#### DMCA

## How much should we trust differences-in-differences estimates? (2003)

### Cached

### Download Links

Citations: | 828 - 1 self |

### Citations

5188 | An introduction to the bootstrap.
- Efron, Tibshirani
- 1993
(Show Context)
Citation Context ...tart by examining a simulation-based technique. We then examine three other techniques that can be more readily implemented using standard statistical packages. IV.B. Block Bootstrap Block bootstrap [=-=Efron and Tibshirani 1994-=-] is a variant of bootstrap which maintains the autocorrelation structure by keeping all the observations that belong to the same group (e.g., state) together. In practice, we bootstrap the t-statisti... |

3544 | Econometric Analysis of Cross Section and Panel Data. - Wooldridge - 2002 |

3211 | A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity.” Econometrica XLVIII: - White - 1980 |

1196 | Problems with Instrumental Variables Estimation when the Correlation between the Instruments and the Endogenous Explanatory Variable Is Weak’. - Bound, Jaeger, et al. - 1995 |

1000 |
Biases in dynamic models with fixed effects.
- Nickell
- 1981
(Show Context)
Citation Context ...N-DIFFERENCES? 273 literature. First, “naive” parametric corrections, which do not take into account the bias in the estimation of the autocorrelation parameters in short time series, perform poorly [=-=Nickell 1981-=-]. Furthermore, the time series lengths typical to DD applications are generally too short to reliably estimate more �exible data generating processes and misspeci�cation of the process leads to incon... |

695 |
An illustration of a pitfall in estimating the effects of aggregate variables on micro units.
- Moulton
- 1990
(Show Context)
Citation Context ...of DD estimates. 3. This correction accounts for the presence of a common random effect at the state-year level. For example, economic shocks may affect all individuals in a state on an annual basis [=-=Moulton 1990-=-; Donald and Lang 2001]. Ignoring this grouped data problem can lead to inconsistent standard errors. In most of what follows, we will assume that the researchers estimating equation (1) have already ... |

327 | The Use of Time Series Processes to Model the Error Structure of Earnings in a Longitudinal Data Analysis,” - MaCurdy - 1982 |

316 |
Asymptotic Theory for Econometricians.
- WHITE
- 2001
(Show Context)
Citation Context ...estimate each element of the variance-covari-SHOULD WE TRUST DIFFERENCES-IN-DIFFERENCES? 271 ance matrix in this case, but we can use a generalized White-like formula to compute the standard errors [=-=White 1984-=-; Arellano 1987; Kezdi 2002]. 23 This estimator for the variance-covariance matrix is given by W 5 ~V9V! 21 S O jD N u9 ju j51 ~V9V!21, where N is the total number of states, V is matrix of independen... |

243 |
Computing robust standard errors for within-groups estimators.
- Arellano
- 1987
(Show Context)
Citation Context ...h element of the variance-covari-SHOULD WE TRUST DIFFERENCES-IN-DIFFERENCES? 271 ance matrix in this case, but we can use a generalized White-like formula to compute the standard errors [White 1984; =-=Arellano 1987-=-; Kezdi 2002]. 23 This estimator for the variance-covariance matrix is given by W 5 ~V9V! 21 S O jD N u9 ju j51 ~V9V!21, where N is the total number of states, V is matrix of independent variables (ye... |

196 | Unnatural Experiments? Estimating the Incidence of Endogenous Policies.” - Besley, Case - 1994 |

192 | What We Know and Do Not Know About the Natural Rate of Unemployment’, - Blanchard, Katz - 1997 |

172 | Sendhil Mullainathan, "How Much Should We Trust Differences-in-Differences Estimates - Bertrand, Duflo |

167 | Semiparametric Difference-in-Differences Estimators,” - Abadie - 2006 |

132 | An Introduction to the Bootstrap, Monographs on Statistics and Applied Porability, - Efron, Tibshirani - 1993 |

128 | Causal Parameters and Policy Analysis in Economics: - Heckman - 2000 |

95 |
Robust Standard Error Estimation in Fixed-Effects Panel Models.” Hungarian Statistical Review 9:95-116
- Kezdi
- 2004
(Show Context)
Citation Context ...e variance-covari-SHOULD WE TRUST DIFFERENCES-IN-DIFFERENCES? 271 ance matrix in this case, but we can use a generalized White-like formula to compute the standard errors [White 1984; Arellano 1987; =-=Kezdi 2002-=-]. 23 This estimator for the variance-covariance matrix is given by W 5 ~V9V! 21 S O jD N u9 ju j51 ~V9V!21, where N is the total number of states, V is matrix of independent variables (year dummies, ... |

84 | Identification and inference in nonlinear difference-indifferences models. - Athey, Imbens - 2006 |

61 | Bias Reduction in Standard Errors for Linear Regression with Multi-Stage Samples.” - Bell, McCaffrey - 2002 |

33 |
Inference with Difference
- Donald, Lang
- 2007
(Show Context)
Citation Context ...s. 3. This correction accounts for the presence of a common random effect at the state-year level. For example, economic shocks may affect all individuals in a state on an annual basis [Moulton 1990; =-=Donald and Lang 2001-=-]. Ignoring this grouped data problem can lead to inconsistent standard errors. In most of what follows, we will assume that the researchers estimating equation (1) have already accountedSHOULD WE TR... |

29 |
Estimation of fixed effect models for time series of cross-sections with arbitrary intertemporal covariance,
- Kiefer
- 1980
(Show Context)
Citation Context ... to compute standard errors. Under the assumption that there is no heteroskedasticity, this method will produce consistent estimates of the standard error as N (the number of groups) goes to in�nity [=-=Kiefer 1980-=-]. Table VII investigates how well this technique performs in practice in the CPS and AR(1) Monte Carlo studies. The method performs well when the number of states is large (N 5 50). The rejection rat... |

29 | Natural and quasi-natural experiments in economics. - Meyer - 1995 |

24 |
Econometric Analysis (New
- GREENE
- 1993
(Show Context)
Citation Context ...r depends on (1) the typical length of the time series used, (2) the serial correlation of the most commonly used dependent variables, and (3) whether any procedures have been used to correct for it [=-=Greene 2002-=-]. Since these factors are inherently empirical, we collected data on all DD papers published in six journals between 1990 and 2000. 5 We classi�ed a paper as “DD” if it focuses on speci�c interventio... |

14 | Hodges-Lehmann Point Estimates of Treatment Effect in Observational Studies - Rosenbaum - 1993 |

13 | Estimating autocorrelations in fixed-effects models. - Solon - 1984 |

8 | Observational studies and nonrandomized experiments. - Rosenbaum - 1999 |

3 |
Observational Studies and Nonrandomized Experiments, Handbook of Statistics (Amsterdam
- Rosenbaum
- 1996
(Show Context)
Citation Context ... laws were randomly assigned, the distribution of the parameter estimates obtained using these placebo laws could be used to form a randomization inference test of the signi�cance of the DD estimate [=-=Rosenbaum 1996-=-]. 12. Note that we are randomizing the treatment variable while keeping the set of outcomes �xed. In general, the distribution of the test statistic induced by such randomization is not a standard no... |

1 | SemiparametricDifference-in-DifferencesEstimators,” Working Paper - Abadie - 2000 |

1 | R., “An Illustration of a Pitfall in Estimating the Effects of Aggregate Variables in Micro Units,” Review of Economics and Statistics, LXXII - Meyer - 1995 |

1 | Unless mentioned otherwise under ”Modifications,” reported in the last two columns are the OLS rejection rates of the null hypothesis of no effect (at the 5 percent significance level) on the intervention variable for randomly generated placebo interventi - aNotes |