تبعیض قیمت مبتنی بر سابقه و ورود در بازار با تغییر هزینه ها: یک تجزیه و تحلیل رفاه اجتماعی

We analyze history-based price discrimination in an asymmetric industry, where an incumbent, protected by switching costs, faces an entrant who does not have access to information about consumers’ purchase histories. We demonstrate that consumer surplus is higher with uniform pricing than with history-based price discrimination. We find that the entry decision is invariant to whether the incumbent implements history-based pricing or uniform pricing. This implies that the potential abuse of market dominance imposed by history-based price discrimination is exploitation, not exclusion. Finally, we establish that the profit gain to the incumbent from history-based pricing exceeds the associated loss to consumers.

Business practices with prices set on the basis of observed customer purchasing behavior are now widely observed. It has become common in, for example, telecommunication, service industries and energy markets to differentiate the prices directed to old and new customers. Typically new customers are targeted by aggressive price offers (introductory offers or poaching), which are designed to attract new customers or to induce rival firms’ customers to switch even when those customers are already locked-in in another customer relationship. This form of competition is one manifestation of history-based pricing. In this study we will explore the antitrust implications of history-based pricing by focusing on the following questions: Can history-based pricing be viewed as an instrument for a dominant firm to induce exclusion of a potential entrant? Does it prevent consumers from enjoying the benefits of competition? Does history-based pricing make dominance persistent? What is precisely the relationship between history-based pricing and uniform pricing by an incumbent firm? Which are the effects of history-based pricing for consumers? A number of European antitrust cases have established how history-based price discrimination might facilitate predation in a way which would, according to competition authorities or courts, qualify as an abuse of a dominant market position. The seminal case exemplifying this is the ECS-AKZO case1 where AKZO targeted selective price cuts to ECS's customers with the intention of excluding ECS from the market. According to the decision of the European Commission “the anticompetitive effect of AKZO's differential pricing involved not so much indirect injury to customers but rather a serious impact on the structure of competition at the level of supply by reason of its exclusionary effect” (Section 83 of the European Commission's Decision on the ECS/AKZO case).2 Another example is the Irish Sugar case, where the Commission fined Irish Sugar in 1997 for abuse of its dominant position in the national sugar market. The fined corporation applied a scheme of target rebates such that the rebate was more favorable to particular customers of competing sugar packers. This ruling was upheld by the European Court of First Instance in the case Irish Sugar vs. Commission, where the Court supported the Commission's finding that the selective price cut by Irish Sugar to its rival's customers, had it been proven, would have been considered an abuse of a dominant position. For a more extensive and systematic account of European competition law towards price discrimination we refer to Geradin and Petit (2005). The Swedish Competition Authority vs. TeliaSonera is a national competition case from year 2005 illustrating how selective poaching offers by a dominant firm to a small rival's customers may qualify as an abuse of market dominance. This case focuses on fixed line telecommunications as the relevant market, where TeliaSonera had a dominant position in the Swedish market. In this case TeliaSonera directed selective poaching offers exclusively to customers of Bredbandsbolaget, a small regional rival.3 In this study we analyze the effects of history-based price discrimination in an asymmetric industry, where an incumbent, protected by switching costs, faces an entrant who does not have access to information about consumers’ purchase histories. As for the effects on market shares we find the persistence of market dominance for an incumbent firm to be invariant across the regimes with history-based pricing and uniform pricing. Nevertheless, consumers would benefit from a policy which bans history-based price discrimination insofar as consumer surplus is higher with uniform pricing than with history-based pricing. According to our analysis, the entry decision of a firm introducing a competing brand is invariant to whether the incumbent firm implements history-based pricing or uniform pricing. This result has strong implications for the implementation of Article 82 in Europe. It implies that the potential abuse of consumers imposed by history-based price discrimination is exploitation, not exclusion. Finally, we establish that history-based price discrimination benefits society in the sense that the profit gain from history-based pricing to the incumbent exceeds the associated loss to consumers. An important sequence of studies has earlier explored the effects of price discrimination across separated markets on entry and welfare. Armstrong and Vickers (1993) focused on a framework where the incumbent firm operates in an exogenously determined sheltered segment as well as a segment subject to potential competition, and they found that price discrimination across markets tends to discourage entry. Cheung and Wang (1999) extended this approach in certain respects and found that price discrimination across markets may encourage or discourage entry, depending on the price elasticity of the competitive market and that of the captive market. This approach has been further developed by Jorge and Pires (2007). Armstrong (2008) and Motta (2004, Section 7.4) present general perspectives on this literature. Bouckaert et al. (2008) have extended the analysis of the effects of price discrimination across separated markets in a two-period framework. Chen (2008) also analyses history-based pricing in asymmetric duopoly and its implications for consumer surplus. While his focus is on exclusionary practices by the dominant firm in order to induce exit, we focus on market entry in a setting of extreme initial asymmetry.4 The studies mentioned above all focus on price discrimination within a framework where the dominant firm operates in an exogenously determined sheltered segment as well as a segment subject to competition. Contrary to these approaches, we explore the consequences of history-based price discrimination within a framework where the loyal segment of a dominant firm is endogenously determined. Contrary to these studies we design a model focusing exclusively on history-based price discrimination within one market. This means, in particular, that the loyal segment of the dominant firm is determined endogenously. Our study is structured as follows: Section 2 presents a model of history-based price discrimination and entry. It also explores the effects of history-based pricing on market shares. Section 3 analyzes the similar market configuration for the case in which the incumbent is restricted to charge a uniform price to all buyers. Section 4 investigates the implications of history-based pricing on consumer surplus, profits of the incumbent and the entrant, and total welfare. Section 5 concludes.

In this study we analyzed the effects of history-based price discrimination on entry and welfare in a model where an incumbent, protected by switching costs, faces an entrant who does not have access to information about consumers’ purchase histories. We initially explored the effects of history-based pricing on market shares and found the persistence of market dominance for the incumbent to be invariant across the regimes with history-based pricing and uniform pricing. Nevertheless, a detailed welfare analysis revealed that consumers would benefit from a policy which bans history-based price discrimination insofar as consumer surplus is higher with uniform pricing than with history-based pricing. According to our analysis, history-based price discrimination was found to neither discourage nor encourage entry. This implies that the potential abuse of market dominance imposed by history-based price discrimination is exploitation, not exclusion. This is remarkable because the standard attempts to ban price discrimination typically focus on exclusion as a manifestation of abuse of market dominance. For example, the European Commission “has issued on 3 December 2008 Guidance on its enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings.”7 Our results imply that the potential abuse associated with history-based pricing by a dominant firm is exploitation without exclusion. Finally, we established that the profit gain from history-based pricing to the incumbent exceeds the associated loss to consumers. Of course, we are by no means saying that exclusionary concerns are not important when analyzing the antitrust effects of history-based price discrimination. In this respect we are not implying that enforcement policies of the European Commission are necessarily wrong, but we add an important concern largely overlooked in the current debate. We argue that under simple and theoretically appealing configurations it is well possible that the antitrust concern is exploitation, not exclusion. It should also be emphasized that there are important theoretical studies in support of the Commission's enforcement priorities. For example, Chen (2008) focuses on an asymmetric duopoly, where history-based pricing by a dominant firm poses a anticompetitive threat precisely when it facilitates predation based on the exit of the small rival. Theoretically Chen (2008) analyzes the effects of history-based pricing in an environment which is very different from ours. We focus on a standard Hotelling model, whereas Chen offers an analysis with an arbitrarily long time horizon and with a segmented market where the firms do not compete head-to-head when they apply uniform pricing. Our study has the policy implication that the potential abuse of market dominance imposed by history-based pricing is exploitation, not exclusion. This is highlighted, since in our example the entrant's profit is invariant across the pricing regimes with discriminatory history-based prices and uniform prices. Of course, our example draws heavily on the linear structure of the standard Hotelling framework. That framework has the implication that the optimal pricing scheme of the incumbent with one or two price instruments will always generate the same incumbent market share for any feasible price quote of the entrant. In more general frameworks the invariance results will not generally prevail. In those cases some element of exclusionary concern will be added to the exploitative issue that we focus on. In any case, policy recommendations will depend on the relative strengths of these concerns in the relevant markets. Throughout this study we have analyzed the effects of history-based pricing on entry and welfare within the framework of a very limited horizon. If entry takes place, the strategic interaction between the incumbent and the entrant could continue for many periods. After entry both competitors could then apply history-based pricing. Within such a framework one could investigate the dynamics of dominance and, in particular, characterize the market shares towards which the process would converge and also explore the associated welfare consequences.8 In ongoing work we analyze the welfare effects of history-based price discrimination within the context of an asymmetric duopoly model, where one firm has inherited a dominant market position, but where both firms can apply history-based pricing.9 It remains an interesting challenge for future research to incorporate such an analysis into an entry model with a longer horizon. How will the ability of firms to exploit historical information about their customers for a longer horizon affect pricing and market shares? Would this process converge to perfect price discrimination in the limit, and what would be the consequences of a ban on using such information and other privacy restrictions? Entry barriers can be introduced in several ways. Following the large literature in industrial economics, the traditional approach to entry barriers is to view them as stemming from having sunk and fixed costs of entry. This literature has analyzed in great detail strategic entry deterrence. In Karlinger and Motta (2007) the presence of network externalities generates an entry barrier, and within the context of such a model they demonstrate that price discrimination may induce exclusion which is not the case in the present analysis. In the present analysis, the switching costs constitute the entry barrier. This could be particularly relevant in, for example, service industries such as online shopping, the airlines industry, and banking, where the informational advantages created through an established customer relationship might represent a significant entry barrier.