«Exercising Market Power in Proprietary Aftermarkets SEVERIN BORENSTEIN Haas School of Business University of California Berkeley, CA 94720-1900 ...»
Exercising Market Power in
Haas School of Business
University of California
Berkeley, CA 94720-1900
JEFFREY K. MACKIE-MASON
School of Information
Department of Economics and School of Public Policy
University of Michigan
Ann Arbor, MI 48109-1092
JANET S. NETZ
Department of Economics Purdue University West Lafayette, IN 47907-1310 email@example.com Over 20 recent antitrust cases have turned on whether competition in complex durable-equipment markets prevents manufacturers from exercising market power over proprietary aftermarket products and services. We show that the price in the aftermarket will exceed marginal cost despite competi- tion in the equipment market. Absent perfectly contingent long-term con- tracts, firms will balance the advantages of marginal-cost pricing to future generations of consumers against the payoff from monopoly pricing for current, locked-in equipment owners. The result holds for undifferentiated Bertrand competition, differentiated duopoly, and monopoly equipment mar- kets. We also examine the effects of market growth and equipment durability.
1. Introduction In the past decade, many independent providers of service for high- technology products have sued equipment manufacturers for al- legedly excluding them from providing maintenance services. Over twenty antitrust cases have been brought against manufacturers such Q 2000 Massachusetts Institute of Technology.
Journal of Economics & Management Strategy, Volume 9, Number 2, Summer 2000, 157 ] 188 158 Journal of Economics & Management Strategy as Kodak, Prime Computer, Data General, Northern Telecom, Picker, Unisys, Xerox, Rolm, Hewlett-Packard, EDS, General Electric, and Siemens.1 The common feature in these cases is that the defendants manufacture complex durable equipment for which customers de- mand service, support, parts, and r or upgrades for many years after the initial sale. The economic interaction between the original equip- ment market and aftermarkets is central to the analysis.
Most of the cases have a similar plot. The manufacturer sells one brand of complex equipment in a market that is fairly competitive ( e.g., the market for minicomputers ). In addition, the manufacturer sells aftermarket products to customers who purchased the original equipment. Examples of aftermarket products include hardware maintenance contracts, spare parts, and software upgrades. Due to proprietary rights, the original manufacturer is often the exclusive seller of at least one aftermarket product, such as replacement parts or upgrades to the operating-system software. Plaintiffs charge that the manufacturer exploits its aftermarket position in violation of antitrust laws, typically by tying the purchase of an aftermarket product that is also available from the plaintiffs ( usually service ) to the manufacturer’s proprietary good, that is, by leveraging a monopoly over parts into a monopoly over service.
Once a customer purchases a particular brand of complex, durable equipment, she is likely to be ‘‘locked in’’ to that manufacturer to some extent. There are often significant costs of switching to another brand: retraining, sunk investments in custom software, capital losses on the sale of the used equipment, etc. It would seem that these switching costs could provide the manufacturer with room to collect some monopoly rents by raising aftermarket prices above cost.2 However, a manufacturer that exploits locked-in customers with high aftermarket prices might make its equipment less attractive when it competes with other manufacturers.
The question we address is fundamental: does substantial competition in the durable-equipment market necessarily discipline manufacturers so that they will not exercise market power in the aftermarket? This has been the central claim in the many antitrust cases cited, supported by expert testimony from a number of indusEach of the authors has advised parties to aftermarket antitrust cases.
2. A number of papers demonstrate how the presence of switching costs can endow a firm with market power after consumers make their initial choice. See, e.g., Klemperer ( 1987 ) and Farrell and Shapiro ( 1987 ). Well-known examples include computers with proprietary operating software and printer toner cartridges for laser and inkjet printers.
Market Power in Proprietary Aftermarkets 159 trial organization economists.3 Although there is a recent flurry of articles concerning aftermarket market power, none provide a formal answer to this basic question.4 In this paper, we show that equipment competition will not prevent firms from charging supracompetitive prices in their proprietary aftermarkets; the incentive to exercise at least some degree of market power in the aftermarket is unambiguous.
Of course, this is not the only issue in the antitrust cases we have cited, and we are not taking a position with regard to any specific allegations of aftermarket monopolization. Shapiro ( 1995) has wisely noted that the social welfare costs of some aftermarket inefficiencies may be small, and of course there are costs to fashioning a policy remedy. Chen and Ross ( 1993) have shown that even when the equipment market is monopolized, the welfare consequences of some types of aftermarket exploitation are ambiguous. Our objective is to focus attention on the use of aftermarket strategies by durable-goods manufacturers. As a theoretical matter, the assertion that a competitive equipment market prevents a firm from exercising market power in the aftermarket is wrong. As a policy issue, we believe the discussion should focus on the facts of each case and on an assessment of how much harm will occur in each particular situation.
In Section 2 we review the recent legal history, emphasizing how the central economic question has emerged from important fact contexts. In particular, we discuss the role of lock-in and reputational effects. We develop a differentiated Bertrand duopoly model in Section 3 to address the incentive to exercise market power in the aftermarket and its effect on welfare. This model not only approaches perfect competition in the equipment market as the degree of differentiation approaches zero, but it also allows for the existence of some economic profits, which we argue as necessary in order to study how reputation might affect the incentives to exploit aftermarkets. We find that a reputation for low aftermarket prices does have value and that firms will therefore price the aftermarket product below its monopoly level. However, we also find that firms always price above cost for the aftermarket product.
3. For example, Kodak argued before the Supreme Court that this proposition must hold as a matter of theory: ‘‘even if it concedes monopoly share of the relevant parts market, it cannot actually exercise the necessary market power for a Sherman Act violation... equipment competition precludes any finding of monopoly power in the derivative aftermarkets’’ w Kodak, 112 S. Ct. 2072 ( 1992 ) at 2081] 2182, emphasis in original x.
4. See, e.g., Borenstein et al. ( 1995), Chen and Ross ( 1993, 1999), and Shapiro ( 1995).
160 Journal of Economics & Management Strategy Assertions that equipment market competition will necessarily discipline aftermarket behavior tend to refer to the equipment market as ‘‘highly’’ competitive. Therefore, within the context of our differentiated duopoly, we examine two issues. First, we consider the effects on the aftermarket price as the degree of differentiation between the two firms approaches zero, so that we approach standard Bertrand competition. We continue to find that service price remains higher than cost.5 Then, we consider how the option to scrap used equipment and buy new rather than purchase aftermarket service affects the aftermarket price. We show that if the option to scrap and buy new is binding on the margin, the aftermarket price is increasing in the firm’s degree of equipment market power. These results emphasize our main point for antitrust analysis: the exercise of aftermarket market power is not ruled out by ‘‘highly competitive’’ equipment markets, but is a matter of degree. Rather than merely assume that aftermarket market power is nonexistent or is significant as a matter of theory, some degree of factual inquiry into the market conditions is appropriate.
We extend the analysis in Section 4 by studying the aftermarket pricing behavior of an equipment monopolist. The fundamental result is the same: we show that the aftermarket price is bounded away from marginal cost, but also is not at the full monopoly aftermarket level. We also show that the monopolist charges less in the aftermarket than does a duopolist, at least in our spatial competition model.
We argue that this result has an intuitive explanation: the monopolist can extract a larger share of the surplus created when it lowers the aftermarket price ( by raising the equipment price ) than can a duopolist, so the monopolist has a stronger incentive to keep the aftermarket price down.
We conclude in Section 5 by summarizing our findings and discussing some other incentives to monopolize aftermarkets—some of which may benefit and others of which may harm consumers—that are not modeled here.
5. Alternatively, in the appendix, available at www.umich.edu r ; jmm r papers r aftmkt-appendix.pdf, we develop a model of a perfectly competitive equipment market in which each equipment firm monopolizes its service market. We find once again that service price will be above cost. Competition in the equipment market causes firms to make zero profits overall, but a welfare loss occurs due to an aftermarket price above cost.
Market Power in Proprietary Aftermarkets 161
2. Aftermarket Economic Power in the Courts There are numerous cases before the federal courts that involve claims of antitrust violations in aftermarkets for service products.
Two have recently reached the Supreme Court. In the first, firms selling service for Kodak high-volume photocopiers and micrographic equipment alleged that Kodak adopted a restrictive policy on the availability of spare parts, including tying sales of spare parts to the purchase of other maintenance services from Kodak. The Court upheld the Circuit Court’s denial of Kodak’s motion for summary judgment, concluding that ‘‘it is clearly reasonable to infer that Kodak has market power to raise prices and drive out competition in the aftermarkets... w and x... to infer that Kodak chose to gain immediate profits by exerting that market power where locked-in customers, high information costs, and discriminatory pricing limited and perhaps eliminated any long-term loss’’ w Kodak, 112 S. Ct. 2072 ( 1992) at 2088 x. The Kodak case then went to trial; Kodak lost on this issue at trial and on appeal w Image Tech Services v. Eastern Kodak Co., 125 F3d. 1995 ( Ninth Circuit, 1997) x. In another case, an independent service company alleged that Prime Computer had tied the sale of software support and upgrades to the purchase of hardware maintenance from Prime. Prior to the Supreme Court decision in Kodak, the Sixth Circuit had accepted Prime’s argument that competition in the equipment market would necessarily discipline aftermarket prices.
The Supreme Court overturned this decision shortly after deciding Kodak. The Sixth Circuit then decided that sufficient evidence had been presented to support a finding that it was profitable for Prime to monopolize the service aftermarket w Virtual Maintenance v. Prime Computer, 11 F3d. 660 ( Sixth Circuit, 1993) x.6 Two main features that distinguish aftermarkets have emerged in the many antitrust cases before the courts: the role of customer lock-in establishing market power and the possibility that reputation effects will prevent manufacturers from profitably exploiting whatever economic power they have in service aftermarkets. We discuss these factors now, before proceeding to our formal model.
6. Three recent cases have narrowly interpreted or partially conflicted with the Supreme Court in Kodak. See MacKie-Mason and Metzler ( 1999 ) for a discussion of these.
162 Journal of Economics & Management Strategy
2.1 Customer Lock-In The availability of substitutes limits a manufacturer’s ability to charge above-competitive prices for its aftermarket products. An aftermarket customer would choose to sell or scrap the used equipment and purchase anew from a different manufacturer if the original seller raised the service price sufficiently. The extent to which switching costs discourage a customer from changing brands affects how much room the manufacturer has to raise the service price.
The equipment involved in most of the recent antitrust cases is quite sophisticated. The products include minicomputers, hospital CT scanners, telephone PBX switches, high-volume photocopiers, and micrographic reproduction equipment. In every case, users and experts have testified to the high costs of switching.7 Evidence introduced in the Wang 8 case showed that typically about 80 percent of minicomputer consumers buy the same brand when they replace their equipment.
Previous work has demonstrated the role of switching costs in creating market power, but it has focused on a single product to which the consumer becomes locked in.9 There has been little attention to a firm that sells equipment in a competitive market but sells service to locked-in customers.10 When there are two interrelated markets, the central question becomes the ability of the manufacturer to profitably exercise economic power in one market without a larger adverse impact on profits in the other market.
2.2 Reputation and Imperfect Competition Manufacturers face two types of customers: those who already own equipment and those who are purchasing for the first time. Although customers who already own equipment may face significant costs of switching brands and thus provide the manufacturer with an opportunity to price supracompetitively, de novo customers do not. Do potential new customers provide sufficient competitive discipline in the aftermarket? The answer depends on reputation effects. It may
7. For example, a senior design systems manager for Ford Motor Co. testified in Virtual that switching from Prime minicomputers to another brand would shut Ford down. See also Kodak, 112 S. Ct. 2072 ( 1992) at 2087.