Handling endogenous regressors by joint estimation using copulas

Sungho Park, Sachin Gupta

Research output: Contribution to journalArticle

76 Citations (Scopus)

Abstract

We propose a new statistical instrument-free method to tackle the endogeneity problem. The proposed method models the joint distribution of the endogenous regressor and the error term in the structural equation of interest (the structural error) using a copula method, and it makes inferences on the model parameters by maximizing the likelihood derived from the joint distribution. Similar to the "exclusion restriction" in instrumental variable methods, extant instrument-free methods require the assumption that the unobserved instruments are exogenous, a requirement that is difficult to meet. The proposed method does not require such an assumption. Other benefits of the proposed method are that it allows the modeling of discrete endogenous regressors and offers a new solution to the slope endogeneity problem. In addition to linear models, the method is applicable to the popular random coefficient logit model with either aggregate-level or individual-level data. We demonstrate the performance of the proposed method via a series of simulation studies and an empirical example.

Original languageEnglish (US)
Pages (from-to)567-586
Number of pages20
JournalMarketing Science
Volume31
Issue number4
DOIs
StatePublished - Jul 2012

Fingerprint

Endogenous regressors
Joint estimation
Copula
Joint distribution
Endogeneity
Simulation study
Structural equations
Instrumental variables
Inference
Exclusion
Modeling
Logit model
Random coefficients

Keywords

  • Copula method
  • Endogeneity
  • Instrumental variables
  • Linear regression model
  • Logit model
  • Random coefficient
  • Two-stage least squares

ASJC Scopus subject areas

  • Business and International Management
  • Economics and Econometrics
  • Marketing

Cite this

Handling endogenous regressors by joint estimation using copulas. / Park, Sungho; Gupta, Sachin.

In: Marketing Science, Vol. 31, No. 4, 07.2012, p. 567-586.

Research output: Contribution to journalArticle

Park, Sungho ; Gupta, Sachin. / Handling endogenous regressors by joint estimation using copulas. In: Marketing Science. 2012 ; Vol. 31, No. 4. pp. 567-586.
@article{ba6183c738ea4911b8b11a19e0c633ff,
title = "Handling endogenous regressors by joint estimation using copulas",
abstract = "We propose a new statistical instrument-free method to tackle the endogeneity problem. The proposed method models the joint distribution of the endogenous regressor and the error term in the structural equation of interest (the structural error) using a copula method, and it makes inferences on the model parameters by maximizing the likelihood derived from the joint distribution. Similar to the {"}exclusion restriction{"} in instrumental variable methods, extant instrument-free methods require the assumption that the unobserved instruments are exogenous, a requirement that is difficult to meet. The proposed method does not require such an assumption. Other benefits of the proposed method are that it allows the modeling of discrete endogenous regressors and offers a new solution to the slope endogeneity problem. In addition to linear models, the method is applicable to the popular random coefficient logit model with either aggregate-level or individual-level data. We demonstrate the performance of the proposed method via a series of simulation studies and an empirical example.",
keywords = "Copula method, Endogeneity, Instrumental variables, Linear regression model, Logit model, Random coefficient, Two-stage least squares",
author = "Sungho Park and Sachin Gupta",
year = "2012",
month = "7",
doi = "10.1287/mksc.1120.0718",
language = "English (US)",
volume = "31",
pages = "567--586",
journal = "Marketing Science",
issn = "0732-2399",
publisher = "INFORMS Inst.for Operations Res.and the Management Sciences",
number = "4",

}

TY - JOUR

T1 - Handling endogenous regressors by joint estimation using copulas

AU - Park, Sungho

AU - Gupta, Sachin

PY - 2012/7

Y1 - 2012/7

N2 - We propose a new statistical instrument-free method to tackle the endogeneity problem. The proposed method models the joint distribution of the endogenous regressor and the error term in the structural equation of interest (the structural error) using a copula method, and it makes inferences on the model parameters by maximizing the likelihood derived from the joint distribution. Similar to the "exclusion restriction" in instrumental variable methods, extant instrument-free methods require the assumption that the unobserved instruments are exogenous, a requirement that is difficult to meet. The proposed method does not require such an assumption. Other benefits of the proposed method are that it allows the modeling of discrete endogenous regressors and offers a new solution to the slope endogeneity problem. In addition to linear models, the method is applicable to the popular random coefficient logit model with either aggregate-level or individual-level data. We demonstrate the performance of the proposed method via a series of simulation studies and an empirical example.

AB - We propose a new statistical instrument-free method to tackle the endogeneity problem. The proposed method models the joint distribution of the endogenous regressor and the error term in the structural equation of interest (the structural error) using a copula method, and it makes inferences on the model parameters by maximizing the likelihood derived from the joint distribution. Similar to the "exclusion restriction" in instrumental variable methods, extant instrument-free methods require the assumption that the unobserved instruments are exogenous, a requirement that is difficult to meet. The proposed method does not require such an assumption. Other benefits of the proposed method are that it allows the modeling of discrete endogenous regressors and offers a new solution to the slope endogeneity problem. In addition to linear models, the method is applicable to the popular random coefficient logit model with either aggregate-level or individual-level data. We demonstrate the performance of the proposed method via a series of simulation studies and an empirical example.

KW - Copula method

KW - Endogeneity

KW - Instrumental variables

KW - Linear regression model

KW - Logit model

KW - Random coefficient

KW - Two-stage least squares

UR - http://www.scopus.com/inward/record.url?scp=84864974354&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864974354&partnerID=8YFLogxK

U2 - 10.1287/mksc.1120.0718

DO - 10.1287/mksc.1120.0718

M3 - Article

AN - SCOPUS:84864974354

VL - 31

SP - 567

EP - 586

JO - Marketing Science

JF - Marketing Science

SN - 0732-2399

IS - 4

ER -