EViews.com

Hi,

I currently have a panel dataset. I am using standard OLS (fixed-effects) estimation for my regressions but I believe there may endogeneity (reverse causality) between my dependent variable and explanatory variable. I understand there is a Hausman test to test for endogeneity but how can I do this? Also, if there is endogeneity, am I right in saying GMM must be utilised for IV regression?

Any help is much appreciated

Unfortunately EViews does not have the Hausman endogeneity test built in at the moment, however you should be able to perform it manually fairly easily since it involves running auxiliary regressions and using the residuals.

I was wondering if you could help.. I want to test if there is endogeneity in my fixed effects (country effects) model and was just wondering how I can go about this? Also, I'm not sure when I should use GMM if I need to carry out an IV regression?