Panel Data Estimation Technique

[Amo Lewis]

Hello Everyone,
I am conducting some panel data analysis using fixed effects and system gmm. My total sample size is 510. When I run fixed effect model the reported sample size is 437, but the gmm reported sample size is 356. My question is why does the sample size varies so largely with the same data but different estimator?

[Carlo Lazzaro]

Amo:
with no example/excerpt of your data, it is difficull (for me, at least) to reply (more) positively.
My guess is that you're using two really different estimators. Probably, the different sample are due to the fact that, under -fe-, Stata (as always) automatically exclude all the observations with missing values in any variables, whereas -gmm- adds to it observations excluded due to lag calculation.
As an aside, I fail to get from your post why you're interested in comparing results from static and dynamic panel data regression.