My question is in what ways the predicted values of a dependent variable are more accurate than the actual values. The issue is that I want to investigate the effect of X on Y.
In addition to X the right hand side contains other factors Z that influence both Y and X. I was recommended to use the predicted values of X based on a model where Zs are in the right hand side. Later, I use the predicted values of X to investigate its effect on Y.
Why this is a better way than using the actual values of X in an econometric wordings.