Multiple Instrumental Variable

[Giorgio Depolli]

Hey everyone,

I have the following equation:

y=aX1+bX2+cX3+dX4+e

I want to instrument Z1 for X1 and Z2 for X2. Is this command appropriate?

ivreg2 y (X1 X2= Z1 Z2) X3 X4

The results show that instruments pass under identification test and are strong instrument. Is shea R^2 relevant in this case? What is considered as low shea R^2?

Thank you for your time.

[Joro Kolev]

Yes, the command you use is appropriate.

Show the output of your regression, and we can think through what it means.

Shea's R-squared is an R-squared as the name suggests. There is no good answer to the question of "what is considered a low R-squared" in any context and for any R-squared.

Hey everyone,

I have the following equation:

y=aX1+bX2+cX3+dX4+e

I want to instrument Z1 for X1 and Z2 for X2. Is this command appropriate?

ivreg2 y (X1 X2= Z1 Z2) X3 X4

The results show that instruments pass under identification test and are strong instrument. Is shea R^2 relevant in this case? What is considered as low shea R^2?

Thank you for your time.