ALTERNATIVE NESTED LOGIT-MODELS - STRUCTURE, PROPERTIES AND ESTIMATION

Two distinctly different nested legit models have been widely used in both research and applications. The differences, not widely recognized , between these models will substantially influence estimation results , behavioral interpretation and policy analysis. The McFadden nested l egit model is derived from random utility theory; the Daly or non-norm alized nested legit model is based on probability relationships and is not consistent with utility maximization. This paper describes and co mpares the model structure and properties of these different nested le git models identifying important differences between the different mod el structures. An empirical application demonstrates that the differen t nested legit models produce dramatically different results with resp ect to nesting structure and the relative importance of utility compon ents. Thus, the selection of one or another of the nested legit models has important consequences for model interpretation and prediction wi th consequent impacts on policy analysis. The authors prefer the McFad den model because of its basis in utility theory, intuitively reasonab le elasticity relationships and a clear interpretation of utility func tion parameters across alternatives. (C) 1998 Elsevier Science Ltd. Al l rights reserved.