Product Differentiation Research

Abstract

Product differentiation has been an important field in industrial organization economics and has inspired theoretical and empirical research on competition in strategic management and marketing, among other disciplines. To better understand this important field, we review the literature in a critical manner. Here we have two purposes. First, we update some recent developments in the last decade or two. One interesting aspect of recent developments is an active incorporation of consumer heterogeneity, elastic demand in the model, two-side markets with network externalities, among others. Second, we assess the sensitivity of equilibrium outcomes on their assumptions. A critical review of the literature allows us to argue that the assumptions of quadratic transportation costs, elastic demand, and concave consumer distribution are influential in determining the level of differentiation. These assumptions, individually or together, increase product differentiation, possibly to the maximum level, by making consumers quite sensitive to (even small) price changes. And we also argue that when we assume multiple products and asymmetric cost positions, we have results similar to those suggested by resource endowments argument in strategic management. Based on the review, we offer several suggestions for future research. First, given the influential effect of the assumption of quadratic transportation costs, we suggest that researchers should make it clear why they choose to assume quadratic transportation costs over linear transportation costs. We believe this issue could be solved rather easily if researchers are more explicit about the applications of derived results rather than the derivation itself. Second, the impact of elastic demand on the degree of differentiation suggests that researchers should be particularly cautious in empirical research. Any condition that triggers price rigidity should be explicitly represented in the equation. Otherwise the degree of differentiation would not be properly estimated due to the endogeneity caused by the absence of this condition on the right side of the equation. Third, regarding unknown distribution of consumer density, researchers should be advised to include income distribution in their empirical model. Unlike unknown distribution of consumer heterogeneity, income distribution could be available if the target of the application of derived results is known. Lastly, it may prove to be quite fruitful if researchers build upon some of recent developments. Researchers may want to extend vertical differentiation in two-side markets with network externalities that may have interesting implications on on-demand economy or researchers may want to take up the notion of variable elasticity of substitution that could open up new possibilities of expanding research streams on supply-side (e.g., production costs) of product differentiation.