Instrumental variable regression

Hi all,

I am an econ student doing some econometric analysis and was wondering how good is this fit:

The STATA command given was: ivreg x (y = z)
Where x is the output, y the treatment and z the instrument.

I am confused since the r-squared is extremely low but the p value is very high.
I am trying to prove causality between x and y (y causes x).

Any suggestions?


Not a robit
I dont see any significance and i see a negative adjusted R^2, see impossiple, but must be possible with a small corrected value.

Did you help eve you met IV assumptions?