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

Re: st: xtivreg2


From   Daniel Simon <dhs29@cornell.edu>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: xtivreg2
Date   Wed, 16 Nov 2005 15:31:35 -0500

Mark - thanks for the suggestion. Dropping the cluster option did not change things, but when I then dropped the robust option, the problem goes away... unfortunately, I think that I need the cluster option for my data (panel data on magazines over time).

once again, any suggestions would be greatly appreciated. Daniel

At 08:16 PM 11/16/2005 +0000, you wrote:

Daniel,

Just to check - by "number of instruments" I meant both included and
excluded instruments.  But from your output below, it seems like you are
OK.

You could try running the estimation as below, but dropping the -cluster-
option and replacing it with -robust-, and with nothing.  This would help
tie down whether the number of clusters or something else to do with the
cluster-robust approach is or isn't the problem.

--Mark

> Mark - I think i'm ok with the number of instruments and the number of
> clusters. I get the following output:
>
> Number of clusters     N_clust     =        497
> Number of observations N           =       2393
> Number of regressors   K           =        239
> Number of instruments  L           =        244
> Number of excluded instruments L2  =          6
>
> If you have any other thoughts, or if I'm misunderstanding something,
> please let me know. Thanks. Daniel
>
> At 07:27 PM 11/16/2005 +0000, you wrote:
>>Daniel,
>>
>> > Hi - when I run xtivreg2 and include fixed effects other than those
>> > captured with the fe option,  I get the following message:
>> > Error: covariance matrix of moment conditions not of full rank;
>> > overidentification statistic not reported, and standard errors and
>> model
>> > tests should be interpreted with caution and may be meaningless
>> > Possible causes: singleton dummy variable (dummy with 1 one and N-1
>> zeros
>> > or visa-versa)
>> >
>> > the output then provides none of the overid/endog tests that I am
>> > interested in. Is there any way around this problem?
>> >
>> > Specifically, I run the following model, where i.groupyr are the fixed
>> > effects. When I do not include i.groupyr, the program runs fine.
>> >
>> > xi:xtivreg2 lncirc  currentsite (=pubonecurrentsites2 pubcurrentsites2
>> > pubcurrentsites2_sq pubonecurrentrivsites1 pubcurrentrivsites1
>> > pubcurrentrivsites1_sq) pgcurrentsites1 pgcirc1 i.groupyr , fe i(mag1)
>> > ffirst cluster(mag1) orthog (currentsite)
>>
>>You need to investigate the relationships between the fixed effects, your
>>extra fixed effects, and the cluster variable.  For example, you have to
>>have more clusters than instruments for the cluster-robust vcv to be full
>>rank.  This could be the problem if the number of groupyr fixed effects
>> is
>>big.
>>
>>Hope this helps.
>>
>>Cheers,
>>Mark
>>
>> >
>> >
>> > Thanks for any help. Daniel
>> >
>> >
>> > Daniel Simon
>> > Assistant Professor
>> > Department of Applied Economics and Management
>> > Cornell University
>> > (607) 255-1626
>> >
>> > *
>> > *   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/
>> >
>>
>>
>>Prof. Mark Schaffer
>>Director, CERT
>>Department of Economics
>>School of Management & Languages
>>Heriot-Watt University, Edinburgh EH14 4AS
>>tel +44-131-451-3494 / fax +44-131-451-3294
>>email: m.e.schaffer@hw.ac.uk
>>web: http://www.sml.hw.ac.uk/ecomes
>>
>>
>>
>>__________________________________________________________________
>>
>>DISCLAIMER:
>>
>>This e-mail message is subject to http://www.hw.ac.uk/disclaim.htm
>>__________________________________________________________________
>>
>>*
>>*   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/
>
> Daniel Simon
> Assistant Professor
> Department of Applied Economics and Management
> Cornell University
> (607) 255-1626
>
> *
> *   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/
>


Prof. Mark Schaffer
Director, CERT
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3294
email: m.e.schaffer@hw.ac.uk
web: http://www.sml.hw.ac.uk/ecomes



__________________________________________________________________

DISCLAIMER:

This e-mail message is subject to http://www.hw.ac.uk/disclaim.htm
__________________________________________________________________

*
*   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/
Daniel Simon
Assistant Professor
Department of Applied Economics and Management
Cornell University
(607) 255-1626
*
* 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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index