st: ivreg2, ivreg29 and ranktest updates

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Hi all.  This is just a note to say that updates to -ivreg2-, -ivreg29-
and -ranktest- by Baum/Schaffer/Stillman and Kleibergen/Schaffer are now
available from ssc-ideas in the usual way (with thanks as aways to Kit
Baum).

The updates are minor bug fixes that address various issues, some of
which were raised on the list recently by Umid Aliev and John Antonakis.
Full details are inside the -ivreg2- and -ranktest- ado files.  The bug
fixes worth noting:

- When presented with reduced form eqns involving very large matrices,
-ivreg2- would sometimes have trouble generating first-stage stats
(John's problem).  This was partly because of an inefficient and
unnecessary (but not wrong) calculation that is now fixed.

- If data were tsset on one variable and cluster-robust SEs on a
different variable were requested, the detailed first-stage statistics
could be wrong.  This is now fixed.  (I doubt anyone actually
encountered this, because the detailed first-stage stats affected by the
bug would contradict the correct weak and underidentification stats in
the main -ivreg2- output and the problem would be obvious.)

- Automatic bandwidth selection with panel data has been disabled.  This
was already done for -xtivreg2-, but we added it for -ivreg2- as well.

- The calculation of the Stock-Wright S statistic in -ivreg2- and
-ivreg29- has been modified (e.g., it is now invariant to
partialling-out with the -partial- option).

Cheers,
Mark

Prof. Mark Schaffer FRSE
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3296
http://ideas.repec.org/e/psc51.html
 


-- 
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/