[Limdep Nlogit List] Welfare Measures in the Error Components ML model with Heteroscedastic Error Components

[Mikołaj Czajkowski]

Dear All,

I find it difficult to derive welfare measures for the Error Components 
ML model. I suspect that the usual Hanemann's (1982) formula is not 
enough in this case, since the alternatives are not independent. The 
probability of choosing a particular alternative is known, however I 
don't see how a change in expected maximum utility would simplify to the 
usual log-sum expression in this case. Would it?

I searched through the literature, I was not able to find the formula 
for CV derived explicitly for the case of EC-MNL. Could you possibly 
suggest some papers on this or suggest a way to deal with this problem?

I will greatly appreciate any comments / suggestions.

Best regards,

-- 
     Mikołaj Czajkowski

     Warsaw Ecological Economics Center
     University of Warsaw
     http://www.woee.pl/