Dear Statalist,
I am estimating a regression model using system GMM through xtabond2.
I have an unbalanced panel dataset of about 50,000 firm-year
observations. A simplified version of my model has the following form:
y = a +bx1 + cx2 +e. I treat both x1 and x2 as endogenous using their
2nd and 3rd lags as instruments. I do not inlcude l1.y as a regressor.
To my surprise, I noticed that the standard error are exactly the same
whether or not I cluster them at the firm level. Does anybody have any
idea of why this could be the case?
Thanks,
Erasmo
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
