hi all
stata's online faq section makes the following claim :
suppose i have panel observations on x and y and i compute within-group
averages of x (call this "avgx") and deviations of x from avgx (call this
"deltax"), and i run
xtreg y avgx deltax, re ....................(1)
then the coefficient on avgx should be the between estimator (obtained by
running "xtreg y x, be") and the coefficient on deltax should be the within
estimator (obtained by running "xtreg y x, fe").
however, when i try (1), i get different estimators (numerically, i.e.) from
the ones i get by using the "fe" and "be" options. has anyone encountered
this before? if so, do you have any explanation for these differences? does
it have anything to do with the fact that i am working with an unbalanced
panel? many thanks.
deep
