Identification and inference on regressions with missing covariate data

Esteban Aucejo, Federico A. Bugni, V. Joseph Hotz

Research output: Contribution to journalArticle

Abstract

This paper examines the problem of identification and inference on a conditional moment condition model with missing data, with special focus on the case when the conditioning covariates are missing. We impose no assumption on the distribution of the missing data and we confront the missing data problem by using a worst case scenario approach. We characterize the sharp identified set and argue that this set is usually too complex to compute or to use for inference. Given this difficulty, we consider the construction of outer identified sets (i.e. supersets of the identified set) that are easier to compute and can still characterize the parameter of interest. Two different outer identification strategies are proposed. Both of these strategies are shown to have nontrivial identifying power and are relatively easy to use and combine for inferential purposes.

Original languageEnglish (US)
Pages (from-to)196-241
Number of pages46
JournalEconometric Theory
Volume33
Issue number1
DOIs
StatePublished - Feb 1 2017
Externally publishedYes

Fingerprint

regression
conditioning
scenario
Inference
Missing data
Covariates
Moment conditions
Scenarios
Conditional moments
Conditioning

ASJC Scopus subject areas

  • Social Sciences (miscellaneous)
  • Economics and Econometrics

Cite this

Identification and inference on regressions with missing covariate data. / Aucejo, Esteban; Bugni, Federico A.; Hotz, V. Joseph.

In: Econometric Theory, Vol. 33, No. 1, 01.02.2017, p. 196-241.

Research output: Contribution to journalArticle

Aucejo, Esteban ; Bugni, Federico A. ; Hotz, V. Joseph. / Identification and inference on regressions with missing covariate data. In: Econometric Theory. 2017 ; Vol. 33, No. 1. pp. 196-241.
@article{dd1f337a8855408b83eae2116a935cbd,
title = "Identification and inference on regressions with missing covariate data",
abstract = "This paper examines the problem of identification and inference on a conditional moment condition model with missing data, with special focus on the case when the conditioning covariates are missing. We impose no assumption on the distribution of the missing data and we confront the missing data problem by using a worst case scenario approach. We characterize the sharp identified set and argue that this set is usually too complex to compute or to use for inference. Given this difficulty, we consider the construction of outer identified sets (i.e. supersets of the identified set) that are easier to compute and can still characterize the parameter of interest. Two different outer identification strategies are proposed. Both of these strategies are shown to have nontrivial identifying power and are relatively easy to use and combine for inferential purposes.",
author = "Esteban Aucejo and Bugni, {Federico A.} and Hotz, {V. Joseph}",
year = "2017",
month = "2",
day = "1",
doi = "10.1017/S0266466615000250",
language = "English (US)",
volume = "33",
pages = "196--241",
journal = "Econometric Theory",
issn = "0266-4666",
publisher = "Cambridge University Press",
number = "1",

}

TY - JOUR

T1 - Identification and inference on regressions with missing covariate data

AU - Aucejo, Esteban

AU - Bugni, Federico A.

AU - Hotz, V. Joseph

PY - 2017/2/1

Y1 - 2017/2/1

N2 - This paper examines the problem of identification and inference on a conditional moment condition model with missing data, with special focus on the case when the conditioning covariates are missing. We impose no assumption on the distribution of the missing data and we confront the missing data problem by using a worst case scenario approach. We characterize the sharp identified set and argue that this set is usually too complex to compute or to use for inference. Given this difficulty, we consider the construction of outer identified sets (i.e. supersets of the identified set) that are easier to compute and can still characterize the parameter of interest. Two different outer identification strategies are proposed. Both of these strategies are shown to have nontrivial identifying power and are relatively easy to use and combine for inferential purposes.

AB - This paper examines the problem of identification and inference on a conditional moment condition model with missing data, with special focus on the case when the conditioning covariates are missing. We impose no assumption on the distribution of the missing data and we confront the missing data problem by using a worst case scenario approach. We characterize the sharp identified set and argue that this set is usually too complex to compute or to use for inference. Given this difficulty, we consider the construction of outer identified sets (i.e. supersets of the identified set) that are easier to compute and can still characterize the parameter of interest. Two different outer identification strategies are proposed. Both of these strategies are shown to have nontrivial identifying power and are relatively easy to use and combine for inferential purposes.

UR - http://www.scopus.com/inward/record.url?scp=84940702400&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84940702400&partnerID=8YFLogxK

U2 - 10.1017/S0266466615000250

DO - 10.1017/S0266466615000250

M3 - Article

VL - 33

SP - 196

EP - 241

JO - Econometric Theory

JF - Econometric Theory

SN - 0266-4666

IS - 1

ER -