The core methods in today's econometric toolkit are linear regression for statistical control, instrumental variables methods for the analysis of natural experiments, and differences-in-differences methods that exploit policy changes. In the modern experimentalist paradigm, these techniques address clear causal questions such as: Do smaller classes increase learning? Should wife batterers be arrested? How much does education raise wages? Mostly Harmless Econometrics shows how the basic tools of applied econometrics allow the data to speak.
In addition to econometric essentials, Mostly Harmless Econometrics covers important new extensions--regression-discontinuity designs and quantile regression--as well as how to get standard errors right. Joshua Angrist and Jörn-Steffen Pischke explain why fancier econometric techniques are typically unnecessary and even dangerous. The applied econometric methods emphasized in this book are easy to use and relevant for many areas of contemporary social science.
- An irreverent review of econometric essentials
- A focus on tools that applied researchers use most
- Chapters on regression-discontinuity designs, quantile regression, and standard errors
- Many empirical examples
- A clear and concise resource with wide applications
Author(s): Joshua D. Angrist, Jorn-Steffen Pischke
Publisher: Princeton University Press
Year: 2008
Language: English
Commentary: this is a draft
Pages: 290
Preface......Page 11
Acknowledgments......Page 13
Organization of this Book......Page 15
I Introduction......Page 17
Questions about Questions......Page 19
The Experimental Ideal......Page 25
The Selection Problem......Page 26
Random Assignment Solves the Selection Problem......Page 28
Regression Analysis of Experiments......Page 32
II The Core......Page 35
Making Regression Make Sense......Page 37
Regression Fundamentals......Page 38
Economic Relationships and the Conditional Expectation Function......Page 39
Linear Regression and the CEF......Page 42
Asymptotic OLS Inference......Page 46
Saturated Models, Main Effects, and Other Regression Talk......Page 52
The Conditional Independence Assumption......Page 54
The Omitted Variables Bias Formula......Page 60
Bad Control......Page 63
Regression Meets Matching......Page 67
Control for Covariates Using the Propensity Score......Page 75
Propensity-Score Methods vs. Regression......Page 79
Weighting Regression......Page 82
Limited Dependent Variables and Marginal Effects......Page 85
Why is Regression Called Regression and What Does Regression-to-the-mean Mean?......Page 96
Appendix: Derivation of the average derivative formula......Page 97
Instrumental Variables in Action: Sometimes You Get What You Need......Page 99
IV and causality......Page 100
Two-Stage Least Squares......Page 105
The Wald Estimator......Page 110
Grouped Data and 2SLS......Page 116
The Limiting Distribution of the 2SLS Coefficient Vector......Page 119
Over-identification and the 2SLS Minimand......Page 121
Two-Sample IV and Split-Sample IV......Page 125
IV with Heterogeneous Potential Outcomes......Page 127
Local Average Treatment Effects......Page 128
The Compliant Subpopulation......Page 133
IV in Randomized Trials......Page 135
Counting and Characterizing Compliers......Page 139
LATE with Multiple Instruments......Page 146
Covariates in the Heterogeneous-effects Model......Page 147
Average Causal Response with Variable Treatment Intensity......Page 152
2SLS Mistakes......Page 157
Peer Effects......Page 160
Limited Dependent Variables Reprise......Page 163
The Bias of 2SLS......Page 169
Appendix......Page 176
Individual Fixed Effects......Page 181
Differences-in-differences......Page 185
Regression DD......Page 190
Fixed Effects versus Lagged Dependent Variables......Page 198
Appendix: More on fixed effects and lagged dependent variables......Page 200
III Extensions......Page 203
Sharp RD......Page 205
Fuzzy RD is IV......Page 212
Quantile Regression......Page 219
The Quantile Regression Model......Page 220
Censored Quantile Regression......Page 224
The Quantile Regression Approximation Property......Page 226
Tricky Points......Page 229
Quantile Treatment Effects......Page 230
The QTE Estimator......Page 232
Nonstandard Standard Error Issues......Page 237
The Bias of Robust Standard Errors......Page 238
Clustering and the Moulton Factor......Page 247
Serial Correlation in Panels and Difference-in-Difference Models......Page 252
Fewer than 42 clusters......Page 254
Appendix: Derivation of the simple Moulton factor......Page 257
Last words......Page 261
Acronyms......Page 263
Empirical Studies Index......Page 267
Notation......Page 269
