. As an introductory review, this work defers detailed implementation procedures for the estimation methods to future studies.

(with Priscilla E. Greenwood, Nancy Heckman and Wolfgang Wefelmeyer)

Statistical Inference for Stochastic Processes, 2017, 20:237-252.

We consider estimation of the drift function of a stationary diffusion process when we observe high-frequency data with microstructure noise over a long time interval. We propose to estimate the drift function at a point by a Nadaraya-Watson estimator that uses observations that have been pre-averaged to reduce the noise. We give conditions under which our estimator is consistent and asympotically normal. Its rate and asymptotic bias and variance are the same as those without microstructure noise. To use our method in data analysis, we propose a data-based cross-validation method to determine the bandwidth in the Nadaraya-Watson estimator. Via simulation, we study several methods of bandwidth choices, and compare our estimator to several existing estimators. In terms of mean squared error, our new estimator outperforms existing estimators.