Hausman Endogeneity Test
Posted: Tue Apr 07, 2009 4:28 am
Hi,
I wanted to run the hausman endogeneity test manually as outlined in the eviews helpfile by regressing the endogenous variable on exogenous regressors including a valid instrument. I am using panel data so OLS fixed effects. I was just wondering when I find the residuals from the first OLS regression.. do I just add as an additional regressor into the OLS fixed effects model or does the "second stage" have to be normal OLS? Also, how can I test if my instrument I am using for this Hausman test is strong?
Any help would be appreciated..
Thanks
I wanted to run the hausman endogeneity test manually as outlined in the eviews helpfile by regressing the endogenous variable on exogenous regressors including a valid instrument. I am using panel data so OLS fixed effects. I was just wondering when I find the residuals from the first OLS regression.. do I just add as an additional regressor into the OLS fixed effects model or does the "second stage" have to be normal OLS? Also, how can I test if my instrument I am using for this Hausman test is strong?
Any help would be appreciated..
Thanks