An Unconditional Competing Risk Hazard Model of Consumer Store-Choice Dynamics

Abstract

In this paper an unconditional competing risk hazard model of consumer store-choice dynamics is developed and tested as an alternative to the negative binomial and Dirichlet models of store choice introduced in the urban planning literature by Wrigley and Dunn. The hazard model is less restrictive in terms of its assumptions regarding duration effects. It is also more flexible in that various distributions can be incorporated into the model, leading to different store choice dynamics. An empirical example, based on Nielsen scanner panel data for Springfield, MO, is provided to illustrate the modelling approach. Results indicate that the model represents the observed store-choice dynamics satisfactorily.