[Limdep Nlogit List] System ML with Ordered Probit

[David Tufte]

Thanks for the program Dr. Greene - this is exactly the sort of 
bridge I needed!

One question though: in this part of the program:
maximize
;start = beta0,mu0,a0,s0,r0
;labels= beta1,beta2,beta3,beta4,mu1,a1,a2,a3,sgma,rhoeu
;fcn   = eps = h - a1'z |
          dr = 1/sqr(1 - rhoeu*rhoeu) |
          index = dr*(beta1'x - rhoeu/sgma*eps) |
          p0 =     phi(       - index)      |
          p1 =     phi(mu1*dr - index) - p0 |
          p2 = 1 - phi(mu1*dr - index)      |
          log(y0*p0 + y1*p1 + y2*p2)
          + log(n01(eps/sgma)/sgma)
          ; maxit=10 ; output=3  $
          ; fix=rhoeu $
What is the purpose of the last line (which won't run by itself as written)?

Is this just to show how to code the identifying restriction that 
rhoeu=0 if I need to include it (i.e., in a two-step estimation the 
residual from the first step can be included in the second step 
without the problem being obvious as in Barro [1977], but if you're 
going to estimate a system with the residual from the one equation as 
an independent variable in the second Pesaran [1987] showed that you 
must restrict rhoeu=0 to achieve identification)?



Regards,

David Tufte
Associate Professor
Department of Economics and Finance
School of Business
Southern Utah University
Cedar City, UT 84720

http://www.suu.edu/faculty/tufte/