Modeling Interdependent Consumer Preferences

Abstract

A consumer's preference for an offering can be influenced by the preferences of others in many ways, ranging from social identification and inclusion to the benefits of network externalities. In this article, the authors introduce a Bayesian spatial autoregressive discrete-choice model to study the preference interdependence among individual consumers. The autoregressive specification can reflect patterns of heterogeneity in which influence propagates within and across networks. These patterns cannot be modeled with standard random-effect specifications and can be difficult to capture with covariates in a linear model. The authors illustrate their model of interdependent preferences with data on automobile purchases and show that preferences for Japanese-made cars are related to geographically and demographically defined networks.