Estimation of nonseparable models with censored dependent variables and endogenous regressors

Luke Taylor, Taisuke Otsu

Research output: Contribution to journalArticle

Abstract

In this article we develop a nonparametric estimator for the local average response of a censored dependent variable to endogenous regressors in a nonseparable model where the unobservable error term is not restricted to be scalar and where the nonseparable function need not be monotone in the unobservables. We formalize the identification argument put forward in Altonji, Ichimura, and Otsu (2012), construct a nonparametric estimator, characterize its asymptotic property, and conduct a Monte Carlo investigation to study its small sample properties. Identification is constructive and is achieved through a control function approach. We show that the estimator is consistent and asymptotically normally distributed. The Monte Carlo results are encouraging.

Original languageEnglish
Pages (from-to)4-24
Number of pages21
JournalEconometric Reviews
Volume38
Issue number1
DOIs
Publication statusPublished - 2019 Jan 2
Externally publishedYes

Fingerprint

Endogenous regressors
Estimator
Nonseparable models
Control function
Asymptotic properties
Small sample properties

Keywords

  • Average derivatives
  • censored dependent variables
  • endogeneity
  • nonparametric estimation
  • nonseparable models

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Estimation of nonseparable models with censored dependent variables and endogenous regressors. / Taylor, Luke; Otsu, Taisuke.

In: Econometric Reviews, Vol. 38, No. 1, 02.01.2019, p. 4-24.

Research output: Contribution to journalArticle

@article{0cdc286347824a21926dd24eca433845,
title = "Estimation of nonseparable models with censored dependent variables and endogenous regressors",
abstract = "In this article we develop a nonparametric estimator for the local average response of a censored dependent variable to endogenous regressors in a nonseparable model where the unobservable error term is not restricted to be scalar and where the nonseparable function need not be monotone in the unobservables. We formalize the identification argument put forward in Altonji, Ichimura, and Otsu (2012), construct a nonparametric estimator, characterize its asymptotic property, and conduct a Monte Carlo investigation to study its small sample properties. Identification is constructive and is achieved through a control function approach. We show that the estimator is consistent and asymptotically normally distributed. The Monte Carlo results are encouraging.",
keywords = "Average derivatives, censored dependent variables, endogeneity, nonparametric estimation, nonseparable models",
author = "Luke Taylor and Taisuke Otsu",
year = "2019",
month = "1",
day = "2",
doi = "10.1080/07474938.2016.1235310",
language = "English",
volume = "38",
pages = "4--24",
journal = "Econometric Reviews",
issn = "0747-4938",
publisher = "Taylor and Francis Ltd.",
number = "1",

}

TY - JOUR

T1 - Estimation of nonseparable models with censored dependent variables and endogenous regressors

AU - Taylor, Luke

AU - Otsu, Taisuke

PY - 2019/1/2

Y1 - 2019/1/2

N2 - In this article we develop a nonparametric estimator for the local average response of a censored dependent variable to endogenous regressors in a nonseparable model where the unobservable error term is not restricted to be scalar and where the nonseparable function need not be monotone in the unobservables. We formalize the identification argument put forward in Altonji, Ichimura, and Otsu (2012), construct a nonparametric estimator, characterize its asymptotic property, and conduct a Monte Carlo investigation to study its small sample properties. Identification is constructive and is achieved through a control function approach. We show that the estimator is consistent and asymptotically normally distributed. The Monte Carlo results are encouraging.

AB - In this article we develop a nonparametric estimator for the local average response of a censored dependent variable to endogenous regressors in a nonseparable model where the unobservable error term is not restricted to be scalar and where the nonseparable function need not be monotone in the unobservables. We formalize the identification argument put forward in Altonji, Ichimura, and Otsu (2012), construct a nonparametric estimator, characterize its asymptotic property, and conduct a Monte Carlo investigation to study its small sample properties. Identification is constructive and is achieved through a control function approach. We show that the estimator is consistent and asymptotically normally distributed. The Monte Carlo results are encouraging.

KW - Average derivatives

KW - censored dependent variables

KW - endogeneity

KW - nonparametric estimation

KW - nonseparable models

UR - http://www.scopus.com/inward/record.url?scp=84997638269&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84997638269&partnerID=8YFLogxK

U2 - 10.1080/07474938.2016.1235310

DO - 10.1080/07474938.2016.1235310

M3 - Article

VL - 38

SP - 4

EP - 24

JO - Econometric Reviews

JF - Econometric Reviews

SN - 0747-4938

IS - 1

ER -