Research Statement

My research is in the fields of econometrics and applied econometrics, with emphasis on microeconometrics. I am interested in developing econometric methods that have a broad range of applications and can lead to collaborations with colleagues conducting empirical studies in various fields of applied economics. I currently have three themes in my research agenda. The first is identification, estimation, and inference of semiparametric discrete choice models in both cross-sectional and panel data settings. The second theme is semiparametric estimation and inference with generated regressors. The last theme is the analysis of M-estimators for high-dimensional models in which the number of parameters is comparable to or larger than the sample size.

[Download full PDF statement]

Job Market Paper

Semiparametric Estimation of Multinomial Choice Models with Rank-Order Property

[Download the latest version of my JMP]

Abstract: In this paper, I propose a new semiparametric identification and estimation approach to multinomial choice models using cross-sectional data. The approach relies on the rank-order property proposed by Manski (1975) and employed by recent studies such as Fox (2007) and Yan (2013), which is a distribution-free restriction on the random utility framework underlying a multinomial choice model. From the rank-order property, a novel reparameterization provides a multivariate nonlinear least squares (population) criterion identifying the structural parameters. This identification result then motivates a sieve-based estimation procedure, which is the first in the semiparametric literature to allow joint estimation of regression coefficients and reduced-form parameters such as choice probabilities and marginal effects. Asymptotic properties of two functional estimators are developed. A Monte Carlo study indicates that these functional estimators perform well in finite samples. I illustrate the implementation of the estimation procedure via estimating a model of college major choice using UCOP data of 1998-2003. As extensions, I also propose estimators for the model using a choice-based sample and the model with ranking information.

Working Papers

Adaptive Rank Inference in Semiparametric Multinomial Response Models (with Shakeeb Khan and Elie Tamer) [draft coming soon]

Abstract: We consider identification, estimation, and inference on regression coefficients in semiparametric multinomial response models. Our identification result is constructive, and estimation is based on a localized rank objective function, loosely analogous to that used in Abrevaya, Hausman, and Khan (2010). We show this achieves sharp identification which is in contrast to existing procedures in the literature such as Ahn, Powell, Ichimura, and Ruud (2014). Our procedure is adaptive (Khan and Tamer (2009)) in the sense that it provides an estimator of the sharp set when point identification does not hold, and a consistent point estimator when it does. Furthermore, our rank procedure extends to panel data settings for inference in models with fixed effects, including dynamic panel models with lagged dependent variables as covariates. A simulation study establishes adequate finite sample properties of our new procedures.

Semiparametric Estimation of Panel Data Binary Choice Models with Lagged Dependent Variables [draft coming soon]

Abstract: This paper proposes a new approach to the semiparametric analysis of panel binary choice models with fixed effects and lagged dependent variables. The same random utility framework as in Honoré andKyriazidou (2000) is employed with two distinctions: (1) at least five observations per individual are observed by econometrician; (2) the joint distribution of explanatory variables is time stationary. With these extra restrictions, the identification of the model proceeds in two steps and only requires matching the value of an index function of explanatory variables over time, as opposed to that of each explanatory variable. The identification approach motivates an easy-to-implement two-stage estimator whose rate of convergence does not decrease as the number of explanatory variables increases, which is in contrast to the Honoré and Kyriazidou’s (2000) estimator. Asymptotic properties are established. Monte Carlo evidence indicates that the proposed estimator performs well in finite samples.

Local Nonlinear Least Squares Estimator for Multinomial Choice Models with Choice-Based Samples and Nonparametrically Generated Regressors [draft coming soon]

Abstract: This paper proposes a semiparametric estimator of preference parameters in multinomial choice models having the following characteristics: (1) only choice-based samples can be used by econometrician; (2) there exist one or more utility determinants being uncertain at an agent’s choice-making stage, and so her decision rule depends on conditional expectations of these determinants. The estimator does not impose parametric restrictions on the distribution of the error terms and allows general forms of (conditional) heteroskedasticity. The estimation is easy to implement and proceeds in two stages: the uncertain utility determinants are nonparametrically estimated in the first stage, and then are used as plug-in terms in a local nonlinear least squares (LNLS) estimation of preference parameters in the second stage. Asymptotic properties are developed. A Monte Carlo study indicates that the proposed estimator performs well in finite samples.

Work in Progress

U-Processes with Nonparametrically Generated Covariates (with Yichong Zhang)

Confidence Intervals for High-dimensional Probit Models

Maximum Rank Correlation Estimation with Uncertainty: A Study of Strategic Drilling Behavior (with Ashley Vissing)

Abstract: Over the last twenty years, the oil and gas industry has increased production of resources by developing and implementing unconventional extraction techniques. We study how firms’ profit functions vary across conventionally and unconventionally drilled plays located in the Barnett Shale region of Texas. We find evidence that, as technology evolved, firms’ spatial strategic decisions change whereby they are less likely to competitively drill wells and choose locations spaced farther away from their competitors’ existing wells. We then estimate a structural model under uncertainty in which firms’ decision-making over portfolios of permitted wells is affected by the conditional expectation of their future revenue. Our empirical model captures the differential effects of complementary well locations before and after the shale boom while allowing for correlated unobservables across the wells in the firms’ portfolios.