(3) Demand Functions: From Linear Demand Functions to Spatial Utility Function-based Demand Models. Due to their high mathematical tractability, linear aggregate demand functions have been extensively used in the analytical models literature (80 papers) (e.g., Choi, 1991; Ingene and Parry, 1995b;McGuire and Staelin, 1983; Shugan and Jeuland, 1988; Trivedi, 1998). Choi (1991), however, was also the first to show that the type of demand function assumed – linear or nonlinear – could have a major impact on conclusions about the optimal decentralized channel structure under various power balance structures. His results underscore the importance of employing a correct demand function for a channel decision. Accordingly, in recent years, we have seen an increase in the number of papers deriving the demand functions from a theoretically sound, individual-level utility function model, based, for example, on flexible spatial models (e.g., Yoo and Lee, 2011; Lee et al., 2013). Such models allow consistent analyses and comparisons of numerous channel structures under a large number of underlying market environments.