Wild Bootstrap Inference for Wildly Different Cluster Sizes

Summary The cluster robust variance estimator (CRVE) relies on the number of clusters being sufficiently large. Monte Carlo evidence suggests that the ‘rule of 42’ is not true for unbalanced clusters. Rejection frequencies are higher for datasets with 50 clusters proportional to US state populations than with 50 balanced clusters. Using critical values based on the wild cluster bootstrap performs much better. However, this procedure fails when a small number of clusters is treated. We explain why CRVE t statistics and the wild bootstrap fail in this case, study the ‘effective number’ of clusters and simulate placebo laws with dummy variable regressors. Copyright © 2016 John Wiley & Sons, Ltd.

[1]  Joshua D. Angrist,et al.  Mostly Harmless Econometrics: An Empiricist's Companion , 2008 .

[2]  Andrew V. Carter,et al.  Asymptotic Behavior of a t-Test Robust to Cluster Heterogeneity , 2017, Review of Economics and Statistics.

[3]  Ulrich K. Müller,et al.  t-Statistic Based Correlation and Heterogeneity Robust Inference , 2007 .

[4]  H. White,et al.  Some heteroskedasticity-consistent covariance matrix estimators with improved finite sample properties☆ , 1985 .

[5]  Rembert De Blander,et al.  Mostly Harmless Econometrics: An Empiricist's Companion , 2011 .

[6]  Inference with Di erence-inDi erences Revisited ∗ , 2013 .

[7]  J. MacKinnon,et al.  Econometric Theory and Methods , 2003 .

[8]  Michal Kolesár,et al.  Robust Standard Errors in Small Samples: Some Practical Advice , 2012, Review of Economics and Statistics.

[9]  S. Zeger,et al.  Longitudinal data analysis using generalized linear models , 1986 .

[10]  J. Angrist,et al.  Rural Windfall or a New Resource Curse? Coca, Income, and Civil Conflict in Colombia , 2005, The Review of Economics and Statistics.

[11]  Teun Kloek,et al.  OLS Estimation in a Model Where a Microvariable Is Explained by Aggregates and Contemporaneous Disturbances Are Equicorrelated , 1979 .

[12]  Douglas L. Miller,et al.  A Practitioner’s Guide to Cluster-Robust Inference , 2015, The Journal of Human Resources.

[13]  Brent R. Moulton.  An Illustration of a Pitfall in Estimating the Effects of Aggregate Variables on Micro Unit , 1990 .

[14]  J. MacKinnon,et al.  Bootstrap tests: how many bootstraps? , 2000 .

[15]  J. MacKinnon.  Bootstrap Methods in Econometrics , 2006 .

[16]  Stephen G. Donald,et al.  Inference with Difference-in-Differences and Other Panel Data , 2007, The Review of Economics and Statistics.

[17]  Timothy G. Conley,et al.  Inference with “Difference in Differences” with a Small Number of Policy Changes , 2005, The Review of Economics and Statistics.

[18]  H. White.  Asymptotic theory for econometricians , 1985 .

[19]  James G. MacKinnon,et al.  THE SIZE DISTORTION OF BOOTSTRAP TESTS , 1999, Econometric Theory.

[20]  E. Duflo,et al.  How Much Should We Trust Differences-in-Differences Estimates? , 2001 .

[21]  J. MacKinnon,et al.  Bootstrap Confidence Sets with Weak Instruments , 2014 .

[22]  Timothy G. Conley,et al.  Inference with dependent data using cluster covariance estimators , 2011 .

[23]  Matthew D. Webb.  Reworking wild bootstrap‐based inference for clustered errors , 2014, Canadian Journal of Economics/Revue canadienne d'économique.

[24]  Emmanuel Flachaire,et al.  The wild bootstrap, tamed at last , 2001 .

[25]  Daniel F. McCaffrey,et al.  Bias reduction in standard errors for linear regression with multi-stage samples , 2002 .