I am a Ph.D. candidate in economics at Northwestern University, on the 2026-2027 job market.
My research is in econometrics, with a current focus on nonparametric methods for settings with endogeneity and individual heterogeneity.
This paper proposes a new way to estimate nonparametric structural functions by adapting to the unknown magnitude of endogeneity. Instrumental variables can address endogeneity, but their use comes at a substantial cost in estimation accuracy: nonparametric instrumental variables (NPIV) estimators can converge at very slow rates. By contrast, nonparametric least squares (NP-OLS) estimates the conditional mean function at a much faster rate, but this function generally differs from the structural function under endogeneity. We formalize this trade-off as a minimax estimation problem over the class of data generating processes corresponding to each magnitude of endogeneity. We first show that the optimal rate of convergence ranges from the faster NP-OLS rate to the slower NPIV rate as the magnitude of endogeneity increases. We then construct a feasible shrinkage estimator combining NP-OLS and NPIV that nearly attains the optimal rate in each regime without prior knowledge of the magnitude of endogeneity. The key to this adaptation is a one-sided downward-adjusted estimate of the endogeneity bias component. Simulations calibrated to gasoline demand data show that our adaptive estimator exploits substantial accuracy gains under limited endogeneity while performing comparably to NPIV under severe endogeneity.
This paper proposes nonparametric empirical Bayes confidence intervals (NP-EBCIs) for unobservable individual effects in a normal means model. The oracle intervals are constructed from posterior quantiles under a point-identified, fully nonparametric prior; feasible intervals replace these quantiles with nonparametric estimates. The NP-EBCIs are asymptotically exact in the sense that both their conditional and marginal coverage probabilities converge to the nominal level. The flexibility of this nonparametric construction has an unavoidable statistical cost. We demonstrate that posterior quantiles, unlike posterior means, inherit the severe ill-posedness of nonparametric deconvolution: the minimax optimal estimation rate is logarithmic. This logarithmic rate is minimax optimal for errors in the conditional coverage probability, and the resulting errors in the marginal coverage probability also vanish at the same logarithmic rate. Despite these slow asymptotic rates, simulations show that the NP-EBCIs remain close to nominal coverage when the prior is non-Gaussian, and deliver substantial length reductions relative to intervals that treat each unit in isolation.
We provide an analytical characterization of the model flexibility of the synthetic control method (SCM) in the familiar form of degrees of freedom. We obtain estimable information criteria, which may be used to circumvent cross-validation when selecting either the tuning parameter in penalized variants of SCM or the weighting matrix in the SCM with covariates. We assess the impact of car license rationing in Tianjin; while a natural match is available, both it and other donors are noisy, inviting the use of SCM to average over approximately matching donors. The very large number of candidate donors calls for penalized variants of SCM and we observe that model selection using information criteria outperforms that based on cross-validation.