Hi John, and fellow Stata users,
Thank you for your quick response. Yes, I am familiar with the GLS option and
have used that, but prefer to do simple OLS with "robust" standard errors that
are consistently estimated in the presence of both heteroskedasticity and serial
correlation. I have also used the "xtpcse" command in STATA 7, that the HELP
menu reports as using the Prais-Winsten regression. I am not familiar with this
estimation method, and would appreciate help in this regard as well.
Best,
Stuti
"John A Karikari"
<[email protected]> To: <[email protected]>,
Sent by: <[email protected]>
[email protected] cc:
arvard.edu Subject: Re: st: Newey West errors with panel
data
09/26/2002 02:18 PM
Please respond to
statalist
As an alternative to Newey West, if you do not have lagged dependent
variables in your model, you could use the XTGLS, which estimates
cross-sectional time-series models using the FGLS technique--it allows
estimation with AR(1) within panels and heteroscedasticity across
panels.
For the IV estimation, you could use a 2SLS approach with the XTGLS
>>> [email protected] 09/26/02 01:58PM >>>
I would very much appreciate receiving advice on how to estimate
Newey-West
standard errors in a panel data setting in Stata, and with
instrumental
variables (Stata command: IVREG) and three stage least squares (Stata
command:
REG3)
Thank you,
Stuti Khemani
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
