Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
AW: AW: st: RE: Lagged dependent variable with fixed effects regression
From
"Dithmer, Jan" <[email protected]>
To
"[email protected]" <[email protected]>
Subject
AW: AW: st: RE: Lagged dependent variable with fixed effects regression
Date
Tue, 10 Jul 2012 09:37:04 +0200
Erhan,
Nickell, S. (1981): Biases in dynamic models with fixed effects, Econometrica, 49, 1417-1426.
...and every recent book on panel data analysis will cover this issue, e.g.:
Hsiao,Cheng, 2003. "Analysis of Panel Data," Cambridge Books, Cambridge University Press.
For me it looks like a large N, small T case. Including the lagged dependent variable you have a dynamic panel model and may use Difference or System GMM,
see for example xtabond, xtabond2 by David Roodman or xtdpdsys
Hope this helps...
Jan
-----Ursprüngliche Nachricht-----
Von: [email protected] [mailto:[email protected]] Im Auftrag von Erhan Kilincarslan
Gesendet: Monday, July 09, 2012 5:33 PM
An: [email protected]
Betreff: RE: AW: st: RE: Lagged dependent variable with fixed effects regression
Thanks, Jan. I have around 850 fims with data from 2001-2011. So, T is 11 years. Even in fixed effects specifcation, as first year is omiited, T = 10 years. Hence, in this case, you are saying I would definitily get problems with endogenity??? if it is the case, should I use RE specification instead to report my results?
p.s. Can you please recommend me a book/books explaining endogenity problems related to lagged dependent variable and fixed effects?? Because, I couldnt find any
Regards
Erhan
> From: [email protected]
> To: [email protected]
> Date: Mon, 9 Jul 2012 16:15:59 +0200
> Subject: AW: st: RE: Lagged dependent variable with fixed effects
> regression
>
> Erhan,
>
> including the lagged dependent variable gives rise to dynamic panel or
> Nickell bias as it will be correlated with the error term in the fixed effects specification. However this bias diminishes with increasing T. Thus, if your time horizon is rather short, you may get problems with endogeneity.
>
> Best, Jan
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]] Im Auftrag von Erhan
> Kilincarslan
> Gesendet: Monday, July 09, 2012 3:39 PM
> An: [email protected]
> Betreff: RE: st: RE: Lagged dependent variable with fixed effects
> regression
>
> Well, my basic model is reg cashdiv earnings l.cashdiv and then i will add more IV and control variables. Howeever, when I run OLS pooled , FE and RE regressions, OLS and RE give similar results, FE gives similar but a bit different with very low coeeficients. I wonder it is because the lagged dependent variable in the model? Hausman test indicates FE rules over RE and B/P Lagrange test indicates RE rules over OLS. So, in this case should I go for RE? Is including Lagged dependent variable in FE is not appropriate in FE?
>
> ----------------------------------------
> > From: [email protected]
> > To: [email protected]
> > Subject: st: RE: Lagged dependent variable with fixed effects
> > regression
> > Date: Mon, 9 Jul 2012 12:43:00 +0000
> >
> > Erhan,
> >
> > That might depend on whether you have identification of the regression equation with the lagged dependent variables being included. If you have any endogeneity concerns, I hope you have some IV's.
> >
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]] On Behalf Of Erhan
> > Kilincarslan
> > Sent: Monday, July 09, 2012 8:36 AM
> > To: [email protected]
> > Subject: st: Lagged dependent variable with fixed effects regression
> >
> > Would you please tell me that if I include lagged dependent variable in the model and use fixed effects panel regression, would it be working correctly?
> > *
> > * For searches and help try:
> > * http://www.stata.com/help.cgi?search
> > * http://www.stata.com/support/statalist/faq
> > * http://www.ats.ucla.edu/stat/stata/
> >
> > *
> > * For searches and help try:
> > * http://www.stata.com/help.cgi?search
> > * http://www.stata.com/support/statalist/faq
> > * http://www.ats.ucla.edu/stat/stata/
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/