Bootstrap-Based Improvements for Inference with Clustered Errors Export

@article{cameron:gelbach:miller:2008, abstract = {Researchers have increasingly realized the need to account for within-group dependence in estimating standard errors of regression parameter estimates. The usual solution is to calculate cluster-robust standard errors that permit heteroskedasticity and within-cluster error correlation, but presume that the number of clusters is large. Standard asymptotic tests can over-reject, however, with few (five to thirty) clusters. We investigate inference using cluster bootstrap-t procedures that provide asymptotic refinement. These procedures are evaluated using Monte Carlos, including the example of Bertrand, Duflo, and Mullainathan (2004). Rejection rates of 10\% using standard methods can be reduced to the nominal size of 5\% using our methods}, author = {Cameron, Colin A. and Miller, Douglas and Gelbach, Jonah B.}, citeulike-article-id = {4510149}, citeulike-linkout-0 = {http://ideas.repec.org/p/ecl/ucdeco/06-21.html}, citeulike-linkout-1 = {http://www.mitpressjournals.org/doi/pdfplus/10.1162/rest.90.3.414}, journal = {The Review of Economics and Statistics}, keywords = {bootstrap, ce\_jsm09\_svy\_bstrap, resampling\_inference, survey\_resampling, wild\_bootstrap}, number = {3}, pages = {414--427}, posted-at = {2009-05-13 15:21:04}, priority = {2}, title = {Bootstrap-Based Improvements for Inference with Clustered Errors}, url = {http://ideas.repec.org/p/ecl/ucdeco/06-21.html}, volume = {90}, year = {2008} }

TY - JOUR ID - cameron:gelbach:miller:2008 L3 - citeulike-article-id:4510149 N2 - Researchers have increasingly realized the need to account for within-group dependence in estimating standard errors of regression parameter estimates. The usual solution is to calculate cluster-robust standard errors that permit heteroskedasticity and within-cluster error correlation, but presume that the number of clusters is large. Standard asymptotic tests can over-reject, however, with few (five to thirty) clusters. We investigate inference using cluster bootstrap-t procedures that provide asymptotic refinement. These procedures are evaluated using Monte Carlos, including the example of Bertrand, Duflo, and Mullainathan (2004). Rejection rates of 10% using standard methods can be reduced to the nominal size of 5% using our methods IS - 3 JF - The Review of Economics and Statistics EP - 427 TI - Bootstrap-Based Improvements for Inference with Clustered Errors VL - 90 SP - 414 KW - bootstrap KW - ce_jsm09_svy_bstrap KW - resampling_inference KW - survey_resampling KW - wild_bootstrap AU - Cameron, Colin AU - Miller, Douglas AU - Gelbach, Jonah PY - 2008/// UR - http://ideas.repec.org/p/ecl/ucdeco/06-21.html ER -