Inference with Difference-in-Differences Revisited

Brewer, Mike and Crossley, Thomas and Joyce, Robert (2013) Inference with Difference-in-Differences Revisited. In: 8th IZA Conference on Labor Market Policy Evaluation co-hosted by the IFS and PEPA., 18 and 19 October 2013, London.

[thumbnail of IZA_paper_MB_TC_RJ.pdf]
IZA_paper_MB_TC_RJ.pdf - Published Version

Download (414kB) | Preview


A growing literature on inference in difference-in-differences (DiD) designs with grouped errors has been pessimistic about obtaining hypothesis tests of the correctsize, 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

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: difference in diferences, hypothesis test, power, cluster robust, feasible GLS
Subjects: 1. Frameworks for Research and Research Designs > 1.11 Evaluation Research
Depositing User: PEPA User
Date Deposited: 20 Nov 2013 10:24
Last Modified: 14 Jul 2021 13:58

Actions (login required)

View Item
View Item