RE: st: RE: Cragg-Donald Wald Statistic and Stock and Yogo Critical values when there are more than 2 endogenous variables that are likely weak

 From "Volpi, Massimiliano" To "statalist@hsphsun2.harvard.edu" Subject RE: st: RE: Cragg-Donald Wald Statistic and Stock and Yogo Critical values when there are more than 2 endogenous variables that are likely weak Date Thu, 11 Jun 2009 13:54:39 +0100

Dear Statalisters

I take up again this thread, as I was wondering what kind of solution could work, when there are more than 3 endogenous variables.

In a previous answer to this thread, Kit Baum pointed out that xtivreg2 could deal with 3 instruments, at most (while all usual tests remain available at least).
As I moved from having 3 to 4 variables that might potentially need being instrumented, I am wondering if using reg3 could solve the issue (estimating a system of equation instead of a single equation).

I assume that it would be necessary to check that all four equation are correctly specified, to be able to defend the specification with reg3, but haven't got a chance to really delve into this estimation procedure.

So I am asking for the experts' judgement. Is this a meaningful possibility, worth pursuing and reading more on, or is it just non-sense?
If meaningful, can I ask you for advice on where I should start reading from, for a practical implementation? The literature on the matter seems so large I am a bit "scared"/overwhelmed.

A closely related question: would it make sense to use xtivreg2 to test whether indeed all 4 variables really need being instrumented or not? If I could prove this way that only 3 variables are endogenous, while the fourth is exogenous, then I am fine (and lucky) and I don't need to look for any other solution. Otherwise, the answer to the above really becomes critical to my work.

Thanks and Best Regards

Max Volpi

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Stephen Armah
Sent: 02 June 2009 11:17
To: statalist@hsphsun2.harvard.edu
Subject: Re: st: RE: Cragg-Donald Wald Statistic and Stock and Yogo Critical values when there are more than 2 endogenous variables that are likely weak

Dear max,
Thanks so much for the prompt response.  however I believe ivreg2 and
xtivreg2 both do not  report Cragg-Donald Wald Statistic and Stock and
Yogo Critical values when there are " more than two endogenous
variables" that are likely weak.  This may be because Stock and Yogo
only worked with two endogenous variables.  I might be wrong.  Your
help is appreciated

I read Stock Wright and Yogo (2002) "A Survey of Weak Instruments and
Weak Identi' cation in Generalized Method of Moments". Journal of
Business & Economic Statistics October 2002, Vol. 20, No. 4   but I am
still not sure for the  more than two endogenous variables case.

Thanks
and be well

--
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.

*
*   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/