Re: st: Spooky: Different Results with Repeated FE Estimation

[Miyu Lee <miyu.lee1@googlemail.com>]

Hi, thanks for your feedback so far:-)

I apologise for this general inquiry. The do file is just very long so
I did not think that replicating it might make sense. I was also
hoping that people would recognise this problem on a rather general
level.

Since I work with panel data, I first use xtset and then use the l.
operation to work with lags.

Apart from that as I said, I only use summation, multiplication, log
transformations throughout the do file.

But Dave mentioned that the lagging might be problematic. I lag
variables which I newly construct in my do file and I also lag
variables which are given in my dataset. Do you think that might be a
problem?

Guo, I get different coefficients, standard errors and constants. I use the
xtreg , fe r command throughout all estimations and with the same
computer and program.
*
*   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/