st: orthogonal regression

[alvaro <aa314@cam.ac.uk>]

Dear Dr Baum,

Thank you so much for your previous answer and especially for the 
point on the interpretation of the procedure.

One of the problems I see in this solution is that in the first model 
iv1 is not independent of the errors (it is just the ranking for y1) 
and similarly for iv2 in the second model.

Having thought about it, I guess a question which remains is why 
should orthogonal regressions depend on the choice of instruments. If 
it is just a procedure which minimizes a distance to the fitted line, 
why is it that I need to define instruments before having the 
estimation results. Isn't it possible to estimate this method not 
having to define instruments? Parameters change a lot with different 
instruments.


Thank you very much, again.


All the best,

alvaro

*
*   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/