Hi Kit:
Thanks for your note.
In fact, reg3 with the 2sls option provides precisely the same estimates
as does ivreg/ivreg2. Whether using the 2sls or default 3sls, the
cross-equation disturbance is estimated both in reg3 and ivreg2 (in an
exchange awhile back when discussing the endog option in ivreg2 I
establish that the cross-equation distrubance is estimated, remember?).
reg3 does the job of a limited information 2sls estimator when 2sls is
invoked in precisely the same way as does ivreg2, so I sincerely hope
you don't disable the option to use overid with other estimators, like
2sls in reg3. In fact, the old overid provided precisely the same overid
statistic (and of course, estimates) as does the ivreg2 routine. I think
that you may have changed something in that regard in the new overid
code. For example:
. use http://fmwww.bc.edu/ec-p/data/hayashi/griliches76.dta
. xi i.year
. ivreg2 lw s expr tenure rns smsa _I* (iq=med kww age mrt)
IV (2SLS) estimation
--------------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs
= 758
F( 12, 745) =
45.91
Prob > F =
0.0000
Total (centered) SS = 139.2861498 Centered R2 =
0.4255
Total (uncentered) SS = 24652.24662 Uncentered R2 =
0.9968
Residual SS = 80.0182337 Root MSE =
.3249
------------------------------------------------------------------------------
lw | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
iq | .0001747 .0039035 0.04 0.964 -.007476
.0078253
s | .0691759 .0129366 5.35 0.000 .0438206
.0945312
expr | .029866 .0066393 4.50 0.000 .0168533
.0428788
tenure | .0432738 .0076271 5.67 0.000 .0283249
.0582226
rns | -.1035897 .029481 -3.51 0.000 -.1613715
-.0458079
smsa | .1351148 .0266573 5.07 0.000 .0828674
.1873623
_Iyear_67 | -.052598 .0476924 -1.10 0.270 -.1460734
.0408774
_Iyear_68 | .0794686 .0447194 1.78 0.076 -.0081797
.1671169
_Iyear_69 | .2108962 .0439336 4.80 0.000 .1247878
.2970045
_Iyear_70 | .2386338 .0509733 4.68 0.000 .1387281
.3385396
_Iyear_71 | .2284609 .0437436 5.22 0.000 .1427251
.3141967
_Iyear_73 | .3258944 .0407181 8.00 0.000 .2460884
.4057004
_cons | 4.39955 .2685443 16.38 0.000 3.873213
4.925887
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic):
52.436
Chi-sq(4) P-val =
0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic):
13.786
Stock-Yogo weak ID test critical values: 5% maximal IV relative bias
16.85
10% maximal IV relative bias
10.27
20% maximal IV relative
bias 6.71
30% maximal IV relative
bias 5.34
10% maximal IV size
24.58
15% maximal IV size
13.96
20% maximal IV size
10.26
25% maximal IV
size 8.31
Source: Stock-Yogo (2005). Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments):
87.655
Chi-sq(3) P-val =
0.0000
------------------------------------------------------------------------------
Instrumented: iq
Included instruments: s expr tenure rns smsa _Iyear_67 _Iyear_68 _Iyear_69
_Iyear_70 _Iyear_71 _Iyear_73
Excluded instruments: med kww age mrt
------------------------------------------------------------------------------
. reg3 (lw =iq s expr tenure rns smsa _I*) (iq=med kww age mrt s expr
tenure rns smsa _Iyear
> _67 _Iyear_68 _Iyear_69 _Iyear_70 _Iyear_71 _Iyear_73)
Three-stage least-squares regression
----------------------------------------------------------------------
Equation Obs Parms RMSE "R-sq" chi2 P
----------------------------------------------------------------------
lw 758 12 .3249076 0.4255 560.57 0.0000
iq 758 15 11.09542 0.3354 384.36 0.0000
----------------------------------------------------------------------
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
lw |
iq | .0001747 .0039035 0.04 0.964 -.007476
.0078253
s | .0691759 .0129366 5.35 0.000 .0438206
.0945312
expr | .029866 .0066393 4.50 0.000 .0168533
.0428788
tenure | .0432738 .0076271 5.67 0.000 .0283249
.0582226
rns | -.1035897 .029481 -3.51 0.000 -.1613715
-.0458079
smsa | .1351148 .0266573 5.07 0.000 .0828674
.1873623
_Iyear_67 | -.052598 .0476924 -1.10 0.270 -.1460734
.0408774
_Iyear_68 | .0794686 .0447194 1.78 0.076 -.0081797
.1671169
_Iyear_69 | .2108962 .0439336 4.80 0.000 .1247878
.2970045
_Iyear_70 | .2386338 .0509733 4.68 0.000 .1387281
.3385396
_Iyear_71 | .2284609 .0437436 5.22 0.000 .1427251
.3141967
_Iyear_73 | .3258944 .0407181 8.00 0.000 .2460884
.4057004
_cons | 4.39955 .2685443 16.38 0.000 3.873213
4.925887
-------------+----------------------------------------------------------------
iq |
med | .2721714 .159856 1.70 0.089 -.0411405
.5854834
kww | .4427703 .0691238 6.41 0.000 .3072901
.5782505
age | -.9913636 .2201974 -4.50 0.000 -1.422943
-.5597846
mrt | -.9452683 .9319855 -1.01 0.310 -2.771926
.8813898
s | 2.624565 .2823757 9.29 0.000 2.071119
3.178011
expr | .0564757 .25055 0.23 0.822 -.4345933
.5475448
tenure | .690506 .2700872 2.56 0.011 .1611447
1.219867
rns | -2.581106 .9398181 -2.75 0.006 -4.423116
-.7390961
smsa | .0621573 .9124468 0.07 0.946 -1.726205
1.85052
_Iyear_67 | .7564285 1.638293 0.46 0.644 -2.454566
3.967423
_Iyear_68 | .2275306 1.557616 0.15 0.884 -2.82534
3.280401
_Iyear_69 | 2.026613 1.505147 1.35 0.178 -.9234201
4.976646
_Iyear_70 | 5.605308 1.677934 3.34 0.001 2.316619
8.893998
_Iyear_71 | 5.213041 1.545298 3.37 0.001 2.184313
8.241768
_Iyear_73 | 4.444228 1.467061 3.03 0.002 1.56884
7.319616
_cons | 68.58568 4.056543 16.91 0.000 60.63501
76.53636
------------------------------------------------------------------------------
Endogenous variables: lw iq
Exogenous variables: s expr tenure rns smsa _Iyear_67 _Iyear_68 _Iyear_69
_Iyear_70 _Iyear_71 _Iyear_73 med kww age mrt
------------------------------------------------------------------------------
. overid
Number of equations : 2
Total number of exogenous variables in system : 16
Number of estimated coefficients : 29
Hansen-Sargan overidentification statistic : 87.655
Under H0, distributed as Chi-sq(3), pval = 0.0000
. which overid
H:\stata\ado\plus\o\overid.ado
*! overid V2.0.3 4Feb2007
*! Authors C F Baum, Vince Wiggins, Steve Stillman, Mark Schaffer
Regards,
John.
____________________________________________________
Prof. John Antonakis
Associate Dean Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland
Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305
Faculty page:
http://www.hec.unil.ch/people/jantonakis
Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________
On 24.11.2009 04:17, Kit Baum wrote:
> <>
> John used the 2sls option on reg3 on a two-equation system and commented:
>
> What is a little odd, though, is that the Hansen Sargan statistic
> returned by overid (using a simple two-equation model with 2SLS) is not
> the same as ivreg2:
>
>
> Not really odd, in that the overid statistic from reg3 (even if you
have estimated the model ignoring cross-equation correlation) is
calculated for the entire system of two equations and five coefficients,
rather than the single equation containing only two of those
coefficients. As the -overid- help file describes, the overid stat for
3SLS takes account of the entire system estimates. Perhaps we should
disable its use if you use ols, sure, 2sls or mvreg options in reg3, as
it is not obvious that it is appropriate in those circumstances. It is
meant to be used after 3sls estimation.
>
> Kit
>
> Kit Baum | Boston College Economics & DIW Berlin |
http://ideas.repec.org/e/pba1.html
> An Introduction to Stata Programming
| http://www.stata-press.com/books/isp.html
> An Introduction to Modern Econometrics Using Stata |
http://www.stata-press.com/books/imeus.html
>
>
> *
> * 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/
