I now have a question on the endogeneity test with -ivreg2-. I know that
one can use the Hausman test. i.e. running the regression first with the
instruments, then running the regression treating all variables as
exogenous and then doing the Hausman test.
I now read in an earlier posting that it is also possible to use the
-orthog- option to test for endogeneity, i.e. if I suspect y2 to be
endogenous:
ivreg2 y1 y2 ( = x4 x5) x1 x2 x2, orthog(y2)
Is this procedure of testing for endogeneity of y2 equivalent to using
the
Hausman test?
