Cluster SE Bootstrap wp

Cluster SE Bootstrap wp - Working Paper Series...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
Working Paper Series Bootstrap-Based Improvements for Inference with Clustered Errors A. Colin Cameron University of California, Davis Douglas Miller University of California, Davis Jonah B. Gelbach Department of Economics, University of Maryland and College of Law, Florida State University July 14, 2006 Paper # 06-21 Microeconometrics researchers have increasingly realized the essential need to account for any within-group dependence in estimating standard errors of regression parameter estimates. The typical preferred solution is to calculate cluster-robust or sandwich standard errors that permit quite general heteroskedasticity and within-cluster error correlation, but presume that the number of clusters is large. In applications with few (5-30) clusters, standard asymptotic tests can over-reject considerably. We investigate more accurate inference using cluster bootstrap-t procedures that provide asymptotic refinement. These procedures are evaluated using Monte Carlos, including the much-cited differences-in-differences example of Bertrand, Mullainathan and Duflo (2004). In situations where standard methods lead to rejection rates in excess of ten percent (or more) for tests of nominal size 0.05, our methods can reduce this to five percent. In principle a pairs cluster bootstrap should work well, but in practice a Wild cluster bootstrap performs better. Department of Economics One Shields Avenue Davis, CA 95616 (530)752-0741 http://www.econ.ucdavis.edu/working_search.cfm
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Bootstrap-Based Improvements for Inference with Clustered Errors A.Colin Cameron & , Jonah B. Gelbach y and Douglas L. Miller z June 14, 2006 Abstract Microeconometrics researchers have increasingly realized the essen- tial need to account for any within-group dependence in estimating standard errors of regression parameter estimates. The typical pre- ferred solution is to calculate cluster-robust or sandwich standard er- rors that permit quite general heteroskedasticity and within-cluster er- ror correlation, but presume that the number of clusters is large. In ap- plications with few (5-30) clusters, standard asymptotic tests can over- reject considerably. We investigate more accurate inference using clus- ter bootstrap-t procedures that provide asymptotic re&nement. These procedures are evaluated using Monte Carlos, including the much-cited di/erences-in-di/erences example of Bertrand, Mullainathan and Du±o (2004). In situations where standard methods lead to rejection rates in excess of ten percent (or more) for tests of nominal size 0 : 05 , our methods can reduce this to &ve percent. In principle a pairs cluster bootstrap should work well, but in practice a Wild cluster bootstrap performs better. Keywords: clustered errors; random e/ects; cluster robust; sand- wich; bootstrap; bootstrap-t; clustered bootstrap; pairs bootstrap; wild bootstrap.
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 12/26/2011 for the course ECON 245a taught by Professor Staff during the Fall '08 term at UCSB.

Page1 / 52

Cluster SE Bootstrap wp - Working Paper Series...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online