Econometrics help--is this multicollinearity or something else?

Question

Suppose we have our dependent variable, Y, and two groups of independent variables X and Z. Both X and Z are determinants of Y, but we also believe that Z might be a determinant of X ( but not the other way ). Is this a case of multicollinearity or is it something else? How can I address this?

Answer 1

The issue is that of endogeneity.

You can use granger causality to test the direction of causality to confirm the direction of causality if it exists. If there is an issue, then you may need to search for an instrument variable, orthognalize your explanitory variables with a two step model, or simultaineous system of eqns to ensure that your estimators are unbiased.