Dear Forum,
I have a panel of firm-year observations from different countries. A given firm will be constantly attributed to one specific country at all times and furthermore be assigned to a treatment or control group ("Treat"=0 or 1 respectively).
Years pre-treatment are coded as Post=0 and years post-treatment with Post=1
I want to run a difference-in-difference accounting for country specific characteristics that influence the level of the dependent variable. Therefore my idea is to include country fixed-effects.
The countries feature a varying amount of firms and also the proportion of firms with Treat=0 or = 1 varies between countries. idc is unique per firm.
My formal model looks like this:
Array
I estimate it using reghdfe and cluster standard errors on firm level:
xtset idc year
reghdfe DepV Post Treat Interaction (list of controls), absorb (country) vce(cluster idc)
The regression runs fine. My question is if it is alright to use fixed effects in a DID model, since demeaning also influences the Treat, Post and Interaction. And if it is okay, if reghdfe is a proper way to do so?
Thank you for your support!
Related Posts with Inclusion of country-fixed effects in a difference-in-difference model, specified on firm-year level and using REGHDFE
Adding Independent Variables in RegressionHi there! I am new to Stata and am eager to learn. I am currently doing an OLS regression on a cros…
Test for math \[ \frac{2}{3}+ \mathbf{\pi} \] \[ y = \beta*x + \gamma *x * \varepsilon \] …
Estimating TFP by industry using prodestDear Stata users, I estimate TFP using prodest command (ssc install prodest) and have one question …
Shaping dataHello, I would like to transform my data from this way : order station up down 1 A 48 0 …
create dummy showing 36&60 months after event dateDear statalists, I am trying to create a variable to flag the 36th and 60th months after each event…
Subscribe to:
Post Comments (Atom)
0 Response to Inclusion of country-fixed effects in a difference-in-difference model, specified on firm-year level and using REGHDFE
Post a Comment