Deart Stata users,
I'm estimating the acces to : 0- Public beach, 1- Private beach and 2-semi private beach...
I first estimated a multinomial logit but found that it is not very robust due to the IIA hypothesis.
I'm wondering if it will sound better to estimate a conditional multinomial logit instead, taking the income as the variuable that may affect differently the acces to one of the beach.
Thanks for your help...
Modeste SENOU