st: logit with correlated residuals

["A D'Ignazio" <est1900@hotmail.com>]

Dear Statalist,

I have a dataset consisting of individual characteristics for N firms. For 
each pair of them, I also know whether they are in a business relationship 
or not.

I am modelling the likelihood of having a business relationship, for each 
pair, based on the firms individual characteristics (difference in size, 
etc...).

By using a logit, the results are presumably affected by the fact that the 
residuals are going to be correlated across observations, and standard 
errors could be underestimated.

As a first solution, I've added N-1 dummies (for each pair, I have two 
dummies assuming value 1) and estimated the new model with fixed effects.

Could you suggest me another way to address the problem at hand? For 
instance, how to cluster the residual around the firms identifiers in a 
logit model?

Many thanks!

Alex

_________________________________________________________________
Push the button! Crea il tuo blog e fatti vedere...����  
http://pushthebutton2006.spaces.live.com/

*
*   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/