Hi,
I would really appreciate if you would help me on the following problem. I
have the following equation:
ivreg y (x1 x2 = z1 z2) X
that is, I use z1 to instrument x1 and z2 to instrument x2.
Now I want to test the endogeneity of x2. I am not worried about x1 at this
point.
One option is to use the argmented regression (Davidson and MacKinnon,
reference A-J page 541).
My question is: how to obtain the residuals from the first stage for x1 and
x2 after running the above -ivreg- ?
Another option, which I am not so sure is correct, is to do the following:
ivreg y (x1 x2 = z1 z2) X
est store myest1
ivreg y (x1 = z1) x2 X
est store myest2
test whether the coefficient of x1 in estimates myest1 and myest2 differ by
a basically t-stat.
My question: is the above test testing for whether I should model x2 as
endogenous? should I test for the entire set of coefficients?
Your generous help is very very much appeciated. I am grateful to your
attention and help!
Bo MacInnis
UC Berkeley
*
* 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/
