Ruoyao Shi | University of California, Riverside (original) (raw)

Papers by Ruoyao Shi

Research paper thumbnail of Uniform Asymptotic Risk of Averaging GMM Estimator Robust to Misspeci…cation

Social Science Research Network, 2015

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. It is an alternative to pre-test estimators that switch between the conservative and aggressive estimators based on model speci…cation tests. This averaging estimator is robust in the sense that it uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation, whereas the pre-test estimators reduce the risk in parts of the parameter space and increase it in other parts. To establish uniform dominance of one estimator over another, we establish asymptotic theories on uniform approximations of the …nite-sample risk di¤erences between two estimators. These asymptotic results are developed along drifting sequences of data generating processes (DGPs) that model various degrees of local misspeci…cation as well as global misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model.

Research paper thumbnail of Essays on Econometrics - eScholarship

Research paper thumbnail of Uniform Asymptotic Risk of Averaging GMM Estimator Robust to Misspecification, Second Version

RePEc: Research Papers in Economics, Mar 25, 2015

This paper studies the averaging generalized method of moments (GMM) estimator that combines a co... more This paper studies the averaging generalized method of moments (GMM) estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. It is an alternative to pre-test estimators that switch between the conservative and agressive estimators based on model speci…cation tests. This averaging estimator is robust in the sense that it uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation, whereas the pre-test estimators reduce the risk in parts of the parameter space and increase it in other parts. To establish uniform dominance of one estimator over another, we establish asymptotic theories on uniform approximations of the …nite-sample risk di¤erences between two estimators. These asymptotic results are developed along drifting sequences of data generating processes (DGPs) that model various degrees of local misspeci…cation as well as global misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model.

Research paper thumbnail of Identification and Estimation of Nonparametric Hedonic Equilibrium Model with Unobserved Quality

RePEc: Research Papers in Economics, 2018

Research paper thumbnail of Synthetic Control and Inference

RePEc: Research Papers in Economics, 2016

We examine properties of permutation tests in the context of synthetic control. Permutation tests... more We examine properties of permutation tests in the context of synthetic control. Permutation tests are frequently used methods of inference for synthetic control when the number of potential control units is small. We analyze the permutation tests from a repeated sampling perspective and show that the size of permutation tests may be distorted. Several alternative methods are discussed.

Research paper thumbnail of A Note on the GRS Test

SSRN Electronic Journal, 2020

The Gibbons, Ross, and Shanken (1989) F-test of mean-variance efficiency of asset returns is stat... more The Gibbons, Ross, and Shanken (1989) F-test of mean-variance efficiency of asset returns is stated incorrectly for the multi-factor case. We first derive the correct formula for the test statistic for the general case of K factors and N test assets, then highlight the impact of the error in common applications. The ranking of competing models can be scrambled if the original (incorrect) formula is used, and tests of factor models over-reject. While the impact is material only for horizons of less than 20 or so years of monthly data, given the theoretical interpretation of the (correctly) calculated GRS statistic, we recommend that researchers use the correct formula regardless of sample size.

Research paper thumbnail of Utilizing Two Types of Survey Data to Enhance the Accuracy of Labor Supply Elasticity Estimation

We argue that despite its nonclassical measurement errors, the hours worked in the Current Popula... more We argue that despite its nonclassical measurement errors, the hours worked in the Current Population Survey (CPS) can still be utilized to enhance the overall accuracy of the estimator of the labor supply parameters based on the American Time Use Survey (ATUS), if done properly. We propose such an estimator that is a weighted average between the two stage least squares estimator based on the CPS and a non-standard estimator based on the ATUS.

Research paper thumbnail of Supplemental Appendix of “ An Averaging GMM Estimator Robust to Misspecification ”

In this supplemental appendix, we present supporting materials for Cheng, Liao and Shi (2018) (ci... more In this supplemental appendix, we present supporting materials for Cheng, Liao and Shi (2018) (cited as CLS hereafter in this Appendix): • Section D provides primitive conditions for Assumptions 3.1, 3.2 and 3.3 and the proof of Lemma 3.1 of CLS. • Section E provides the proof of (4.3) in Section 4 and the proof of some Lemmas in Appendix B.1 of CLS. The proof of Lemma A.1 in Appendix A of CLS is also included in this section. • Section F studies the bounds of asymptotic risk difference of the pre-test GMM estimator. • Section G contains simulation results under the truncated risk for the simulation designs in Section 6 of CLS. • Section H includes extra simulation studies.

Research paper thumbnail of Essays on Econometrics

Author(s): Shi, Ruoyao | Advisor(s): Hahn, Jinyong; Liao, Zhipeng | Abstract: This dissertation s... more Author(s): Shi, Ruoyao | Advisor(s): Hahn, Jinyong; Liao, Zhipeng | Abstract: This dissertation studies econometric questions in the context of three different methods that are frequently used by empirical economists.Chapter 1 provides a short introduction to the contexts, questions, methods and results studied in Chapter 2 to Chapter 4.Chapter 2 studies a nonparametric hedonic equilibrium model in which certain product characteristics are unobserved. Unlike most previously studied hedonic models, both the observed and unobserved agent heterogeneities enter the structural functions nonparametrically. Prices are endogenously determined in equilibrium. Using both within-market and cross-market price variation, I show that all the structural functions of the model are nonparametrically identified up to normalization. In particular, the unobserved product quality function is identified if the relative prices of the agent characteristics differ in at least two markets. Following the cons...

Research paper thumbnail of Identification and Estimation of Nonparametric Hedonic Equilibrium Model with Unobserved Quality

This paper studies a nonparametric hedonic equilibrium model in which certain product characteris... more This paper studies a nonparametric hedonic equilibrium model in which certain product characteristics are unobserved. Unlike most previously studied hedonic models, both the observed and unobserved agent heterogeneities enter the structural functions nonparametrically. Prices are endogenously determined in equilibrium. Using both within- and cross-market price variation, I show that all the structural functions of the model are nonparametrically identified up to normalization. In particular, the unobserved product quality function is identified if the relative prices of the agent characteristics differ in at least two markets. Following the constructive identification strategy, I provide easy- to-implement series minimum distance estimators of the structural functions and derive their consistency and uniform rates of convergence. To illustrate the estimation procedure, I estimate the unobserved efficiency of American full-time workers as a function of age and unobserved ability.

Research paper thumbnail of Averaging GMM Estimator Robust to Misspeci cation

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the …nite-sample risk di¤erence between two estimators, which is used to show that the averaging estimator uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The simulation results support our theoretical …ndings. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model. Keywords: Asymptotic Risk, Finite-Sample Risk, Generalized Shrinkage Estimator...

Research paper thumbnail of Testing and Ranking of Asset Pricing Models Using the Grs Statistic

Research paper thumbnail of The Influence Function of Semiparametric Two-step Estimators with Estimated Control Variables

This paper studies semiparametric two-step estimators with a control variable estimated in a firs... more This paper studies semiparametric two-step estimators with a control variable estimated in a first-step parametric or nonparametric model. We provide the explicit influence function of the two-step estimator under an index restriction which is imposed directly on the unknown control variable. The index restriction is weaker than the commonly used identification conditions in the literature, which are imposed on all exogenous variables. An extra term shows up in the influence function of the semiparametric two-step estimator under the weaker identification condition. We illustrate our influence function formula in a mean regression example, a quantile regression example, and a sample selection example where the control variable approach is applied for identification and consistent estimation of structural parameters. JEL Classification: C14, C31, C32

Research paper thumbnail of On uniform asymptotic risk of averaging GMM estimators

Quantitative Economics, 2019

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspecified moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the finite‐sample truncated risk difference between any two estimators, which is used to compare the averaging GMM estimator and the conservative GMM estimator. Under some sufficient conditions, we show that the asymptotic lower bound of the truncated risk difference between the averaging estimator and the conservative estimator is strictly less than zero, while the asymptotic upper bound is zero uniformly over any degree of misspecification. The results apply to quadratic loss functions. This uniform asymptotic dominance is established in non‐Gaussian semiparametric nonlinear models.

Research paper thumbnail of Synthetic Control and Inference

Econometrics, 2017

We examine properties of permutation tests in the context of synthetic control. Permutation tests... more We examine properties of permutation tests in the context of synthetic control. Permutation tests are frequently used methods of inference for synthetic control when the number of potential control units is small. We analyze the permutation tests from a repeated sampling perspective and show that the size of permutation tests may be distorted. Several alternative methods are discussed.

Research paper thumbnail of An Averaging Estimator for Two-Step M-Estimation in Semiparametric Models

Econometric Theory, 2021

In a two step extremum estimation (M estimation) framework with a finite dimensional parameter of... more In a two step extremum estimation (M estimation) framework with a finite dimensional parameter of interest and a potentially infinite dimensional first step nuisance parameter, I propose an averaging estimator that combines a semiparametric estimator based on nonparametric first step and a parametric estimator which imposes parametric restrictions on the first step. The averaging weight is the sample analog of an infeasible optimal weight that minimizes the asymptotic quadratic risk. I show that under mild conditions, the asymptotic lower bound of the truncated quadratic risk difference between the averaging estimator and the semiparametric estimator is strictly less than zero for a class of data generating processes (DGPs) that includes both correct specification and varied degrees of misspecification of the parametric restrictions, and the asymptotic upper bound is weakly less than zero.

Research paper thumbnail of What time use surveys can (and cannot) tell us about labor supply

Journal of Applied Econometrics, 2021

It has been widely acknowledged that the measurement of labor supply in the Current Population Su... more It has been widely acknowledged that the measurement of labor supply in the Current Population Survey (CPS) and other conventional microeconomic surveys has nonclassical measurement error, which will bias the estimates of crucial parameters in labor economics, such as labor supply elasticity. Time diary studies, such as the American Time Use Survey (ATUS), only have accurate measurement of hours worked on a single day, hence the weekly hours worked are unobserved. Despite the missing data problem, we provide several consistent estimators of the parameters in weekly labor supply equation using the information in the time use surveys. The consistency of our estimators does not require more conditions beyond those for a usual two stage least square (2SLS) estimator when the true weekly hours worked are observed. We also show that it is impossible to recover the weekly number of hours worked or its distribution function from time use surveys like the ATUS. In our empirical application we find considerable evidence of nonclassical measurement error in the hours worked in the CPS, and illustrate the consequences of using mismeasured weekly hours worked in empirical studies.

Research paper thumbnail of Constructing Counterfactual Wage Distribution Using A General Equilibrium Labor Market Model with Heterogeneity and Unobservable Efficiency

Welfare analysis of wage inequality requires constructing counterfactual wage distributions. I pr... more Welfare analysis of wage inequality requires constructing counterfactual wage distributions. I propose a method based on a fully nonparametric general equilibrium labor market model in which heterogenous workers and firms trade effective labor. Effective labor depends on two factors, observable hours and unobservable efficiency. Contrary to previous partial equilibrium approaches, counterfactual interventions in my model affect the behaviors of both workers and firms, and hence the market equilibrium. I show nonparametric identification of the structural functions of the model, in particular the unobservable efficiency function. The identified structural functions are used to generate counterfactual wage samples through a simulation method I prescribe. As a preliminary step towards analyzing identification and estimation of counterfactual wages, I introduce the operators that map the structural functions to parameters of interests. My model works under a wide range of counterfactual...

Research paper thumbnail of What Time Use Surveys Can (And Cannot) Tell Us About Labor Supply

Nonclassical measurement errors in conventional microeconomic surveys result in biased estimate... more Nonclassical measurement errors in conventional microeconomic surveys result in biased estimates of weekly labor supply parameters. The American Time Use Survey (ATUS) accurately measures hours worked on a single day. We show that despite the impossibility to recover weekly hours, weekly labor supply parameters can be consistently and efficiently estimated using the ATUS. We propose impute estimator and carefully examine its properties. It is a simple modification of the 2SLS estimator, which imputes both dependent and independent variables using daily subsamples. We apply it to the ATUS and find substantially different elasticity estimates from the CPS, especially for married women.

Research paper thumbnail of On uniform asymptotic risk of averaging GMM estimators

Quantitative Economics, 2019

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspecified moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the finite‐sample truncated risk difference between any two estimators, which is used to compare the averaging GMM estimator and the conservative GMM estimator. Under some sufficient conditions, we show that the asymptotic lower bound of the truncated risk difference between the averaging estimator and the conservative estimator is strictly less than zero, while the asymptotic upper bound is zero uniformly over any degree of misspecification. The results apply to quadratic loss functions. This uniform asymptotic dominance is established in non‐Gaussian semiparametric nonlinear models.

Research paper thumbnail of Uniform Asymptotic Risk of Averaging GMM Estimator Robust to Misspeci…cation

Social Science Research Network, 2015

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. It is an alternative to pre-test estimators that switch between the conservative and aggressive estimators based on model speci…cation tests. This averaging estimator is robust in the sense that it uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation, whereas the pre-test estimators reduce the risk in parts of the parameter space and increase it in other parts. To establish uniform dominance of one estimator over another, we establish asymptotic theories on uniform approximations of the …nite-sample risk di¤erences between two estimators. These asymptotic results are developed along drifting sequences of data generating processes (DGPs) that model various degrees of local misspeci…cation as well as global misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model.

Research paper thumbnail of Essays on Econometrics - eScholarship

Research paper thumbnail of Uniform Asymptotic Risk of Averaging GMM Estimator Robust to Misspecification, Second Version

RePEc: Research Papers in Economics, Mar 25, 2015

This paper studies the averaging generalized method of moments (GMM) estimator that combines a co... more This paper studies the averaging generalized method of moments (GMM) estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. It is an alternative to pre-test estimators that switch between the conservative and agressive estimators based on model speci…cation tests. This averaging estimator is robust in the sense that it uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation, whereas the pre-test estimators reduce the risk in parts of the parameter space and increase it in other parts. To establish uniform dominance of one estimator over another, we establish asymptotic theories on uniform approximations of the …nite-sample risk di¤erences between two estimators. These asymptotic results are developed along drifting sequences of data generating processes (DGPs) that model various degrees of local misspeci…cation as well as global misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model.

Research paper thumbnail of Identification and Estimation of Nonparametric Hedonic Equilibrium Model with Unobserved Quality

RePEc: Research Papers in Economics, 2018

Research paper thumbnail of Synthetic Control and Inference

RePEc: Research Papers in Economics, 2016

We examine properties of permutation tests in the context of synthetic control. Permutation tests... more We examine properties of permutation tests in the context of synthetic control. Permutation tests are frequently used methods of inference for synthetic control when the number of potential control units is small. We analyze the permutation tests from a repeated sampling perspective and show that the size of permutation tests may be distorted. Several alternative methods are discussed.

Research paper thumbnail of A Note on the GRS Test

SSRN Electronic Journal, 2020

The Gibbons, Ross, and Shanken (1989) F-test of mean-variance efficiency of asset returns is stat... more The Gibbons, Ross, and Shanken (1989) F-test of mean-variance efficiency of asset returns is stated incorrectly for the multi-factor case. We first derive the correct formula for the test statistic for the general case of K factors and N test assets, then highlight the impact of the error in common applications. The ranking of competing models can be scrambled if the original (incorrect) formula is used, and tests of factor models over-reject. While the impact is material only for horizons of less than 20 or so years of monthly data, given the theoretical interpretation of the (correctly) calculated GRS statistic, we recommend that researchers use the correct formula regardless of sample size.

Research paper thumbnail of Utilizing Two Types of Survey Data to Enhance the Accuracy of Labor Supply Elasticity Estimation

We argue that despite its nonclassical measurement errors, the hours worked in the Current Popula... more We argue that despite its nonclassical measurement errors, the hours worked in the Current Population Survey (CPS) can still be utilized to enhance the overall accuracy of the estimator of the labor supply parameters based on the American Time Use Survey (ATUS), if done properly. We propose such an estimator that is a weighted average between the two stage least squares estimator based on the CPS and a non-standard estimator based on the ATUS.

Research paper thumbnail of Supplemental Appendix of “ An Averaging GMM Estimator Robust to Misspecification ”

In this supplemental appendix, we present supporting materials for Cheng, Liao and Shi (2018) (ci... more In this supplemental appendix, we present supporting materials for Cheng, Liao and Shi (2018) (cited as CLS hereafter in this Appendix): • Section D provides primitive conditions for Assumptions 3.1, 3.2 and 3.3 and the proof of Lemma 3.1 of CLS. • Section E provides the proof of (4.3) in Section 4 and the proof of some Lemmas in Appendix B.1 of CLS. The proof of Lemma A.1 in Appendix A of CLS is also included in this section. • Section F studies the bounds of asymptotic risk difference of the pre-test GMM estimator. • Section G contains simulation results under the truncated risk for the simulation designs in Section 6 of CLS. • Section H includes extra simulation studies.

Research paper thumbnail of Essays on Econometrics

Author(s): Shi, Ruoyao | Advisor(s): Hahn, Jinyong; Liao, Zhipeng | Abstract: This dissertation s... more Author(s): Shi, Ruoyao | Advisor(s): Hahn, Jinyong; Liao, Zhipeng | Abstract: This dissertation studies econometric questions in the context of three different methods that are frequently used by empirical economists.Chapter 1 provides a short introduction to the contexts, questions, methods and results studied in Chapter 2 to Chapter 4.Chapter 2 studies a nonparametric hedonic equilibrium model in which certain product characteristics are unobserved. Unlike most previously studied hedonic models, both the observed and unobserved agent heterogeneities enter the structural functions nonparametrically. Prices are endogenously determined in equilibrium. Using both within-market and cross-market price variation, I show that all the structural functions of the model are nonparametrically identified up to normalization. In particular, the unobserved product quality function is identified if the relative prices of the agent characteristics differ in at least two markets. Following the cons...

Research paper thumbnail of Identification and Estimation of Nonparametric Hedonic Equilibrium Model with Unobserved Quality

This paper studies a nonparametric hedonic equilibrium model in which certain product characteris... more This paper studies a nonparametric hedonic equilibrium model in which certain product characteristics are unobserved. Unlike most previously studied hedonic models, both the observed and unobserved agent heterogeneities enter the structural functions nonparametrically. Prices are endogenously determined in equilibrium. Using both within- and cross-market price variation, I show that all the structural functions of the model are nonparametrically identified up to normalization. In particular, the unobserved product quality function is identified if the relative prices of the agent characteristics differ in at least two markets. Following the constructive identification strategy, I provide easy- to-implement series minimum distance estimators of the structural functions and derive their consistency and uniform rates of convergence. To illustrate the estimation procedure, I estimate the unobserved efficiency of American full-time workers as a function of age and unobserved ability.

Research paper thumbnail of Averaging GMM Estimator Robust to Misspeci cation

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspeci…ed moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the …nite-sample risk di¤erence between two estimators, which is used to show that the averaging estimator uniformly dominates the conservative estimator by reducing the risk under any degree of misspeci…cation. Extending seminal results on the James-Stein estimator, the uniform dominance is established in non-Gaussian semiparametric nonlinear models. The simulation results support our theoretical …ndings. The proposed averaging estimator is applied to estimate the human capital production function in a life-cycle labor supply model. Keywords: Asymptotic Risk, Finite-Sample Risk, Generalized Shrinkage Estimator...

Research paper thumbnail of Testing and Ranking of Asset Pricing Models Using the Grs Statistic

Research paper thumbnail of The Influence Function of Semiparametric Two-step Estimators with Estimated Control Variables

This paper studies semiparametric two-step estimators with a control variable estimated in a firs... more This paper studies semiparametric two-step estimators with a control variable estimated in a first-step parametric or nonparametric model. We provide the explicit influence function of the two-step estimator under an index restriction which is imposed directly on the unknown control variable. The index restriction is weaker than the commonly used identification conditions in the literature, which are imposed on all exogenous variables. An extra term shows up in the influence function of the semiparametric two-step estimator under the weaker identification condition. We illustrate our influence function formula in a mean regression example, a quantile regression example, and a sample selection example where the control variable approach is applied for identification and consistent estimation of structural parameters. JEL Classification: C14, C31, C32

Research paper thumbnail of On uniform asymptotic risk of averaging GMM estimators

Quantitative Economics, 2019

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspecified moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the finite‐sample truncated risk difference between any two estimators, which is used to compare the averaging GMM estimator and the conservative GMM estimator. Under some sufficient conditions, we show that the asymptotic lower bound of the truncated risk difference between the averaging estimator and the conservative estimator is strictly less than zero, while the asymptotic upper bound is zero uniformly over any degree of misspecification. The results apply to quadratic loss functions. This uniform asymptotic dominance is established in non‐Gaussian semiparametric nonlinear models.

Research paper thumbnail of Synthetic Control and Inference

Econometrics, 2017

We examine properties of permutation tests in the context of synthetic control. Permutation tests... more We examine properties of permutation tests in the context of synthetic control. Permutation tests are frequently used methods of inference for synthetic control when the number of potential control units is small. We analyze the permutation tests from a repeated sampling perspective and show that the size of permutation tests may be distorted. Several alternative methods are discussed.

Research paper thumbnail of An Averaging Estimator for Two-Step M-Estimation in Semiparametric Models

Econometric Theory, 2021

In a two step extremum estimation (M estimation) framework with a finite dimensional parameter of... more In a two step extremum estimation (M estimation) framework with a finite dimensional parameter of interest and a potentially infinite dimensional first step nuisance parameter, I propose an averaging estimator that combines a semiparametric estimator based on nonparametric first step and a parametric estimator which imposes parametric restrictions on the first step. The averaging weight is the sample analog of an infeasible optimal weight that minimizes the asymptotic quadratic risk. I show that under mild conditions, the asymptotic lower bound of the truncated quadratic risk difference between the averaging estimator and the semiparametric estimator is strictly less than zero for a class of data generating processes (DGPs) that includes both correct specification and varied degrees of misspecification of the parametric restrictions, and the asymptotic upper bound is weakly less than zero.

Research paper thumbnail of What time use surveys can (and cannot) tell us about labor supply

Journal of Applied Econometrics, 2021

It has been widely acknowledged that the measurement of labor supply in the Current Population Su... more It has been widely acknowledged that the measurement of labor supply in the Current Population Survey (CPS) and other conventional microeconomic surveys has nonclassical measurement error, which will bias the estimates of crucial parameters in labor economics, such as labor supply elasticity. Time diary studies, such as the American Time Use Survey (ATUS), only have accurate measurement of hours worked on a single day, hence the weekly hours worked are unobserved. Despite the missing data problem, we provide several consistent estimators of the parameters in weekly labor supply equation using the information in the time use surveys. The consistency of our estimators does not require more conditions beyond those for a usual two stage least square (2SLS) estimator when the true weekly hours worked are observed. We also show that it is impossible to recover the weekly number of hours worked or its distribution function from time use surveys like the ATUS. In our empirical application we find considerable evidence of nonclassical measurement error in the hours worked in the CPS, and illustrate the consequences of using mismeasured weekly hours worked in empirical studies.

Research paper thumbnail of Constructing Counterfactual Wage Distribution Using A General Equilibrium Labor Market Model with Heterogeneity and Unobservable Efficiency

Welfare analysis of wage inequality requires constructing counterfactual wage distributions. I pr... more Welfare analysis of wage inequality requires constructing counterfactual wage distributions. I propose a method based on a fully nonparametric general equilibrium labor market model in which heterogenous workers and firms trade effective labor. Effective labor depends on two factors, observable hours and unobservable efficiency. Contrary to previous partial equilibrium approaches, counterfactual interventions in my model affect the behaviors of both workers and firms, and hence the market equilibrium. I show nonparametric identification of the structural functions of the model, in particular the unobservable efficiency function. The identified structural functions are used to generate counterfactual wage samples through a simulation method I prescribe. As a preliminary step towards analyzing identification and estimation of counterfactual wages, I introduce the operators that map the structural functions to parameters of interests. My model works under a wide range of counterfactual...

Research paper thumbnail of What Time Use Surveys Can (And Cannot) Tell Us About Labor Supply

Nonclassical measurement errors in conventional microeconomic surveys result in biased estimate... more Nonclassical measurement errors in conventional microeconomic surveys result in biased estimates of weekly labor supply parameters. The American Time Use Survey (ATUS) accurately measures hours worked on a single day. We show that despite the impossibility to recover weekly hours, weekly labor supply parameters can be consistently and efficiently estimated using the ATUS. We propose impute estimator and carefully examine its properties. It is a simple modification of the 2SLS estimator, which imputes both dependent and independent variables using daily subsamples. We apply it to the ATUS and find substantially different elasticity estimates from the CPS, especially for married women.

Research paper thumbnail of On uniform asymptotic risk of averaging GMM estimators

Quantitative Economics, 2019

This paper studies the averaging GMM estimator that combines a conservative GMM estimator based o... more This paper studies the averaging GMM estimator that combines a conservative GMM estimator based on valid moment conditions and an aggressive GMM estimator based on both valid and possibly misspecified moment conditions, where the weight is the sample analog of an infeasible optimal weight. We establish asymptotic theory on uniform approximation of the upper and lower bounds of the finite‐sample truncated risk difference between any two estimators, which is used to compare the averaging GMM estimator and the conservative GMM estimator. Under some sufficient conditions, we show that the asymptotic lower bound of the truncated risk difference between the averaging estimator and the conservative estimator is strictly less than zero, while the asymptotic upper bound is zero uniformly over any degree of misspecification. The results apply to quadratic loss functions. This uniform asymptotic dominance is established in non‐Gaussian semiparametric nonlinear models.