st: AW: cmp vs cdsimeq to estimate simultaneous equation

["Meulemann Max" <mmeulemann@ethz.ch>]

They do different things. As far as I understand cmp only estimates the correlation of the errors of both equations which corrects the estimates for your observed values of Y2, as explained on p 18 of "Estimating Fully..." Rodman 2009. 

There is an older stata list post about this when both endogenous variables are binary http://www.stata.com/statalist/archive/2009-04/msg00073.html
Maybe this helps to understand.

I think cdsimeq really instruments for Y2 as probably explained here (http://www.polsci.ucsb.edu/faculty/glasgow/research/PA2000.pdf) or in the sources of the help file. 

Good luck 



-----Ursprüngliche Nachricht-----
Von: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Singh, Harpreet
Gesendet: Montag, 20. Januar 2014 07:03
An: statalist@hsphsun2.harvard.edu
Betreff: st: cmp vs cdsimeq to estimate simultaneous equation

I have system of equations in which one dependent variable is binary and other is continuous.

My equations are something like this :
Y2 = Y1 + X+Z1
Y1= Y2 +X +Z2 

Here Y1 is continuous and Y2 is binary. Z1 and Z2 are instrument variables. 

Shall I use cdsimeq or cmp to estimate these system of equations ? 

I am getting different results with these two commands and not sure which one is better.

I will really appreciate any suggestions.



*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/