Dear statalisters
I am estimating a linear model with a potentially endogenous regressor using ivreg2. I have data on 40 countries over 25 years, and so are estiating using heteroskedasticity robust HAC standard errors.
My problem is that in the test of weak identification I get a Kleibergen-Paap Wald rk F statistic of 8.4, and so my instruments might be weak ( i have 3 excluded instruments). How much can I trust tests of endogeneity(of my potentially endogenous regressor) and overidentifying restrictions when my instruments are weak? And do you know of any way to make these tests robust to weak instruments?
Any help would be greatly appreciated
Siri
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
