st: Re: nonlinear constrained estimation and "makecns"

[Kit Baum <baum@bc.edu>]

I don't think this is the appropriate way to deal with this. By  
pooling the data in these two equations you are assuming that there  
is a single error process. Dealing with them as SUR equations,  
without constraints, would clearly not result in equal sigma^2. Even  
though two of the equations' coefficients are forced to be  
proportional by the constraints, the third is not, so I would think  
that the imposition of a single error variance implied by this method  
would be an inappropriate restriction on the DGP.


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


On Apr 28, 2008, at 02:33 , Maarten wrote:

> Actually, I should have seen that I have already done something  
> that is
> very close in the -propcnsreg- package, see -findit propcnsreg-. You
> need to create a singly y variable, and a dummy variable indicating
> whether the a record is refering to y1 or y2 (call this dummy variable
> D). Than you type:
>
> propcnsreg y D X3, constrained(X1 X2) lambda(D)
*
*   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/