A simple OLS estimation provide expected marginal ceteris paribus effects, i.e. expected partial derivatives of the dependent variable wrt a regressor. You don't need the regressors to be mutually independent, just that regressors are independent from your error term (you may need something even weaker, E(xu)=0). I think any basic econometrics textbook can offer a discussion about this point.
However, without further information it is difficult to be sure that your theoretical statements actually fit your data.
Benjamin
-----Mensaje original-----
De: [email protected] [mailto:[email protected]] En nombre de Gordon
Enviado el: Monday, September 29, 2008 2:29 PM
Para: [email protected]
Asunto: st: should I concern the endogeneity problem
Hi,
I am estimating a model:
Y = X1 + X2 + X1*X2
I am confident that X1 and X2 are exogenous to Y, however it is likely
that X1 is influenced by X2. In this case, should I worry about any
endogeneity?
Thanks for your insights!
Gordon
*
* 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/
*
* 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/
