A generalization of the multinomial logit (MNL) model is developed for cases where discrete alternatives are ordered, by allowing stochastic correlation among alternatives in close proximity. The model belongs to the Generalized Extreme Value class and is therefore consistent with random utility maximization. An extension can handle cases where observations have been selected from a truncated choice set. A two-stage procedure using MNL computer software provides a specification test for MNL against the proposed model. Two empirical applications are briefly described. Copyright 1987 by The Econometric Society.
Stored only when both papers are in the corpus — 345 citations exist upstream.