Identification of treatment effects under imperfect matching with an application to Chinese elite schools

Hongliang ZHANG

doi:10.1016/j.jpubeco.2016.03.004

Identification of treatment effects under imperfect matching with an application to Chinese elite schools

Hongliang ZHANG^*

^*Corresponding author for this work

Research output: Contribution to journal › Journal article › peer-review

12 Citations (Scopus)

Abstract

This paper extends the treatment effect framework for causal inference to contexts in which the instrument appears in a data set that can only be linked imperfectly to the treatment and outcome variables contained in another data set. To overcome this problem, I form all pairwise links between information on the instrument and information on the treatment and outcome matched by the commonly recorded personal characteristics in both data sets. I show how these imperfect conditional matches can be used to identify both the average and distributional treatment effects for compliers of the common units of the two data sets. This multiple data source approach is then applied to analyze the effect of attending an elite middle school in a Chinese city where schools’ admissions lottery records can only be linked imperfectly to the administrative student records.

Original language	English
Pages (from-to)	56-82
Number of pages	27
Journal	Journal of Public Economics
Volume	142
DOIs	https://doi.org/10.1016/j.jpubeco.2016.03.004
Publication status	Published - 1 Oct 2016

User-Defined Keywords

Distributional treatment effects
Elite schools
Imperfect matching
Local average treatment effect
Student achievement

Access to Document

10.1016/j.jpubeco.2016.03.004

Cite this

@article{305f3a11b46e4c8e9f3c8281b6a62cdf,

title = "Identification of treatment effects under imperfect matching with an application to Chinese elite schools",

abstract = "This paper extends the treatment effect framework for causal inference to contexts in which the instrument appears in a data set that can only be linked imperfectly to the treatment and outcome variables contained in another data set. To overcome this problem, I form all pairwise links between information on the instrument and information on the treatment and outcome matched by the commonly recorded personal characteristics in both data sets. I show how these imperfect conditional matches can be used to identify both the average and distributional treatment effects for compliers of the common units of the two data sets. This multiple data source approach is then applied to analyze the effect of attending an elite middle school in a Chinese city where schools{\textquoteright} admissions lottery records can only be linked imperfectly to the administrative student records.",

keywords = "Distributional treatment effects, Elite schools, Imperfect matching, Local average treatment effect, Student achievement",

author = "Hongliang ZHANG",

note = "Funding Information: An earlier version of this paper was circulated under the title {"}The mirage of elite schools: Evidence from lottery-based school admissions in China.{"} I thank Joshua Angrist, Abhijit Banerjee, James Berry, Weili Ding, Esther Duflo, Rongzhu Ke, Cynthia Kinnan, Frank Levy, Steven Lehrer, Weifeng Li, Haoming Liu, Alex Mas, Roger Moon, Benjamin Olken, Florian Ploeckl, Karen R. Polenske, Aaron Sojourner, Zhentao Shi, Christopher Taber, William Wheaton, and seminar participants at CUHK, Georgetown University, Hong Kong Baptist University, HKUST, MIT, Nanyang Technological University, Peking University, Shanghai Jiao Tong University, University of Kent, University of Toronto, the 2010 AEA Annual Meeting, and the 2012 Singapore Conference on Evidence-based Public Policy Using Administrative Data for valuable discussions and comments. I am also grateful to the co-editor and three anonymous referees for their helpful comments and feedback, and Amy Ru Chien Tseng for her excellent research assistance. I acknowledge financial support from the Hong Kong Research Grants Council General Research Fund (No. 458610 ). All remaining errors are my own. ",

year = "2016",

month = oct,

day = "1",

doi = "10.1016/j.jpubeco.2016.03.004",

language = "English",

volume = "142",

pages = "56--82",

journal = "Journal of Public Economics",

issn = "0047-2727",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Identification of treatment effects under imperfect matching with an application to Chinese elite schools

AU - ZHANG, Hongliang

N1 - Funding Information: An earlier version of this paper was circulated under the title "The mirage of elite schools: Evidence from lottery-based school admissions in China." I thank Joshua Angrist, Abhijit Banerjee, James Berry, Weili Ding, Esther Duflo, Rongzhu Ke, Cynthia Kinnan, Frank Levy, Steven Lehrer, Weifeng Li, Haoming Liu, Alex Mas, Roger Moon, Benjamin Olken, Florian Ploeckl, Karen R. Polenske, Aaron Sojourner, Zhentao Shi, Christopher Taber, William Wheaton, and seminar participants at CUHK, Georgetown University, Hong Kong Baptist University, HKUST, MIT, Nanyang Technological University, Peking University, Shanghai Jiao Tong University, University of Kent, University of Toronto, the 2010 AEA Annual Meeting, and the 2012 Singapore Conference on Evidence-based Public Policy Using Administrative Data for valuable discussions and comments. I am also grateful to the co-editor and three anonymous referees for their helpful comments and feedback, and Amy Ru Chien Tseng for her excellent research assistance. I acknowledge financial support from the Hong Kong Research Grants Council General Research Fund (No. 458610 ). All remaining errors are my own.

PY - 2016/10/1

Y1 - 2016/10/1

N2 - This paper extends the treatment effect framework for causal inference to contexts in which the instrument appears in a data set that can only be linked imperfectly to the treatment and outcome variables contained in another data set. To overcome this problem, I form all pairwise links between information on the instrument and information on the treatment and outcome matched by the commonly recorded personal characteristics in both data sets. I show how these imperfect conditional matches can be used to identify both the average and distributional treatment effects for compliers of the common units of the two data sets. This multiple data source approach is then applied to analyze the effect of attending an elite middle school in a Chinese city where schools’ admissions lottery records can only be linked imperfectly to the administrative student records.

AB - This paper extends the treatment effect framework for causal inference to contexts in which the instrument appears in a data set that can only be linked imperfectly to the treatment and outcome variables contained in another data set. To overcome this problem, I form all pairwise links between information on the instrument and information on the treatment and outcome matched by the commonly recorded personal characteristics in both data sets. I show how these imperfect conditional matches can be used to identify both the average and distributional treatment effects for compliers of the common units of the two data sets. This multiple data source approach is then applied to analyze the effect of attending an elite middle school in a Chinese city where schools’ admissions lottery records can only be linked imperfectly to the administrative student records.

KW - Distributional treatment effects

KW - Elite schools

KW - Imperfect matching

KW - Local average treatment effect

KW - Student achievement

UR - https://www.scopus.com/pages/publications/84973578478

U2 - 10.1016/j.jpubeco.2016.03.004

DO - 10.1016/j.jpubeco.2016.03.004

M3 - Journal article

AN - SCOPUS:84973578478

SN - 0047-2727

VL - 142

SP - 56

EP - 82

JO - Journal of Public Economics

JF - Journal of Public Economics

ER -

Identification of treatment effects under imperfect matching with an application to Chinese elite schools

Abstract

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this