If I am looking at an
ivreg2 equation where:
ivreg2 y1 x2 x3 x4 (x1 = x5 x6)
and I know that
x6 are endogenous variables (i.e. they are motivated by latent, unmeasurable characteristics) and there are no other viable alternatives, does IV 2sls work? Is it biased? Inconsistent?