Scribd
Upload
124 views6 pages

Matching Regression

This document appears to be the table of contents for a book on causal inference methods including matching, regression discontinuity, difference-in-differences, and triple differences. The table of contents lists 13 chapters that cover the basics and various implementations of these methods. It also includes appendices on supplementary topics like kernel density estimation, bootstrap methods, and software programs for implementing some of the analyses.

Uploaded by

Youssef Ucef
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
124 views6 pages

Matching Regression

This document appears to be the table of contents for a book on causal inference methods including matching, regression discontinuity, difference-in-differences, and triple differences. The table of contents lists 13 chapters that cover the basics and various implementations of these methods. It also includes appendices on supplementary topics like kernel density estimation, bootstrap methods, and software programs for implementing some of the analyses.

Uploaded by

Youssef Ucef
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

Matching, Regression

Discontinuity, Difference in

Differences, and Beyond

Myoung-jae Lee,

OXFORD
UNIVERSITY PRESS

CONTENTS

Preface

xv

1 Basics of Treatment Effect Analysis


1.1 Counterfactual, Intervention, and Causal Relation
1.1.1 Potential Outcomes and Intervention
1.1.2 Causality and Association
1.1.3 Partial Equilibrium Analysis and Remarks
1.2 Various Treatment Effects and No Effects
1.2.1 Various Effects
1.2.2 Three No-Effect Concepts
1.2.3 Remarks
1.3 Group-Mean Difference and Randomization
1.3.1 Group-Mean Difference and Mean Effect
1.3.2 Consequences of Randomization
1.3.3 Checking Out Covariate Balance
1.4 Overt Bias, Hidden Bias, and Selection Problems
1.4.1 Overt and Hidden Biases
1.4.2 Selection on Observables and Unobservables
1.4.3 Linear Models and Biases
1.5 Estimation with Group Mean Difference and LSE
1.5.1 Group-Mean Difference and LSE
1.5.2 Job Training Example
1.5.3 Linking Counterfactuals to Linear Models
1.6 Structural Form, Assignment, and Marginal Model
1.6.1 Structural versus Reduced Forms for Response
1.6.2 Treatment Structural Form and Assignment
1.6.3 Marginal Structural Model
1.7 Simpsons Paradox and False Covariate Control

1
1
1
3
4
5
5
6
7
8
8
10
12
13
13
14
16
18
18
19
21
22
22
23
24
25

2 Matching
2.1 Basics of Matching and Various Effects
2.1.1 Main Idea
2.1.2 Effect on Treated and Effect on Population
2.1.3 Dimension and Support Problems
2.1.4 Variables to Control

28
28
28
30
31
32

vii

viii Contents
2.2 Implementing Matching
2.2.1 Decisions to Make in Matching
2.2.2 Matching Estimators
2.2.3 Asymptotic Variance Estimation
2.2.4 Labor Union Effect on Wage
2.3 Propensity Score Matching (PSM)
2.3.1 Propensity Score as a Balancing Score
2.3.2 Removing Overt Bias with Propensity Score
2.3.3 Implementing PSM and Bootstrap
2.3.4 PSM Empirical Examples
2.3.5 Propensity Score Specification Issues*
2.4 Further Remarks
2.4.1 Covariate Balance Check
2.4.2 Matching for Hidden Bias
2.4.3 Prognostic Score and More*
3 Nonmatching and Sample Selection
3.1 Weighting
3.1.1 Weighting Estimator for Effect on Population
3.1.2 Other Weighting Estimators and Remarks
3.1.3 Asymptotic Distribution of Weighting Estimators*
3.1.4 Job Training Effect on Unemployment
3.1.5 Doubly Robust Estimator*
3.1.6 Weighting for Missing Data*
3.2 Regression Imputation
3.2.1 Linear Regression Imputation
3.2.2 Regression Imputation with Propensity Score
3.2.3 Regression Imputation for Multiple Treatment
3.2.4 Regression Imputation for Continuous Treatment*
3.2.5 Military Service Effect on Wage
3.3 Complete Pairing with Double Sum
3.3.1 Discrete Covariates
3.3.2 Continuous Covariates
3.3.3 Nonparametric Distributional Effect Tests*
3.4 Treatment Effects under Sample Selection
3.4.1 Difficulties with Sample Selection Models
3.4.2 Parti cipation, Invisible, and Visible Effects
3.4.3 Identification of Three Effects with Mean Differences
3.4.4 Religiosity Effect on Affairs
3.5 Effect Decomposition in Sample Selection Models*
3.5.1 Motivation for Decomposition
3.5.2 Decomposition with Linear Selection Model
3.5.3 Four Special Models
3.5.4 Race Effect on Wage

35
35
37
40
44
46
46
47
48
50
52
54
54
56
58
61
61
61
63
65
66
67
68
69
70
71
72
73
74
76
77
79
80
84
85
86
87
88
90
90
91
92
94

ix Contents
4 Regression Discontinuity
4.1 Introducing RD with Before-After
4.1.1 BAExamples
4.1.2 BA Identification Assumption
4.1.3 FromBAtoRD
4.2 RD Identification and Features
4.2.1 Sharp RD (SRD) and Fuzzy RD (FRD)
4.2.2 Identification at Cutoff
4.2.3 RD Main Features
4.2.4 Class Size Effect on Test Score
4.3 RD Estimators
4.3.1 LSE for Level Equation
4.3.2 IVE for Right-Left Differenced Equation
4.3.3 Bandwidth Choice and Remarks
4.3.4 High School Completion Effect on Fertility
4.4 Specification Tests
4.4.1 Breaks in Conditional Means
4.4.2 Continuity in Score Density
4.5 RD Topics*
4.5.1 Spatial Breaks
4.5.2 RD for Limited Dependent Variables
4.5.3 Measurement Error in Score
4.5.4 Regression Kink (RK) and Generalization
4.5.5 SRD with Multiple Scores
4.5.6 Quantile RD

97
97
97
98
99
100
101
102
104
106
109
109
110
112
113
116
116
117
119
119
120
121
123
126
129

5 Difference in Differences
5.1 DD Basics
5.1.1 Examples for DD
5.1.2 Time-Constant and Time-Varying Qualifications
5.1.3 Data Requirement and Notation
5.2 DD with Repeated Cross-Sections
5.2.1 Identification
5.2.2 Identification with Parametric Models
5.2.3 Schooling Effect on Fertility: 'Fuzzy DD'
5.2.4 Linear Model Estimation for Two Periods or More
5.2.5 Earned Income Tax Credit Effect on Work
5.2.6 Time-Varying Qualification*
5.3 DD with Panel Data
5.3.1 Identification
5.3.2 Identification and Estimation with Parametric Models
5.3.3 Daylight Saving Time Effect on Energy
5.4 Panel Stayer DD for Time-Varying Qualification
5.4.1 Motivation

131
131
132
133
135
136
136
140
142
144
147
148
150
150
152
157
158
158

x Contents
5.4.2 Effect on In-Stayers Identified by Stayer DD
5.4.3 Identification and Estimation with Panel Linear Models
5.4.4 Pension Effect on Health Expenditure

159
160
162

6 Triple Difference and Beyond


6.1 TD Basics and More
6.2 TD with Repeated Cross-Sections
6.2.1 Identification
6.2.2 Identification and Estimation with Linear Models
6.2.3 Mandated Benefit Effect on Wage
6.3 TD with Panel Data
6.3.1 Identification
6.3.2 Estimation with Panel Linear Model
6.3.3 Tax-Inclusive Price Effect on Demand
6.4 GDD and Beyond
6.4.1 Motivation for GDD and Beyond
6.4.2 Identification for GDD and QD
6.4.3 Identified Effects When Panel Linear Model Holds
6.4.4 LSE for DD and GDD and Testing for DD Condition
6.4.5 Sulfa Drug Effect on Mortality: Is DD Trustworthy?
6.5 Clustering Problems and Inference for DD and TD
6.5.1 Single Clustering
6.5.2 Clustering in Panel Data
6.5.3 DD and TD with Cluster-Specific Treatment
6.5.4 Details on Cluster Variance Estimator*

165
165
166
166
169
172
174
174
175
177
178
179
180
181
182
184
187
188
194
199
202

A APPENDIX
A.l Kernel Density and Regression Estimators
A.l.l Histogram-Type Density Estimator
A.l.2 Kernel Density Estimator
A.l.3 Kernel Regression Estimator
A.l.4 Local Linear Regression
A.2 Bootstrap
A.2.1 Review on Usual Asymptotic Inference
A.2.2 Bootstrap to Find Quantiles
A.2.3 Percentile-t and Percentile Methods
A.2.4 Nonparametric, Parametric, and Wild Bootstraps
A.3 Confounder Detection, IVE, and Selection Correction
A.3.1 Coherence Checks
A.3.2 IVE and Compiler Effect
A.3.3 Selection Correction Approach
A.4 Supplements for DD Chapter
A.4.1 Nonparametric Estimators for Repeated Cross-Section DD
A.4.2 Nonparametric Estimation for DD with Two-Wave Panel Data

209
209
209
210
211
213
213
215
216
218
219
220
220
225
230
232
233
233

xi Contents
A.4.3 Panel Linear Model Estimation for DD with One-Shot
Treatment
A.4.4 Change in Changes
References
Index
Online GAUSS Programs:
Pair Matching with PS for Union on Wage (PairMatchUnionOnWage)
Regression Imputation with PS-Based Nonparametrics (EeglmpPsNprSim)
Complete Pairing with PS for Union on Wage (CpUnionOnWage)
RD Program (RdSim)
Repeated Cross-Section DD (DdReCroVary4WavesSim)
Panel DD for Differenced Model (DdPanel6WavesSim)
Repeated Cross-Section TD (TdReCro2WavesSim)
Panel DD and GDD for Differenced Model (DdGddPanel5WavesSim)
Panel DD, GDD and QD for Sulfa Drug (DdGddQdSulf aDrug)
Bootstrap for Sample Mean (BootAvgSim)
Selection Correction for Work on Doctor Visits (SelCorcWorkOnVisit)
Panel LSE, WIT and BET (PanelLseWitBetSim)

You might also like