Adding pre and post treatment dummies in DID regression


I want to estimate the impact of economic sanctions on GDP and other macroeconomic indicators on Italy in 1935 with DID. I have observations of GDP of Italy (treated) and of other countries (control) from 1929 to 1938. I created a time dummy equal to 1 if the Year is >= 1935 and a treatment dummy that is equal to 1 if the observation is for Italy (treatment group). I want to add pre and post treatment dummies for 3 years before and after the sanctions. My questions are:

1)Should I generate a dummy for every year pre and post treatment I want to consider? And how should the pre and post treatment dummies look like? For example, the pre-treatment dummy for 1934 (t= -1) should be equal to 1 only in 1934 and 0 in all other years, or it should be equal to 1 if year >=1934? And the post-treatment dummy for 1936 is equal to 1 only in 1936 or if Year >= 1936?

2)What should my regression look like with pre and post treatment dummies?

My dataset looks like this:

If you are doing a classical DiD, then just one dummy for before/after, one dummy for Italy/not Italy, and the interaction of the two.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.