Hausmann-Test Interpretation

[Lorenz Wagner]

Hey Guys, i just performed a Hausman and a Durbin-Wu-Hausman-Test on my Instrumental variable estimation.
This is my Code:

  **Hausman-Test
quietly regress lnscore experience nord west east city age age_sq
estimates store inconsistent
quietly ivregress 2sls lnscore (experience = vgebuehr) nord west east city age age_sq
estimates store consistent
hausman consistent inconsistent
  **Durbin-Wu-Hausman-Test *Ohne robuste Standardfehler
quietly regress lnscore experience nord west east city age age_sq
predict e, resid
regress lnscore experience nord west east city age age_sq e
test e

The first one returns a p-value of 0,99 the other one 0,25. How do I interpret these results

[Carlo Lazzaro]

Lorenz:
I would say that there's no evidence of endogeneity.

[Lorenz Wagner]

Lorenz:
I would say that there's no evidence of endogeneity.

Thanks. Does that mean that the instrument is exogenous or that there wasnt even a endogenity-problem to begin with?

[Carlo Lazzaro]

Lorenz:
probably there wasn't an endogeneity problem at the origin.

[Lorenz Wagner]

Lorenz:
probably there wasn't an endogeneity problem at the origin.

Thanks. Just one more question. Whenever i run a weak instrument test, i get an F-vaule of about 7,9. But when i introduce a second instrument it jumps to 55. Does that mean that both together are good instruments or that one is bad but the other is that good, that it increases the F-value by so much?

[Carlo Lazzaro]

Lorenz:
if you ran -estat firststage-, please share the full outcome. Thanks.

[Lorenz Wagner]

Lorenz:
if you ran -estat firststage-, please share the full outcome. Thanks.

[Carlo Lazzaro]

Lorenz:

Minimum eigenvalue statistic

?