Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: Re: RE: Re: Two way Fixed Effects & Autocorrelation& Heteroskedasticity


From   "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: Re: RE: Re: Two way Fixed Effects & Autocorrelation& Heteroskedasticity
Date   Fri, 31 Mar 2006 14:39:14 +0100

Max, 

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu 
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of 
> kosak@gmx.de
> Sent: 31 March 2006 14:24
> To: statalist@hsphsun2.harvard.edu
> Subject: RE: st: RE: Re: RE: Re: Two way Fixed Effects & 
> Autocorrelation& Heteroskedasticity
> 
> thanks for your reply Mark.
> 
> The problem is that i don`t have an appropriate instrument 
> v.Can i use xtivreg without specifying an instrument 
> variable?

Yes, this is possible with -xtivreg2- (which I wrote).  It's also possible with Stata's official -xtivreg-, but the syntax is different, as you note, i.e. the -regress- option.

If you have Stata 9, you can simply use Stata's official -xtivreg- with cluster, or -areg- with cluster, to get SEs robust to het. and a.c.  The only reason to use -xtivreg2- is if you want, say, Newey-West-type robust standard errors and the like.

Cheers,
Mark

>(i think xtivreg allows for this with by the option 
> regress.) 
> 
> Is areg an alternative which i can choose?like i understood 
> areg, the standard errors will be robut to heteros. and 
> autocorrelation when i choose the cluster option.
> 
> Max
> 
> 
> 
> 
>  --- Ursprüngliche Nachricht ---
> > Von: "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
> > An: <statalist@hsphsun2.harvard.edu>
> > Betreff: RE: st: RE: Re: RE: Re: Two way Fixed Effects & 
> > Autocorrelation& Heteroskedasticity
> > Datum: Fri, 31 Mar 2006 09:38:10 +0100
> > 
> > Max,
> > 
> > > -----Original Message-----
> > > From: owner-statalist@hsphsun2.harvard.edu
> > > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of 
> > > kosak@gmx.de
> > > Sent: 31 March 2006 09:29
> > > To: statalist@hsphsun2.harvard.edu
> > > Subject: Re: st: RE: Re: RE: Re: Two way Fixed Effects & 
> > > Autocorrelation& Heteroskedasticity
> > > 
> > > Rodrigo,Mark,
> > > thank you very much for your advice.I will try xtivreg2.
> > > I have antother short question:
> > > Am i right, that for a one-way fixed effect model with the same 
> > > problems areg be the appropriate solution?
> > > Best,
> > > Max
> > 
> > Unless I am missing something obvious, the same remarks apply to 
> > one-way fixed effect models (and probably more directly, since 
> > xtivreg2 et al. are set up to handle one-way, not two-way, models).
> > 
> > --Mark
> > 
> > > 
> > > --- Ursprüngliche Nachricht ---
> > > > Von: "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
> > > > An: <statalist@hsphsun2.harvard.edu>
> > > > Betreff: st: RE: Re: RE: Re: Two way Fixed Effects &
> > > Autocorrelation&
> > > > Heteroskedasticity
> > > > Datum: Thu, 30 Mar 2006 23:09:50 +0100
> > > > 
> > > > Rodrigo,
> > > > 
> > > > > -----Original Message-----
> > > > > From: owner-statalist@hsphsun2.harvard.edu
> > > > > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf
> > > Of Rodrigo
> > > > > Alfaro
> > > > > Sent: 30 March 2006 23:03
> > > > > To: statalist@hsphsun2.harvard.edu
> > > > > Subject: st: Re: RE: Re: Two way Fixed Effects & 
> > > > > Autocorrelation& Heteroskedasticity
> > > > > 
> > > > > Mark's suggestion is a addendum to my second option. The 4th 
> > > > > proposes to use Least Square without lag dependent
> > > variable but with
> > > > > manual fixed-effects and with correction of the 
> standard errors: 
> > > > > -newey y x1 x2 ind*, lag(1)- Rodrigo.
> > > > 
> > > > That is what I meant.  -xtivreg2- will estimate least
> > > squares models
> > > > as well as IV models:
> > > > 
> > > > xtivreg2 y x1 x2, i(ind) bw(2)
> > > > 
> > > > or, using cluster-robust SEs,
> > > > 
> > > > xtivreg2 y x1 x2, i(ind) cluster(ind)
> > > > 
> > > > Cheers,
> > > > Mark
> > > > 
> > > > > 
> > > > > 
> > > > > ----- Original Message -----
> > > > > From: "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>
> > > > > To: <statalist@hsphsun2.harvard.edu>
> > > > > Sent: Thursday, March 30, 2006 3:44 PM
> > > > > Subject: st: RE: Re: Two way Fixed Effects & Autocorrelation& 
> > > > > Heteroskedasticity
> > > > > 
> > > > > 
> > > > > > Max, Rodrigo,
> > > > > >
> > > > > > Just a brief addendum to Rodrigo's 4th option:
> > > > > >
> > > > > >> -----Original Message-----
> > > > > >> From: owner-statalist@hsphsun2.harvard.edu
> > > > > >> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of 
> > > > > >> Rodrigo Alfaro
> > > > > >> Sent: 30 March 2006 21:28
> > > > > >> To: statalist@hsphsun2.harvard.edu
> > > > > >> Subject: st: Re: Two way Fixed Effects & Autocorrelation& 
> > > > > >> Heteroskedasticity
> > > > > >>
> > > > > >> Dear Max
> > > > > >>
> > > > > > <snip>
> > > > > >>
> > > > > >> (4) A way to solve the problem without a lag dependent
> > > variable
> > > > > >> is using a standard error correction that takes 
> care of the 
> > > > > >> behavior of the error term but does not change the
> > > Least Square
> > > > > >> estimation of the parameters. You can use
> > > > > >> -newey- with manually fixed effect, but you have 
> adjust the 
> > > > > >> degree of freedom (I wrote a command for that I can
> > > send you if
> > > > > >> you are interesting).
> > > > > >
> > > > > > -xtivreg2- will do this.  The fixed effects are handled
> > > > > automatically,
> > > > > > as is the dof adjustment.  However, you need
> > > respectable number of
> > > > > > periods for this to work, since the asymptotics require
> > > t to go to
> > > > > > infinity.  The alternative is cluster-robust standard
> > > > > errors, which are
> > > > > > robust to arbitrary autocorrelation and which will work
> > > > > with any number
> > > > > > of periods (since the asymptotics require only the
> > > number of cross
> > > > > > sections to go to infinity).  -xivreg2- will do this
> > > too.  -findit
> > > > > > xtivreg2- will find it for you.
> > > > > >
> > > > > > Cheers,
> > > > > > Mark
> > > > > >
> > > > > >>
> > > > > >> Rodrigo.
> > > > > >>
> > > > > >>
> > > > > >> ----- Original Message -----
> > > > > >> From: <kosak@gmx.de>
> > > > > >> To: <statalist@hsphsun2.harvard.edu>
> > > > > >> Sent: Thursday, March 30, 2006 11:11 AM
> > > > > >> Subject: st: Two way Fixed Effects & Autocorrelation& 
> > > > > >> Heteroskedasticity
> > > > > >>
> > > > > >>
> > > > > >> > My problem was already discussed before,but i 
> didn`t found
> > > > > >> an appropriate
> > > > > >> > solution to it.-
> > > > > >> > I have a big panel data set (1500 observations over 20
> > > > > >> years).I want to
> > > > > >> > estimate a model with time and group specific 
> fixed effects
> > > > > >> (Hausman Test
> > > > > >> > performed). The Model suffers from autocorrelation and 
> > > > > >> > heteroskedasticity.The problem is that the cluster option
> > > > > >> is not possible
> > > > > >> > for xtregar. Can i use xtgls or areg?As i 
> understood xtgls
> > > > > >> estimates a
> > > > > >> > random effet model.
> > > > > >> > Any suggestions?
> > > > > >> > Max
> > > > > *
> > > > > *   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/
> > > > 
> > > 
> > > --
> > > "Feel free" - 10 GB Mailbox, 100 FreeSMS/Monat ...
> > > Jetzt GMX TopMail testen: http://www.gmx.net/de/go/topmail
> > > *
> > > *   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/
> > 
> 
> --
> E-Mails und Internet immer und überall!
> 1&1 PocketWeb, perfekt mit GMX: http://www.gmx.net/de/go/pocketweb
> *
> *   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/



© Copyright 1996–2020 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index