Hi,

I have a reverse causation system. Y1 = f(Y2, X) and Y2 =g(Y1, Z). Using IV both reverse casuation coefficients (Y2 in Y1 and Y1 in Y2) are statisitically signfcant. However, when I apply the Hausman test to the Y1 equation it is insignficant and suggests no endogenity issue. Does this make sense? Any explanation with supporting literarure would be appreciated .

Cheers, Eddie