[Limdep Nlogit List] Welfare Measures in the Nested Logit Model with Covariance heterogeneity

[Mikołaj Czajkowski]

Dear All,

I am using the Nested Logit Model with Covariance Heterogeneity and I 
would like to estimate welfare measures.

Unfortunately the only paper explicitly deriving welfare measures for 
the nested logit model which I was able to identify is:
- Kling CL, Thomson CJ (1996) 'The Implications of Model Specification 
for Welfare Estimation in Nested Logit Models'. American Journal of 
Agricultural Economics 78(1): 103-114

I have summarized the approach in the attachment, since the formula is 
too complicated to write it inline.

My questions are as follows:

1. Could anyone verify if the formula is ok?

I would also appreciate some suggestions on papers / books where I could 
see how to derive welfare measures for the nested logit model. I 
understand that the utility of each alternative is weighted by the 
probability of selecting it, but I can't seem to be able to derive it 
myself.

2. In case of the nested model with covariance heterogeneity - if I am 
interested in the mean CV, should I use individual-specific or sample 
mean values of attributes explaining inclusive value parameters?

I have specified the CV formula using the WALD command, however I am not 
sure whether variables or their means should be used as covariates of 
inclusive value parameters. Both approaches seem to give different results.

I would appreciate all the comments you might have. Thank you very much 
in advance.

Best regards,


-- 
     Mikołaj Czajkowski

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