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

Re: st: RE: GMM Estimation for Panel Data

From   Mark Schaffer <>
To, "Steven Stillman (LMPG)" <>
Subject   Re: st: RE: GMM Estimation for Panel Data
Date   Sun, 20 Oct 2002 11:39:24 +0100 (BST)

A quick follow-up to Steve's email.  The choice of whether to first-difference or mean-difference 
depends on whether you think the fixed/random effects are endogenous, and whether the transformed 
data are serially correlated or not.

If the effects are exogenous, then ivreg2,gmm on mean-deviation data is what you want; if the 
effects are endogenous, then you eliminate them by first-differencing and then using ivreg2,gmm.

If, after whatever transformation you've settled on, the error term of the resulting data series is 
not serially correlated, you're fine.  If it is serially correlated, then you should use ivreg2,gmm 
with the cluster(id) option, where id identifies a firm.  This will give you coefficients that are 
efficient in the presence of, and standard errors that are consistent to the presence of, arbitrary 
within-cluster correlation, which in your case means consistent to arbitrary serial correlation.

--Mark (another ivreg2 coauthor)

Quoting "Steven Stillman (LMPG)" <>:

> Bersant.  You can run ivreg2, gmm on data that has been first-differenced or
> mean-differenced (for fixed effects) to correctly produce iv-gmm results for
> panel data.  As the standard errors in these models are based on large
> sample properties you do not need to correct the standard errors for the
> reduction in degrees of freedom resulting from differencing the data,
> although there is some debate out there to whether a correction should still
> be done.  An extension to ivreg2 (I am a co-author) to handle all panel data
> models is currently in the planning stages but I am not sure how soon it
> will be available.
> Steve
> > -----Original Message-----
> > From:	Bersant Hobdari []
> > Sent:	Friday, October 18, 2002 7:37 PM
> > To:	statalist
> > Subject:	st: GMM Estimation for Panel Data
> > 
> > Hello to all List Members,
> > 
> > I am trying to test the existence of returns to scale accross firms with
> > different ownership structures. As the existence of simultaneity bias
> > (due to correlation of inputs and the error term) makes OLS results
> > inconsistent, I am using the within, first-differenced and IV-GMM 
> > to account for this problem. 
> > 
> > Here however I run into a problem. The IV-GMM procedures for Stata
> > (ivgmm0 and ivreg2, respectively) are written not for panel data.
> > Xtabond takes account for the panel nature of the data but it considers
> > only dynamic models, that is not the case with the models I am
> > esatimating. My questin is whether the former two procedures can be
> > modified to take into account the panel nature of the data?
> > 
> > Any suggestion would be highly appreciated. 
> > Sincerely,
> > Bersant
> > -- 
> > Bersant Hobdari
> > Center for East European Studies
> > Institute of International Economics and Management
> > Copenhagen Business School
> > Howitzvej 60, 2000 Frederiksberg, Denmark
> > Phone: office:  (45) 38153052
> > Phone: mobile: (45) 26140295
> > Fax:                (45) 3815 2500     
> > *
> > *   For searches and help try:
> > *
> > *
> > *
> The information contained in this document is intended only for the
>  addressee and is not necessarily the views nor the official 
> communication of the Department of Labour.  All final/official papers 
> which are sent from the Department will be sent by non-electronic
> means, on appropriate letterhead, signed by authorised personnel.
> *
> *   For searches and help try:
> *
> *
> *

Prof. Mark Schaffer
Director, CERT
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3008


This e-mail and any files transmitted with it are confidential
and intended solely for the use of the individual or entity to
whom it is addressed.  If you are not the intended recipient
you are prohibited from using any of the information contained
in this e-mail.  In such a case, please destroy all copies in
your possession and notify the sender by reply e-mail.  Heriot
Watt University does not accept liability or responsibility
for changes made to this e-mail after it was sent, or for
viruses transmitted through this e-mail.  Opinions, comments,
conclusions and other information in this e-mail that do not
relate to the official business of Heriot Watt University are
not endorsed by it.
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index