Research Paper IZA Discussion Papers 7742

Inference with difference-in-differences revisited

Authors

Publication date

Nov 2013

Summary

A growing literature on inference in difference-in-differences (DiD) designs with grouped errors has been pessimistic about obtaining hypothesis tests of the correct size, particularly with few groups. We provide Monte Carlo evidence for three points: (i) it is possible to obtain tests of the correct size even with few groups, and in many settings very straightforward methods will achieve this; (ii) the main problem in DiD designs with grouped errors is instead low power to detect real effects; and (iii) feasible GLS estimation combined with robust inference can increase power considerably whilst maintaining correct test size – again, even with few groups.

Subjects

Statistical Mathematics and Econometrics

Links

http://www.iza.org/en/webcontent/research/ra1/publications/papers/viewAbstract?dp_id=7742

Related publications

1. Inference with difference-in-differences revisited

   Mike Brewer, Thomas F. Crossley, and Robert Joyce

   Journal Article, Journal of Econometric Methods - Jan 2018
   1. Statistical Mathematics
   2. Econometrics

#522027