Question

(Econometrics | Instrumental Variables)

Consider the model:

college-BiprivateHS+u

Some researchers have suggested using an instrumental variables (IV) approach to estimating ß1. Specifically, a recent study suggested using the variable Catholic, which is equal to 1 if a student is Catholic and zero if not, as an instrumental variable for privateHS. What two conditions does Catholic have to satisfy in order to deliver consistent estimates of ß1? Which of these can be formally tested, and how? Be specific.

Answer 1

Solution:

To get the consistent estimator of  , the necessary conditions are as follows:

61

a) There should be zero covariance between instrumental variable (Catholic) and error term

b) And instrumental variable effect dependent variable only though regressor