Empirical likelihood for regression discontinuity design

Taisuke Otsu, Ke Li Xu, Yukitoshi Matsushita

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

This paper proposes empirical likelihood based inference methods for causal effects identified from regression discontinuity designs. We consider both the sharp and fuzzy regression discontinuity designs and treat the regression functions as nonparametric. The proposed inference procedures do not require asymptotic variance estimation and the confidence sets have natural shapes, unlike the conventional Wald-type method. These features are illustrated by simulations and an empirical example which evaluates the effect of class size on pupils' scholastic achievements. Furthermore, for the sharp regression discontinuity design, we show that the empirical likelihood statistic admits a higher-order refinement, so-called the Bartlett correction. Bandwidth selection methods are also discussed.

Original languageEnglish
Pages (from-to)94-112
Number of pages19
JournalJournal of Econometrics
Volume186
Issue number1
DOIs
Publication statusPublished - 2015 May 1
Externally publishedYes

Fingerprint

Empirical likelihood
Regression discontinuity design
Inference
Statistics
Confidence set
Simulation
Causal effect
Fuzzy regression
Class size
Asymptotic variance
Bandwidth
Bartlett correction
Variance estimation

Keywords

  • Bartlett correction
  • Empirical likelihood
  • Nonparametric methods
  • Regression discontinuity design
  • Treatment effect

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Empirical likelihood for regression discontinuity design. / Otsu, Taisuke; Xu, Ke Li; Matsushita, Yukitoshi.

In: Journal of Econometrics, Vol. 186, No. 1, 01.05.2015, p. 94-112.

Research output: Contribution to journalArticle

Otsu, Taisuke ; Xu, Ke Li ; Matsushita, Yukitoshi. / Empirical likelihood for regression discontinuity design. In: Journal of Econometrics. 2015 ; Vol. 186, No. 1. pp. 94-112.
@article{81ded2ddab2e46ebb09384daf25dc891,
title = "Empirical likelihood for regression discontinuity design",
abstract = "This paper proposes empirical likelihood based inference methods for causal effects identified from regression discontinuity designs. We consider both the sharp and fuzzy regression discontinuity designs and treat the regression functions as nonparametric. The proposed inference procedures do not require asymptotic variance estimation and the confidence sets have natural shapes, unlike the conventional Wald-type method. These features are illustrated by simulations and an empirical example which evaluates the effect of class size on pupils' scholastic achievements. Furthermore, for the sharp regression discontinuity design, we show that the empirical likelihood statistic admits a higher-order refinement, so-called the Bartlett correction. Bandwidth selection methods are also discussed.",
keywords = "Bartlett correction, Empirical likelihood, Nonparametric methods, Regression discontinuity design, Treatment effect",
author = "Taisuke Otsu and Xu, {Ke Li} and Yukitoshi Matsushita",
year = "2015",
month = "5",
day = "1",
doi = "10.1016/j.jeconom.2014.04.023",
language = "English",
volume = "186",
pages = "94--112",
journal = "Journal of Econometrics",
issn = "0304-4076",
publisher = "Elsevier BV",
number = "1",

}

TY - JOUR

T1 - Empirical likelihood for regression discontinuity design

AU - Otsu, Taisuke

AU - Xu, Ke Li

AU - Matsushita, Yukitoshi

PY - 2015/5/1

Y1 - 2015/5/1

N2 - This paper proposes empirical likelihood based inference methods for causal effects identified from regression discontinuity designs. We consider both the sharp and fuzzy regression discontinuity designs and treat the regression functions as nonparametric. The proposed inference procedures do not require asymptotic variance estimation and the confidence sets have natural shapes, unlike the conventional Wald-type method. These features are illustrated by simulations and an empirical example which evaluates the effect of class size on pupils' scholastic achievements. Furthermore, for the sharp regression discontinuity design, we show that the empirical likelihood statistic admits a higher-order refinement, so-called the Bartlett correction. Bandwidth selection methods are also discussed.

AB - This paper proposes empirical likelihood based inference methods for causal effects identified from regression discontinuity designs. We consider both the sharp and fuzzy regression discontinuity designs and treat the regression functions as nonparametric. The proposed inference procedures do not require asymptotic variance estimation and the confidence sets have natural shapes, unlike the conventional Wald-type method. These features are illustrated by simulations and an empirical example which evaluates the effect of class size on pupils' scholastic achievements. Furthermore, for the sharp regression discontinuity design, we show that the empirical likelihood statistic admits a higher-order refinement, so-called the Bartlett correction. Bandwidth selection methods are also discussed.

KW - Bartlett correction

KW - Empirical likelihood

KW - Nonparametric methods

KW - Regression discontinuity design

KW - Treatment effect

UR - http://www.scopus.com/inward/record.url?scp=84926418721&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84926418721&partnerID=8YFLogxK

U2 - 10.1016/j.jeconom.2014.04.023

DO - 10.1016/j.jeconom.2014.04.023

M3 - Article

VL - 186

SP - 94

EP - 112

JO - Journal of Econometrics

JF - Journal of Econometrics

SN - 0304-4076

IS - 1

ER -