st: Panel data estimator for simultaneous equation model

[john minor <spikefire87@live.com>]

I am using a simultaneous equations model with panel data that has the following form:

(1) Y1=a0+a1*Y2+a2*Y3+a3*K+E1,

(2) Y2=b0+b1*Y1+b2*Y3+b3*L+E2,

(3) Y3=co+c1*Y1+c2*Y2+c3*M+E3

K, L, and M are other exogenous variables and E1, E2, and E3 are independent errors. Y2 and Y3 are the
regressors in (1), Y1 and Y2 are the regressors in (2), and  Y1 and Y2 are the regressors in (3).  

I was considering 3 stage least squares for this estimation, but was looking into other estimators such as GMM. A more specific type of gmm I might be interested in is a xtabond type estimator due to the addition of additional error corrections such as Windmiejer. Would any of these estimators be appropriate? 

Another question I have was: Once I decide on an estimator is there a way to handle individual fixed effects in this simultaneous equation model?

I have used simultaneous equations with 2 equations before, but not 3 which has lead to my doubt about which estimator to use.

Thanks,

John Minor 		 	   		  
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/