0

I am currently conducting a difference in difference (DID) analysis and fail to understand the benefits of fixed effects models in such cases. I ran a regular DID model with three regular independent variables (time, treatment and their interaction), however when I run the model with time (day) and stock fixed effects with the time and treat variable dropped, I essentially get the same results. Note: The dataset is panel data.

I initially tried to run the fixed effects version in order to control for the risk of every single company in my sample, but now I do not understand the benefit. Is the equivalence of results always the case?

Shawn Hemelstrand
  • 13,543
p_mercer123
  • 1
  • Does the “treatment” affect all treated stocks on the same day? – Thomas Bilach Feb 18 '24 at 06:01
  • Yes, the event ("treatment") affects all treated stocks on the same day. – p_mercer123 Feb 18 '24 at 10:58
  • Review my answer here to help you understand the link between DiD and fixed effects. As for the equivalence, it is quite normal to observe similar results when your panel consists of two well-defined treatment/control groups and two well-defined before/after periods. Expect your results to change once you deviate from this setting. My answer here should clear things up. – Thomas Bilach Feb 26 '24 at 23:37

0 Answers0