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]
RE: st: xtnbreg Initial values
From
Rodolphe Desbordes <[email protected]>
To
"'[email protected]'" <[email protected]>
Subject
RE: st: xtnbreg Initial values
Date
Wed, 12 May 2010 10:01:03 +0100
Dear Ivan,
For my part, all credit should go to João M. C. Santos Silva and Silvana Tenreyro.
Rodolphe
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Ivan Png
Sent: mercredi 12 mai 2010 08:30
To: [email protected]
Subject: Re: st: xtnbreg Initial values
Many thanks, Rodolphe and Maarten.
The paper by Santos-Silva and Tenreyro explains when max likelihood
estimation of Poisson or similar specifications might fail. Desbordes
below provides a way to identify the troublesome explanatory variables
and exclude them.
In my case, the solution was simpler. Run OLS and get the
coefficients and use them as starting values in the xtnbreg
Specifically,
----
reg depvar indepvars
matrix _s=e(b)
xtnbreg depvar indepvars , fe vce(oim) iterate(20) from(_s, skip)
Once again, many thanks to STATA-listers!
On Fri, May 7, 2010 at 5:03 PM, Rodolphe Desbordes
<[email protected]> wrote:
> Ivan,
>
> Are you referring to this strategy?
>
> http://cep.lse.ac.uk/pubs/download/dp0932.pdf
>
>
> If that is correct, see http://privatewww.essex.ac.uk/~jmcss/panelpoisson.do
>
> Rodolphe
>
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Ivan Png
> --
> Skype: ipng00
> T: +65 6516 6807
>
*
* 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/