Dear sir or madam:
But I have some question about the mediation effect and endogenous: Is there a tradeoff between Collinearity and endogenous? For example, If we should control the lag dependent variable in panel data model.
Thank you for your time and consideration!
Best wishes for you