Product Differentiation and the Consistency of Monopolistic Competition: A Spatial Perspective

Abstract

IN RECENT years researchers have increasingly recognized the similarities between problems in spatial economics and problems of product differentiation in monopolistic competition. Authors working from the spatial side have tried to outline some of the implications of spatial economics for traditional microtheory. At the same time, other writers such as Lancaster [17] have developed models of product differentiation which have a spatial flavor. It has become apparent that spatial economics can provide considerable insight into monopolistic competition. In this paper we address a long smoldering debate in the theory of the firm concerning whether or not monopolistic competition is a distinctly different case from perfect competition and monopoly. The issue was ignited by Demsetz [8], [9], [10] who questioned both the capacity theorem of monopolistic competition and the consistency of the underlying assumptions. Later Barzel [1] and Schmalensee [22] presented analyses that appeared to have successfully doused Demsetz' criticism of the excess capacity theorem. Recently, however, the debate has been rekindled by Greenhut [14], Ohta [21] and Murphy [20]. Most of the disagreement has been encamped around the excess capacity/efficiency question. A separate but related series of papers in the spatial competition literature (e.g. Eaton and Lipsey [11], [12], [13], and Capozza and Van Order [7]) has challenged the consistency of the free entry and zero profits assumptions. The link between the Demsetz inconsistency arguments and similar reasoning in the spatial competition literature is most apparent if one views product differentiation, a basis of monopolistic competition, as a locational concept (see for example Lovell [18], Eaton and Lipsey [13]). Indeed product differentiation in many large firms is a marketing problem where modern methods call for positioning products in the consumers' n-dimensional characteristics space. We try to demonstrate that monopolistic competition is indeed a distinct structure. We use a spatial model where monopolistic competition is the typical intermediate case while monopoly and perfect competition are extreme cases. We then argue that product differentiation, which has been the usual rationale for monopolistic competition, can be modelled in a manner analogous to the spatial model, implying a large degree of generality than our model.