[Limdep Nlogit List] Endogeniety and Mixed logit model

[sshreay at mail.wsu.edu]

Hello Fred,
Thanks for suggesting this article. It is quite useful. I too used an IV
regression first (using different week price from different stores as
instrument) and then used predicted price in mixed logit model. I was just
worried about Standard error robustness using this way. Unfortunately, I
do not have any information on supply side. LIMDEP does give me
elasticities using mixed logit model. I am just looking some ways to
defend endogeneity of price issue?. Thanks a lot.

Sincerely,
Sanatan Shreay
> You may wish to have a look at this excellent review paper, "Structural
> Modeling in Marketing: Review and Assessment," by Chintagunta, Erdem,
> Rossi
> and Wedel. It has already come out in a journal, but I believe it's still
> available in draft form here:
> http://papers.ssrn.com/sol3/papers.cfm?abstract_id=588903. And it is
> available in final form here:
> http://mktsci.journal.informs.org/cgi/content/abstract/25/6/604.
>
> Fred
>
> =====
>
> Fred Feinberg
> Hallman Fellow and Professor of Marketing
> Ross School of Business
> University of Michigan
> feinf at umich.edu
>
> sshreay at mail.wsu.edu wrote:
>
>> Dr Green,
>> Thank you for your reply and time.
>> Sincerely,
>> Sanatan Shreay
>>
>> > Dear Sanatan. The Berry Levinsohn-Pakes methodology that Aviv Nevo is
>> > writing about
>> > is a (maybe the) way to handle endogenous prices in a logit model for
>> > aggregate
>> > shares.  That I know of, however, the method has not been extended to
>> a
>> > general
>> > random parameters specification.  The BLP method is not yet build into
>> > LIMDEP,
>> > unfortunately.
>> > /B. Greene
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>