st: problem with testing for endogeneity

[Cordula Stolberg <cs32@sussex.ac.uk>]

Dear Statalisters,

I have an unbalanced fixed effects panel and have problems testing for 
endogeneity. My regression looks like this (I used pooled OLS and added the 
fixed effects dummies manually, as there are only four fixed effects 
dummies):

y = b1X1 b2X2 b3dummies

I suspect that X2 might be endogenous and depend on Y. I don't know how to 
test for that. I have tried the Durbin-Wu-Hausman test by first running a 
regression:

X2 = c1Y  c2dummies

predict X2_res, res

and then I tried to run the regression

Y = b1X1 b2X2 b3dummies X2_res

but that didn't work (no standard error etc. reported).

I also tried -ivreg- with

ivreg Y b1X1 b2X2(=Y) b3dummies

but that didn't work either, as I got the error message that too few 
variables were specified.

I would be really grateful for suggestions what to do about this!

Thanks,
Cordula

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/