Zhehui,
Part of the answer is simple, namely that ivreg is reporting t-stats and
ivreg2 is reporting z-stats. This is also why the root MSEs are different;
ivreg is using a small sample correction and ivreg2 isn't. If you use
ivreg2 with the -small- option, the SEs and root MSE should coincide.
My guess is that the difference between ivreg2,gmm and ivgmm0 is
inconsequential and should disappear if you update both programs (the older
versions of ivgmm0 used an extra iteratation for the var-cov matrix), but
this is just a guess because you haven't given us version numbers in your
post. You can get these with the -which- command.
Yours,
Mark
Quoting Zhehui Luo <[email protected]>:
> Dear Listers,
>
> Can someone please tell me why ivreg, robust and ivreg2, robust give
>
> different results? eg, different Root MSE and standard errors, and
> why
> ivreg, robust did not report F-stat. Similarly, the ivgmm0 and
> ivreg2, gmm
> also give different results. What is the difference in their
> estimation
> methods? Thanks. (examples below)
>
> Zhehui
>
> . ivreg bmi93 edudegree91d* lgrproast191 , robust
>
> IV (2SLS) regression with robust standard errors Number of obs
> = 2487
> F(244,
> 2239) = .
> Prob > F
> = 0.0000
> R-squared
> = 0.3372
> Root MSE
> = 3.0303
>
> --------------------------------------------------------------------------
----
> | Robust
> bmi93 | Coef. Std. Err. t P>|t| [95%
> Conf. Interval]
> -------------+------------------------------------------------------------
----
> bmi91 | .8316739 .2122722 3.92 0.000 .415403
> 1.247945
> edudegre~1d2 | -.1254119 .21793 -0.58 0.565 -.552778
> .3019542
> edudegre~1d3 | -.2158311 .215216 -1.00 0.316 -.6378747
> .2062126
> edudegre~1d4 | -.2562037 .2153335 -1.19 0.234 -.6784779
> .1660706
> edudegre~1d5 | -.1086126 .2673033 -0.41 0.685 -.6328007
> .4155755
> edudegre~1d6 | -.0664432 .2827665 -0.23 0.814 -.6209551
> .4880686
> lgrproast191 | .0190294 .0373498 0.51 0.610 -.0542144
> .0922732
>
> . ivreg2 bmi93 edudegree91d* lgrproast191 , robust
>
> IV (2SLS) regression with robust standard errors
> ------------------------------------------------
>
> Number of obs
> = 2487
> F(244, 2186)
> =
>
3683797407504745900000000000000000000000000000000000000000000000000000000000
0000000000000000000000000000000000000000000000000000000000000000000000000000
0000000000000000000000000000000000000000000000000000000000000000000000000000
0000000000000000000000000000000000000000000000000000000000000000000000000000
00.00
> Prob > F
> = 0.0000
> Total (centered) SS = 31020.91547 Centered R2
> = 0.3372
> Total (uncentered) SS = 1215817.365 Uncentered R2
> = 0.9831
> Residual SS = 20559.79007 Root MSE
> = 2.87522
>
> --------------------------------------------------------------------------
----
> | Robust
> bmi93 | Coef. Std. Err. z P>|z| [95%
> Conf. Interval]
> -------------+------------------------------------------------------------
----
> bmi91 | .8316739 .2014106 4.13 0.000 .4369164
> 1.226431
> edudegre~1d2 | -.1254119 .2067789 -0.61 0.544 -.5306911
> .2798673
> edudegre~1d3 | -.2158311 .2042037 -1.06 0.291 -.616063
> .1844008
> edudegre~1d4 | -.2562037 .2043153 -1.25 0.210 -.6566542
> .1442469
> edudegre~1d5 | -.1086126 .2536258 -0.43 0.668 -.60571
> .3884848
> edudegre~1d6 | -.0664432 .2682978 -0.25 0.804 -.5922972
> .4594107
> lgrproast191 | .0190294 .0354387 0.54 0.591 -.0504291
> .0884879
>
>
>
> . ivreg2 bmi93 edudegree91d* lgrproast191 , gmm
>
> GMM estimation
> --------------
>
> Number of obs
> = 2487
> Total (centered) SS = 31020.91547 Centered R2
> = 0.2980
> Total (uncentered) SS = 1215817.365 Uncentered R2
> = 0.9821
> Residual SS = 21776.63206 Root MSE
> = 2.95909
>
> --------------------------------------------------------------------------
----
> | Robust
> bmi93 | Coef. Std. Err. z P>|z| [95%
> Conf. Interval]
> -------------+------------------------------------------------------------
----
> bmi91 | .7992068 .0894907 8.93 0.000 .6238082
> .9746055
> edudegre~1d2 | -.2873432 .2222355 -1.29 0.196 -.7229168
> .1482303
> edudegre~1d3 | -.3704279 .211556 -1.75 0.080 -.7850701
> .0442143
> edudegre~1d4 | -.2362584 .2377846 -0.99 0.320 -.7023075
> .2297908
> edudegre~1d5 | -.0311559 .2842094 -0.11 0.913 -.588196
> .5258842
> edudegre~1d6 | -.183481 .4743958 -0.39 0.699 -1.11328
> .7463177
> lgrproast191 | .0646309 .0317272 2.04 0.042 .0024468
> .126815
>
> . ivgmm0 bmi93 edudegree91d* lgrproast191
>
> Instrumental Variables Estimation via GMM Number of obs =
> 2487
> Root MSE
> = 2.9855
> Hansen J
> = 273.0648
> Chi-sq(56)
> P-val = 0.00000
> --------------------------------------------------------------------------
----
> | GMM
> bmi93 | Coef. Std. Err. z P>|z| [95%
> Conf. Interval]
> -------------+------------------------------------------------------------
----
> bmi91 | .7992069 .0896961 8.91 0.000 .6234058
> .9750079
> edudegre~1d2 | -.2873432 .1770643 -1.62 0.105 -.634383
> .0596965
> edudegre~1d3 | -.3704279 .1617584 -2.29 0.022 -.6874685
> -.0533873
> edudegre~1d4 | -.2362583 .1789966 -1.32 0.187 -.5870852
> .1145685
> edudegre~1d5 | -.0311559 .2170901 -0.14 0.886 -.4566446
> .3943328
> edudegre~1d6 | -.183481 .2393456 -0.77 0.443 -.6525897
> .2856278
> lgrproast191 | .0646309 .0290632 2.22 0.026 .0076681
> .1215937
>
> *
> * 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 +44-131-451-3008
email: [email protected]
web: http://www.sml.hw.ac.uk/ecomes
________________________________________________________________
DISCLAIMER:
This e-mail and any files transmitted with it are confidential
and intended solely for the use of the individual or entity to
whom it is addressed. If you are not the intended recipient
you are prohibited from using any of the information contained
in this e-mail. In such a case, please destroy all copies in
your possession and notify the sender by reply e-mail. Heriot
Watt University does not accept liability or responsibility
for changes made to this e-mail after it was sent, or for
viruses transmitted through this e-mail. Opinions, comments,
conclusions and other information in this e-mail that do not
relate to the official business of Heriot Watt University are
not endorsed by it.
________________________________________________________________
*
* 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/
