Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtpcse vs. xtgls / polsci vs. econ

From   Moritz Schularick <[email protected]>
To   [email protected]
Subject   st: xtpcse vs. xtgls / polsci vs. econ
Date   Fri, 23 Apr 2004 12:01:39 +0200 (CEST)

Dear Statalist,

I've been running both -xtpcse and -xtgls on my
serially correlated and heteroskedastic, unbalanced
panel (N=50, T=30).
Thanks to Beck/Katz (1995) I am prepared to treat the
-xtpcse results as more reliable statistically because
-xtgls produces overconfident test statistics. 

However, I was not prepared to see (for some
variables) markedly different coefficient estimates
(including switching signs) between the two, but
that's exactly what I get. 

Given heteroskedasticity, the coefficient estimates
via GLS (xtgls) could be more efficient than OLS
(xtpcse), but I am not sure how the small T, large N
problem affects the reliability of coefficient
estimates in -xtgls? Does anybody know? 

Btw, xtpcse seems to have emerged as the standard in
political science, but my impression is that in
economics it is much less used. Why?

Thanks in advance

Moritz Schularick
Free University Berlin


Mit sch�nen Gr��en von Yahoo! Mail -
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index