working papers

"Linearized GMM Estimator"  

(This paper will be presented during the poster session of the upcoming AEA meeting in 2025.)

     Abstract: It is well-known that nonlinear generalized method of moments (GMM) estimators often encounter computational issues when the moment conditions are over-identifying and nonlinear. To enhance computational properties, I propose a novel GMM estimator based on linearized moment conditions approximated around an underlying exactly-identified (or over-identified) parameter estimate. This estimator exhibits improved computational properties while maintaining first-order asymptotic efficiency. The enhancement arises from (i) the better-behaved curvature of the GMM objective function (e.g. strict local convexity) for estimating the underlying parameter, and (ii) the availability of a closed-form solution for the final estimate. For any given standard moment condition, I prove the existence of such an underlying parameter, and introduce a straightforward algorithm for its identification. The added dimensions in the underlying exactly-identified parameter can be estimated one element at a time, separately. The proposed method has been applied to Ahn, Lee, and Schmidt's (2013) panel data model with multiple time-varying individual effects.

   PDF file:  (download)

        STATA ado file: (coming soon)

"Short Panel Data Quantile Regression Model with Flexible Correlated Effects"

(revised and resubmitted, Econometric Reviews)

     Abstract: I propose an alternative linear model for short panel data quantile regression. The model assumes a nonparametric correlated effect (CE) that is τ-quantile-specific and time-invariant. The resulting partially linear model is robust to misspecification, and it is characterized as a best linear approximation to the truth under a generalized correlated random effect assumption. At the cost of modeling the individual heterogeneity, the new model is free of incidental parameters, and it does not restrict within-group dependence of idiosyncratic errors at all. The modeled heterogeneity is still well-aligned with the fixed effect approach in the linear mean regression model. For estimation, sieve-approximated CE is regularized by non-convex penalization which enjoys the oracle property against ultra-high dimensionality. Unpenalized sieve estimation is also available. As an empirical application, the proposed method is applied to estimate the distributional effect of smoking on birth weights.

     PDF file: (download)

     STATA ado file: (download)

Publication

"Difference-in-differences Estimator of Quantile Treatment Effect on the Treated" 

(with Jeffrey M. Wooldridge; Journal of Business and Economic Statistics, forthcoming) DOI 

"Asymptotic Efficiency of Joint Estimator relative to Two-stage Estimator under Misspecified Likelihoods"

(Studies in Nonlinear Dynamics and Econometrics, 2024) DOI

"An Alternative Two-step Generalized Method of Moments Estimation based on a Reduced Form Model"

(Economics Letters, Volume 192, 2020, Article 109184) DOI