Has someone ever heard of PCSE (panel corrected standard errors). Is it robust as fixed effects?
Economics Job Market Rumors » Econometrics Job Rumors
Has someone ever heard of PCSE?
(6 posts)-
Posted 5 months ago #
-
Posted 5 months ago #
-
I have heard of panel corrected standard errors...not sure what you mean by "robust as fixed effects" though. They are used when you think there is panel heteroskedasticity and/or autocorrelation (due to the time length of the panel). I'm not sure exactly the difference with robust standard errors, although Stata can do both so my guess is that there IS a difference. Nathaniel Beck at NYU (poli-sci) has all these papers aimed at explaining some of these concepts for poli-sci researchers. My suggestion would be to check out his vita...something might be helpful.
Posted 5 months ago # -
http://www.nyu.edu/gsas/dept/politics/faculty/beck/beckkatz.pdf
You might want to also look at Maddala (1998) Political Analysis
Posted 5 months ago # -
The poli sci guys use this all the time. I think it has something to do with having large T and small N (i.e. 4 countries and 50 years series).
Posted 5 months ago # -
It's the opposite. The 1995 Beck & Katz paper is about how the Parks method can only be used when T is at least as large as N, and only good when T is much greater than N. PCSE are meant to be used when N is large and T is relatively small (e.g. all countries from 1992-2000).
Posted 1 month ago #