All Rights Reserved
AccessEcon LLC 2006, 2008.
Powered by MinhViet JSC

Robert Phillips
''Some Monte Carlo results for a generalized error component model with heteroskedastic disturbances''
( 2003, Vol. 3 No.15 )
This note provides Monte Carlo evidence illustrating that feasible and true GLS estimators of the Baltagi and Griffin (1988) generalized error component model do not have the same sampling behavior. Indeed, while the true GLS estimator is consistent, a feasible GLS estimator need not be, an observation corroborated by the Monte Carlo results.
JEL: C2 - Single Equation Models; Single Variables: General
Manuscript Received : Jun 13 2003 Manuscript Accepted : Jul 16 2003

  This abstract has been downloaded 2082 times                The Full PDF of this paper has been downloaded 153795 times