Hi Mark - thanks for the confirmation about the interpretation of the
endogeneity test.
When I run
ivreg2 y (xstar=instrument1 instrument2) other variables, cluster(mag1)
I get no error message.
Here is the output from the two regressions. thanks for any help you or
anyone else can offer.
Daniel
xi:ivreg2 lncirc currentsite (=pubcurrentrivsites1 pubcurrentsites2)
pctcurrentrivsites, cluster(mag1)
OLS regression with robust standard errors
------------------------------------------
Number of clusters (mag1) = 537 Number of obs =
2377
F( 2, 536) =
3.28
Prob > F =
0.0384
Total (centered) SS = 4643.655397 Centered R2 =
0.0052
Total (uncentered) SS = 74845.78589 Uncentered R2 =
0.9383
Residual SS = 4619.631906 Root MSE =
1.4
------------------------------------------------------------------------------
| Robust
lncirc | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
currentsite | .2428465 .1044253 2.33 0.020 .0381768
.4475163
pctcurre~tes | -.1362768 .1446973 -0.94 0.346 -.4198783
.1473247
_cons | 5.361131 .0762595 70.30 0.000 5.211666
5.510597
------------------------------------------------------------------------------
Hansen J statistic (Lagrange multiplier test of excluded instruments):
78.584
Chi-sq(2) P-val =
0.00000
------------------------------------------------------------------------------
Instruments: pubcurrentrivsites1 pubcurrentsites2 currentsite
pctcurrentrivsites
------------------------------------------------------------------------------
. xi:ivreg2 lncirc (currentsite=pubcurrentrivsites1 pubcurrentsites2)
pctcurrentrivsites, cluster(mag1)
IV (2SLS) regression with robust standard errors
------------------------------------------------
Number of clusters (mag1) = 537 Number of obs =
2377
F( 2, 536) =
0.25
Prob > F =
0.7780
Total (centered) SS = 4643.655397 Centered R2 =
0.0051
Total (uncentered) SS = 74845.78589 Uncentered R2 =
0.9383
Residual SS = 4619.91795 Root MSE =
1.4
------------------------------------------------------------------------------
| Robust
lncirc | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
currentsite | .2160351 .5182878 0.42 0.677 -.7997903
1.23186
pctcurre~tes | -.1154665 .4289222 -0.27 0.788 -.9561385
.7252056
_cons | 5.365396 .1079534 49.70 0.000 5.153812
5.576981
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments):
77.261
Chi-sq(1) P-val =
0.00000
------------------------------------------------------------------------------
Instrumented: currentsite
Instruments: pubcurrentrivsites1 pubcurrentsites2 pctcurrentrivsites
------------------------------------------------------------------------------
> Daniel,
>
>> Hi - I have a couple of questions about ivreg2 and associated
>> endogeneity/exogeneity tests.
>>
>> 1. I want to assess whether a particular variable, xstar, is endogenous,
>> so I run:
>> ivreg2 y xstar (=instrument1 instrument2) other variables,
>> orthog(xstar).
>> the C-statistic allows me to assess endogeity of xstar. Is that correct?
>> This should give me the equivalent results as doing
>> ivreg y (xstar=instrument1 instrument2) other variables, followed by
>> ivendog star.
>> is that correct?
>
> Yes (though I think you mean -ivendog xstar-).
>
>>
>> 2. If so, then I am not sure why I am having the following problem.
>>
>> I run
>> ivreg2 y xstar (=instrument1 instrument2) other variables, orthog(xstar)
>> cluster(id).
>>
>> In other words, I have simply added the cluster option. When I do so,
>> stata
>> tells me:
>> Collinearity/identification problems in restricted equation: C statistic
>> not calculated for orthog option.
>>
>> I only have this problem when I include the cluster option on ivreg2.
>
> The C-stat works here by comparing two estimations, one with xstar
> exogenous and one with it endogenous:
>
> ivreg2 y xstar (=instrument1 instrument2) other variables
>
> ivreg2 y (xstar=instrument1 instrument2) other variables
>
> Something is going wrong with the latter estimation. Try estimating it
> explicitly with the cluster option and see what the error message is.
>
> Hope this helps.
>
> --Mark
>
>> what am I doing wrong? thanks. 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: [email protected]
> 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/
>
*
* 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/
