Micro-Econometrics for Policy, Program and Treatment Effects

Name: Micro-Econometrics for Policy, Program and Treatment Effects
Brand: Oxford University Press
Price: 69.49 EUR
Availability: OnlineOnly

LEE MYOUNG-JAE(Autor*in)

Oxford University Press

1. Auflage

Erschienen am 7. April 2005

263 Seiten

E-Book

PDF mit Adobe-DRM

Systemvoraussetzungen

978-0-19-153316-7 (ISBN)

69,49 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Adobe-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
Contents
1 Tour of the book
2 Basics of treatment effect analysis
2.1 Treatment intervention, counter-factual, and causal relation
2.1.1 Potential outcomes and intervention
2.1.2 Causality and association
2.1.3 Partial equilibrium analysis and remarks
2.2 Various treatment effects and no effects
2.2.1 Various effects
2.2.2 Three no-effect concepts
2.2.3 Further remarks
2.3 Group-mean difference and randomization
2.3.1 Group-mean difference and mean effect
2.3.2 Consequences of randomization
2.3.3 Checking out covariate balance
2.4 Overt bias, hidden (covert) bias, and selection problems
2.4.1 Overt and hidden biases
2.4.2 Selection on observables and unobservables
2.4.3 Linear models and biases
2.5 Estimation with group mean difference and LSE
2.5.1 Group-mean difference and LSE
2.5.2 A job-training example
2.5.3 Linking counter-factuals to linear models
2.6 Structural form equations and treatment effect
2.7 On mean independence and independence*
2.7.1 Independence and conditional independence
2.7.2 Symmetric and asymmetric mean-independence
2.7.3 Joint and marginal independence
2.8 Illustration of biases and Simpson's Paradox*
2.8.1 Illustration of biases
2.8.2 Source of overt bias
2.8.3 Simpson's Paradox
3 Controlling for covariates
3.1 Variables to control for
3.1.1 Must cases
3.1.2 No-no cases
3.1.3 Yes/no cases
3.1.4 Option case
3.1.5 Proxy cases
3.2 Comparison group and controlling for observed variables
3.2.1 Comparison group bias
3.2.2 Dimension and support problems in conditioning
3.2.3 Parametric models to avoid dimension and support problems
3.2.4 Two-stage method for a semi-linear model*
3.3 Regression discontinuity design (RDD) and before-after (BA)
3.3.1 Parametric regression discontinuity
3.3.2 Sharp nonparametric regression discontinuity
3.3.3 Fuzzy nonparametric regression discontinuity
3.3.4 Before-after (BA)
3.4 Treatment effect estimator with weighting*
3.4.1 Effect on the untreated
3.4.2 Effects on the treated and on the population
3.4.3 Effciency bounds and effcient estimators
3.4.4 An empirical example
3.5 Complete pairing with double sums*
3.5.1 Discrete covariates
3.5.2 Continuous or mixed (continuous or discrete) covariates
3.5.3 An empirical example
4 Matching
4.1 Estimators with matching
4.1.1 Effects on the treated
4.1.2 Effects on the population
4.1.3 Estimating asymptotic variance
4.2 Implementing matching
4.2.1 Decisions to make in matching
4.2.2 Evaluating matching success
4.2.3 Empirical examples
4.3 Propensity score matching
4.3.1 Balancing observables with propensity score
4.3.2 Removing overt bias with propensity-score
4.3.3 Empirical examples
4.4 Matching for hidden bias
4.5 Difference in differences (DD)
4.5.1 Mixture of before-after and matching
4.5.2 DD for post-treatment treated in no-mover panels
4.5.3 DD with repeated cross-sections or panels with movers
4.5.4 Linear models for DD
4.5.5 Estimation of DD
4.6 Triple differences (TD)*
4.6.1 TD for qualified post-treatment treated
4.6.2 Linear models for TD
4.6.3 An empirical example
5 Design and instrument for hidden bias
5.1 Conditions for zero hidden bias
5.2 Multiple ordered treatment groups
5.2.1 Partial treatment
5.2.2 Reverse treatment
5.3 Multiple responses
5.4 Multiple control groups
5.5 Instrumental variable estimator (IVE)
5.5.1 Potential treatments
5.5.2 Sources for instruments
5.5.3 Relation to regression discontinuity design
5.6 Wald estimator, IVE, and compliers
5.6.1 Wald estimator under constant effects
5.6.2 IVE for heterogenous effects
5.6.3 Wald estimator as effect on compliers
5.6.4 Weighting estimators for complier effects*
6 Other approaches for hidden bias*
6.1 Sensitivity analysis
6.1.1 Unobserved confounder affecting treatment
6.1.2 Unobserved confounder affecting treatment and response
6.1.3 Average of ratios of biased to true effects
6.2 Selection correction methods
6.3 Nonparametric bounding approaches
6.4 Controlling for post-treatment variables to avoid confounder
7 Multiple and dynamic treatments*
7.1 Multiple treatments
7.1.1 Parameters of interest
7.1.2 Balancing score and propensity score matching
7.2 Treatment duration effects with time-varying covariates
7.3 Dynamic treatment effects with interim outcomes
7.3.1 Motivation with two-period linear models
7.3.2 G algorithm under no unobserved confounder
7.3.3 G algorithm for three or more periods
Appendix
A.1 Kernel nonparametric regression
A.2 Appendix for Chapter 2
A.2.1 Comparison to a probabilistic causality
A.2.2 Learning about joint distribution from marginals
A.3 Appendix for Chapter 3
A.3.1 Derivation for a semi-linear model
A.3.2 Derivation for weighting estimators
A.4 Appendix for Chapter 4
A.4.1 Non-sequential matching with network flow algorithm
A.4.2 Greedy non-sequential multiple matching
A.4.3 Nonparametric matching and support discrepancy
A.5 Appendix for Chapter 5
A.5.1 Some remarks on LATE
A.5.2 Outcome distributions for compliers
A.5.3 Median treatment effect
A.6 Appendix for Chapter 6
A.6.1 Controlling for affected covariates in a linear model
A.6.2 Controlling for affected mean-surrogates
A.7 Appendix for Chapter 7
A.7.1 Regression models for discrete cardinal treatments
A.7.2 Complete pairing for censored responses
References
Index
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
V
W
Z

Systemvoraussetzungen

Als PDF speichern Als Link merken