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]
st: RE: xtivreg
From
"Schaffer, Mark E" <[email protected]>
To
<[email protected]>
Subject
st: RE: xtivreg
Date
Mon, 7 Feb 2011 13:52:32 -0000
Lancine,
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> Lanciné Condé
> Sent: Monday, February 07, 2011 12:04 PM
> To: [email protected]
> Subject: st: xtivreg
>
> Hi
> I am currently working on the impact of the monetary union
> participationdecision. I want to estimate the following panel
> equation:
>
> y= x+cu+ e
>
> where cu is an endogeneous currency union dummy. I would like
> to instrument
> cu by another dummy called h.
> But, the following stata command does not work: xtivreg2 y x
> (cu=h),fe
> first
>
> The return result I receive is:
>
> xtivreg2 y x (cu=h), fd first
>
> equation not identified; must have at least as many instruments not in
> the regression as there are instrumented variables
> r(481);
>
> xtivreg2 y x (cu=h), fe first
>
> equation not identified; must have at least as many instruments not in
> the regression as there are instrumented variables
> r(481);
>
> The iv (h) variable is not recognised, but when I instrument with a
> continuous variable, as instrument, the equation works.
>
> The xtiverg is running, but it does not give the F and J
> statistics of the first step.
>
> I wanted to ask if it is forbitten to instrument a dummy by another
> dummy, or to use a dummy like instrument.
Instrumenting a dummy with a dummy can be legitimate.
My guess is that there is a collinearity problem here. Does the variable h vary over time? If it doesn't, then it will drop out when the within or first-differences transformations are applied. What happens if you do something like
xtreg h, fe
?
Also, when you estimate using official xtivreg, are you using the FE, FD or GLS estimators?
--Mark
>
> Could anyone help me to fix my described concern, and/or to get
> routines which can return all statistics associated to the xtivreg
> command, especially those relevant for the overidentification and the
> weak instruments?
>
> Best regards.
> CONDE
>
> *
> * 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/
>
--
Heriot-Watt University is a Scottish charity
registered under charity number SC000278.
*
* 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/