Endogeneity in data panel
Good afternoon in advance I appreciate your collaboration, I'm working with a short panel data (n> T), in the specification of the error component (a) corresponds to unobservable heterogeneity and (e) idiosyncratic error. When using the haussman test with the option "sigammore and sigmaless" it is convenient to use fixed effects (EF), I understand that when using EF some degree of correlation between (a) and the independent variables is accepted affirming a topic of endogeneity but only With unobservable heterogeneity, my question is:
1. In relation to the Hausman test when I use it without the options of "sigammore and sigmaless" it affirms the use of random effects instead of fixed effects.
2. I do not know how to identify (I need some command) in panel data that endogeneity exists in a general way, that is, that this correlates the idiosyncratic error with the independent variables.
Thanks
1. In relation to the Hausman test when I use it without the options of "sigammore and sigmaless" it affirms the use of random effects instead of fixed effects.
2. I do not know how to identify (I need some command) in panel data that endogeneity exists in a general way, that is, that this correlates the idiosyncratic error with the independent variables.
Thanks
