In stata 12, do nlogit and nlogitrum estimate the same model. I understand before stata 10, nlogit estimates the non-normalized nested logit. That one is not consistent with utility maximization . In the manual for nlogit in stata 12, it says now the command estimates RUM model. So I am assuming the results produced by nlogit and nlogitrum should be same.
Do I understand this problem correctly.
