st: re: xtoverid and ivreg2 version conflict in Stata 8.2

[Kit Baum <kitbaum@mac.com>]

Jacob writes

I’m running Stata 8.2 trying to run ‘xtivreg2’ and subsequently  
‘xtoverid’.
To do this, I first installed (all via ssc) ‘xtivreg2’ (version  
01.0.06) and
then ‘ivreg28’ (version 2.1.22). Running ‘xtivreg2’ produces the 2SLS
results as desired. Then, I installed ‘xtoverid’ (version 2.1.3) and  
tried
to run it after the 2SLS model. This returned the following error:  
“Error -
must have ivreg2 version 2.1.15 or greater installed”. So, I installed
‘ivreg2’ (version 2.2.08). Next, however, re-running the 2SLS with
‘xtivreg2’ and trying ‘xtoverid’ again gives this error: “unrecognized
command:  mata / Error - must have ivreg2 version 2.1.15 or greater
installed”.

There is a bit of confusion here with regard to the interaction of  
these routines. As presently available, -xtoverid- does require that  
the previous command is xtivreg, xtivreg2, xthtaylor or xtreg. But it  
also requires -ivreg2-, and will not accept -ivreg28- as a substitute.

However, it's not clear why Jacob wants to run -xtoverid- if he is  
successfully running -xtivreg2- (calling -ivreg28-). He cannot run - 
ivreg2- on version 8.2. because it requires Mata; so he should ado  
uninstall ivreg2. But the -xtivreg2- output includes a test of  
overidentifying restrictions by default. If the first example in  
'help xtivreg2' is run, the Sargan statistic for overidentification  
is produced. This is the same statistic that would be calculated in  
version 9 or 10 Stata using -xtoverid-. Perhaps -xtoverid- should  
also determine, as -xtivreg2- does, whether -ivreg2- or -ivreg238-  
should be invoked. But there is really no need for -xtoverid- if  
you're using -xtivreg2-.



Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.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/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/