[Limdep Nlogit List] Endogeneity test in case of Bivariate

rhythm Em rhythmem at yahoo.com
Thu Oct 22 02:58:31 EST 2009


Dear all,

I have following general set up:
z = f (x1, x2, x3, y1, y2)

I suspect: y1 and y2 are endogenous. They are binomial (0 or 1) variables.

y1 and y2 are adoption decisions and are related to each other so, I run bivariate probit: 
bivariate; Lhs= y1,y2 ; rh1= x1, x2, x3 ; rh2= x1, x2, x3 ;keep=predy1 ; res= pred y2 $ (first stage)

Please suggest me how to test that y1 and y2 are endogenous or not--- what to plug in OLS (second stage) z= f(.............,....,...) to test it. If you know the procedure of testing endogeniety in bivariate case in LIMDEP, please suggest me. Would you please suggest me if there is related reference regarding this. I really appreciate your help. Thanks.

Regards,
Rhythm


      


More information about the Limdep mailing list