The antitrust assessment of loyalty discounts and rebates

Journal of Competition Law and Economics

Volumn 4, Issue 2, 2008, Pages 375-410

  Faella, Gianluca ^a  

^a Cleary Gottlieb Steen and Hamilton   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 50349083613     PISSN: 17446414     EISSN: 17446422     Source Type: Journal    
DOI: 10.1093/joclec/nhn004     Document Type: Article

References (222)

1. Fidelity discounts may be conditioned on the achievement of certain thresholds, or granted in exchange for an explicit exclusivity obligation, thus giving rise to a form of exclusive dealing.
2. As a general rule, the target is set on the basis of the client's requirements, so as to correspond to all or a large part of his demand. Accordingly, fidelity and target discounts and rebates tend to have similar effects, as recognized by the Commission's decision practice: see Commission decision of 20 March 2001, Case COMP/35.141, Deutsche Post AG, O.J. L 125/27 (2001)
3. and 14 July 1999, Case IV/D-2/34.780, Virgin/British Airways, O.J. L 30/1 (2000). The choice between the two types of practices may depend on the stability and predictability of demand. When demand fluctuates significantly over time, suppliers may prefer to make reference to a certain percentage of the customers' requirements, rather than predetermined individual targets. Both the fidelity and target discount systems may provide for a single threshold or a grid of ascending thresholds, corresponding to increasing discount rates.
4. On the EU experience, see, ex plurimis, Alison Jones & Brenda Sufrin, EC Competition Law: Text, Cases & Materials (3rd edn, Oxford: Oxford University Press, 2007)
5. Valentine Korah, An Introductory Guide to EC Competition Law and Practice (9th edn, Oxford: Hart Publishing, 2007)
6. and Aldo Frignani, "L'abuso di posizione dominante," in Aldo Frignani & Roberto Pardolesi (eds), La concorrenza, in Gianmaria Ajani & Gian Antonio Benacchio (eds), Trattato di Diritto Privato e dell'Unione Europea (Torino: Giappichelli, 2006) 178.
7. See DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses, Brussels, December 2005, available at http://ec.europa.eu/comm/competition/antitrust/art82/ index.html. Interestingly, the most significant changes to the current approach would concern the treatment of discounts and rebates, where the state of the law is commonly perceived as more unsatisfactory.
8. Nonlinear pricing occurs when the price paid by a customer is not strictly proportional to the quantity purchased. See Robert B. Wilson, Nonlinear Pricing (Oxford: Oxford University Press, 1992).
9. Frank P. Maier-Rigaud, Switching Costs in Retroactive Rebates: What's Time Got to Do with it", 26 EUR. COMPETITION L. REV. 272 (2005).
10. On switching costs, see Ciara McSorley, A. Jorge Padilla & Mark Williams, Switching Costs, Department of Trade and Industry - Office of Fair Trading, Discussion Paper No. 5, April 2003, available at http://www.oft.gov.uk/advice_and_resources/publications/reports/ Economic-research/oft655 (visited 30 January 2008).
11. See Maier-Rigaud, supra note 6.
12. Consider for instance a rollback rebate system under which a supplier offers a 5 percent rebate on a unit price of 10 if the quantity purchased is equal to 100 units or greater. When the units purchased increase from 99 to 100 and the threshold is reached, total expenditure falls from 990 to 950, meaning that the incremental price of the last unit is negative. For a discussion and a graphic representation of marginal and average prices and total expenditure under a retroactive discount system, see Giulio Federico, When are Rebates Exclusionary, 26 EUR. COMPETITION L. REV. 477 (2005).
13. For a discussion of purchasers' choices under conditions of bounded rationality, see Martin Beckenkamp & Frank P. Maier-Rigaud, An Experimental Investigation of Article 82 Rebate Schemes, 2(2) THE COMPETITION L. REV. 1 (2006), according to whom retroactive discounts and rebates exert a significant attraction that may enhance the potentially exclusionary effect of the scheme.
14. See also Janusz Ordover & Greg Shaffer, "Exclusionary Discounts," CCP Working Paper 07-13 (2007), available at http://ssrn.com/abstract=995426 (visited 30 January 2008). The authors analyze the exclusionary effects of loyalty discounts arising in a two-period model with two sellers and one buyer, when the buyer's valuations between periods are linked by switching costs and at least one seller is financially constrained.
15. Alberto Heimler, Below Cost Pricing and Loyalty-Inducing Discounts: Are They Restrictive and, if so, When?, 1 COMPETITION POL'Y INT'L 149 (2005)
16. see also OECD, Loyalty and Fidelity Discounts and Rebates (2003), proceedings of the Roundtable on Loyalty and Fidelity Discounts and Rebates held by the Competition Committee in June 2002, available at http://www.oecd.org/dataoecd/18/27/2493106.pdf (visited 30 January 2008).
17. The existence of an assured base of sales may be due to several factors, such as brand loyalty, product differentiation, switching costs, or capacity constraints faced by competitors. See Adrian Majumdar, Simon Bishop, Derek Ridyard, Iestyn Williams & Ugur Akgun, "Selective Price Cuts And Fidelity Rebates," Economic discussion paper prepared for the Office of Fair Trading by RBB Economics, July 2005, available at http://www.oft.gov.uk/shared_oft/reports/comp_policy/ oft804.pdf (visited 30 January 2008), at 117 ff.
18. See T. Randolph Beard, George S. Ford & Lawrence J. Spiwak, "Quantity-Discount Contracts as a Barrier to Entry," Phoenix Center Policy Paper No. 20 (2004), available at http://www.phoenix-center.org/pcpp/PCPP20Final.pdf (visited 30 January 2008)
19. Willard Tom, David Balto & Neil Averitt, Anticompetitive Aspects of Market-Share Discounts and Other Incentives to Exclusive Dealing, 67 ANTITRUST L. J. 615 (2000)
20. Maier-Rigaud, supra note 6.
21. See also David Spector, Loyalty Rebates: An assessment of Competition Concerns and a Proposed Structured Rule of Reason, 1 COMPETITION POL'Y INT'L 89, 96 (2005), at 96, stating that, in some circumstances, loyalty rebates may be equivalent to a cheaper, and thus more efficient, form of predatory pricing.
22. Y. Joseph Lin, The Dampening of Competition Effect of Exclusive Dealing, 39 J. INDUS. ECON. 209 (1990).
23. Paul Klemperer, The Competitiveness of Markets with Switching Costs 18 RAND J. ECON. 138 (1987).
24. See Patrick Greenlee & David Reitman, "Competing with Loyalty Discounts" (2005), available at http://ssrn.com/abstract=502303 (visited 30 January 2008).
25. A possible exception might theoretically be the unusual case in which a group of collectively dominant undertakings used the practice to limit, to a certain extent, their reciprocal competitive pressure.
26. On the possible business justifications for the use of the practice, see David Spector, "Loyalty Rebates and Related Pricing Practices: When Should Competition Authorities Worry?," in David S. Evans & A. Jorge Padilla (eds), Global Competition Policy: Economic Issues and Impact (2004), 317, available at http://www.lecgcp.com/ec/forum/index.aspx?id=20 (visited 30 January 2008)
27. and Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
28. In oligopolistic markets, firms may have low incentives to implement a linear reduction of prices, because such reduction would be more transparent and might be readily replicated by rivals. See John Kallaugher & Brian Sher, Rebates Revisited: Anti-Competitive Effects and Exclusionary Abuse Under Article 82, 25 EUR. COMPETITION L. REV. 263 (2004).
29. The practice allows for the sale of additional units to customers who value the product more than its marginal cost and contribute to covering the fixed costs, thus increasing the incentives of firms to invest and innovate. See Derek Ridyard, Exclusionary Pricing and Price Discrimination Abuses Under Article 82: An Economic Analysis, 23 EUR. COMPETITION L. REV. 286 (2002)
30. Spector, supra note 16.
31. In particular, when it is difficult to identify customers with an elastic demand, a discount system may function as a tool to induce self-selection by buyers. See Alberto Pera & Vito Auricchio, Consumer Welfare, Standard of Proof and the Objectives of Competition Policy, 1 EUR. COMPETITION J. 153 (2005).
32. Sreya Kolay, Greg Shaffer & Janusz Ordover, All-unit Discounts in Retail Contracts, 13 J. ECON. & MGMT. STRATEGY 429 (2004), show that retroactive discounts can be a welfare-improving mechanism to price discriminate between retailers.
33. On the efficiency virtues of exclusive dealing agreements, see, ex plurimis, Howard P. Marvel, Exclusive Dealing, 25 J. L. & ECON. 1 (1982)
34. Richard M. Steuer, Exclusive Dealing in Distribution, 69 CORNELL L. REV. 101 (1983)
35. Stanley I. Ornstein, Exclusive Dealing and Antitrust, 34 ANTITRUST BULL 65 (1989)
36. Paul W. Dobson & Michael Waterson, "Vertical Restraints and Competition Policy," OFT Research Paper 12 (1996), available at http://www.oft.gov.uk/shared_oft/reports/comp_policy/oft177.pdf (visited 30 January 2008), at 16
37. Jan B. Heide, Shantanu Dutta & Mark Bergen, Exclusive Dealing and Business Efficiency: Evidence from Industry Practice, 41 J. L. & ECON. 387 (1998)
38. Richard M. Steuer, Customer-Instigated Exclusive Dealing, 68 ANTITRUST L. J. 239 (2000)
39. Benjamin Klein, Exclusive Dealing as Competition for Distribution on the Merits, 12 GEO MASON L. REV. 119 (2004).
40. On the use of loyalty discounts as an instrument for improving coordination inside the production chain, see Andrew Gavil, Exclusionary Distribution Strategies by Dominant Firms: Striking a Better Balance, 72 ANTITRUST L. J. 3 (2004).
41. Loyalty discounts make the average price offered to distributors higher than the marginal price. Because the marginal price determines the distributors' incentive to undertake promotional and sale efforts to sell additional units of the products concerned, the practice may be a strategic tool to boost sales. See Spector, supra note 16
42. D. E. Mills, "Market Share Discounts" (2006), available at http://virginia.edu/economics/papers/mills/ Market%20Share%20Discounts.pdf (visited 30 January 2008)
43. OECD, supra note 9.
44. Even though some promotional activities may be contracted upon, an incentive system may be a more effective and less intrusive instrument to secure the sale efforts of a distributor, if the distributor is in a better position to establish which initiatives should be adopted or if it is difficult to monitor the fulfillment of contractual obligations. See Mills, "Market Share Discounts;" Kallaugher & Sher, supra note 17
45. Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
46. The manufacturer can use the minimum threshold required to qualify for the discount to induce the retailer to choose the joint profit-maximizing price. The retroactive discount allows the maximized surplus to be divided between the manufacturer and the retailer. See Kolay, Shaffer & Ordover, above n 19.
47. John Temple Lang & Robert O'Donoghue, Defining Legitimate Competition: How to Clarify Pricing Abuses under Article 82 EC, 26 FORDHAM INT'L L. J. 83 (2002).
48. See Spector, supra note 16
49. Marvel, supra note 20.
50. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
51. Virgin Atlantic Airways Ltd. v British Airways PLC, 257 F.3d 256, 265 (2nd Cir. 2001).
52. On the U.S. experience, see Bruce H. Kobayashi, The Economics of Loyalty Discounts and Antitrust Law in the United States, 1 COMPETITION POL'Y INT'L 115 (2005), emphasizing the U.S. courts' "focus on the facts rather than on hypotheticals."
53. In Barry Wright, the Court noted that "the Sherman Act does not make unlawful prices that exceed both incremental and average costs," on the assumption that prohibiting above-cost pricing "threatens to 'chill' highly desirable pro-competitive price cutting." See Barry Wright Corp. v ITT Grinnell Corp., 724 F.2d 235-36 (7th Cir. 1979).
54. Similarly, in Henry v Chloride, Inc., the Court underlined that "nonpredatory price cuts are to be encouraged under antitrust law," adding that "at some point, competitors should know for certain they are pricing legally, and...this point should be average total cost. In other words, prices above average total cost are legal per se." See Henry v Chloride, Inc., 809 F.2d 1334, 1344, 1346 (8th Cir. 1987).
55. In Brooke Group, the Supreme Court explained that, in the case of above-cost discounting, it "is beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling legitimate price cutting." See Brooke Group Ltd. v Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993).
56. In SmithKline and Ortho Diagnostic Systems, the courts applied forms of the equally efficient competitor test, with opposite results. Compare SmithKline Corp. v Eli Lilly & Co., 575 F.2d 1056 (3d Cir. 1978), with Ortho Diagnostic Sys. v. Abbot Labs, Inc., 920 F. Supp. 455 (S.D.N.Y. 1996). The controversial and widely debated judgment delivered by the Court of Appeals for the Third Circuit in Le Page's stated that bundled discounts (that is, discounts conditioned on the achievement of certain thresholds in the purchase of different products) may exclude competitors, if the latter do not provide an equally diverse group of services.
57. See Le Page's v. 3m, 324 F.3d 141 (3rd Cir. 2003). Recently, in line with the report issued by the U.S. Antitrust Modernization Commission established in 2002 (see Antitrust Modernization Commission, Report and Recommendations, April 2007, available at www.amc.gov), the Court of Appeals for the Ninth Circuit underlined the necessity to apply a price-cost test to determine whether, as a result of the bundled discount system, the product open to competition is sold at a below cost incremental price.
58. See Cascade Health Solutions v. PeaceHealth, CV-02-06032 (4 September 2007).
59. For a critique of the application of a price-cost test to bundled rebates, see Timothy J. Brennan, "Bundled Rebates as Exclusionary Not Predatory," J. COMPETITION L. ECON., in press.
60. In Concord Boat Corp. v Brunswick Corp., the Court noted that price cuts represent "the very essence of competition," so that, if the discounted price is above average variable cost, "the plaintiff must overcome a strong presumption of legality by showing other factors indicating that the price charged is anticompetitive." Concord Boat Corp. v Brunswick Corp., 207 F. 3d 1039, 1061 (8th Cir. 2000)
61. see also Morgan v. Ponder, 892 F.2d 1355, 1360 (8th Cir. 1989).
62. In Virgin Atlantic Airways Ltd. v British Airways PLC, supra note 26, a bonus system similar to that condemned by the Commission in the European British Airways case was considered lawful because the plaintiff had not proved that the practice was predatory or had anticompetitive effects.
63. On 10 May 2007, in JBDL Corp. v Wyeth-Ayerst Laboratories Inc., the Sixth Circuit affirmed the dismissal of a class action and individual complaints challenging the loyalty rebate programme implemented by a dominant company, on the grounds that there was no evidence of a causal link between the contested conduct and price increases in the market concerned.
64. See also Stitt Spark Plug Co. v. Champion Spark Plug Co., 840 F.2d 1253 (5th Cir. 1988)
65. R.J. Reynolds Tobacco Co. v. Philip Morris, Inc., - F.3d - (3rd Cir. 2003) (en banc).
66. In Minnesota Mining, the District Court noted that the inelastic base of customers' demand could be used to make it more difficult for rivals to buy from other suppliers. Accordingly, it denied a summary judgment motion. See Minnesota Mining & Mfg. Co. v Appleton Papers Inc., 35 F. Supp. 2d 1138, 1144 (D. Minn. 1999).
67. A summary judgment motion was denied also in Avery Dennison Corp. v ACCO Brands, Inc., 2001-1 Trade Cas. (CCH) 72,882 (C.D. Cal. 2000), and In re Hypodermic Products Antitrust Litigation, No. 05-CV-1602 (JLL/CCC) (D.N.J. June 29, 2007).
68. In particular, it has been stated that "as long as prices are above the relevant measure of cost the discounts cannot exclude an equally efficient rival," because "the penalty for not taking the specified percentage is not a breach of contract suit or termination of a franchise. Rather, it is simply the loss of the discount. But the loss of the discount is not a penalty at all if a rival is willing to match the discounted price." Herbert Hovenkamp, Discounts and Exclusions UTAH L. REV. 857 (2006)
69. see also Thomas A. Lambert, Evaluating Bundled Discounts, 89 MINN L. REV. 1688 (2005)
70. PHILLIP E. AREEDA & H. HOVENKAMP, ANTITRUST LAW 768b2, (New York: Aspen Publishers, 2nd ed., 2002), at 149. However, other commentators seem to be much more skeptical about the competitive impact of so-called "market-share discounts."
71. See, for instance, Tom, Balto & Averitt, supra note 11, according to whom the practice should be analyzed in light of principles on exclusive dealing.
72. Fidelity discounts by dominant firms have been considered illegal almost per se, not only when they are granted in exchange for an explicit exclusivity obligation, but also when they are merely conditioned on the achievement of certain thresholds. See, in particular, Commission decision of 2 January 1973, Case IV/26.918, European Sugar Industry O.J. L 140/17 (1973)
73. ECJ, Joined Cases 40-48, 50, 54-56, 111, 113 and 114/73 Cooperatieve Vereniging Suiker Unie U.A. and others v Commission [1975] ECR 1663; Commission decision of 9 June 1976, Case IV/29.020, Vitamins, O.J. L 223/27 (1976)
74. ECJ, Case 85/76 Hoffmann-La Roche v Commission [1979] ECR 965
75. Commission decision of 5 December 1988, Case IV/31.900, BPB Industries plc, O.J. L 10/50 (1989)
76. CFI, Case T-65/89 BPB Industries plc and British Gypsum Ltd v Commission [1993] ECR II-00389
77. Commission decision of 19 December 1990, Case IV/33.133-C, Soda ash-Solvay, O.J. L 152/21 (1991), and 19 December 1990, Case IV/ 33.133-D, Soda ash-ICI, O.J. L 152/40 (1991)
78. Commission decision, Deutsche Post, above n 2
79. Commission decision of 20 June 2001, Case COMP/E-2/36.041/PO, Michelin, O.J. L 143/1 (2002)
80. CFI, Case T-203/01 Manufacture française des pneumatiques Michelin v Commission [2003] ECR II-04071. Target discounts have been subject to similar treatment.
81. See Commission decision of 7 October 1981, Case IV/29.491, Bandengroothandel Frieschebrug BV/NV Nederlandsche Banden-Industrie Michelin, O.J. L 353/33 (1981)
82. ECJ, Case 322/81 NV Nederlandsche Banden-Industrie Michelin v Commission [1983] ECR 3461
83. Commission decision of 14 May 1997, Case IV/34.621, 35.059/F-3, Irish Sugar plc., O.J. L 258/1 (1997)
84. CFI, Case T-228/97 Irish Sugar v Commission [1999] ECR II-2969
85. Commission decision, Virgin/British Airways, supra note 2
86. CFI, Case T-219/99 British Airways plc v Commission [2003] ECR II-5917
87. Commission decision, Michelin, above; Manufacture française des pneumatiques Michelin v Commission, above.
88. See, inter alia, British Airways plc v Commission, supra note 32, 271-73
89. Commission decision, Michelin, supra note 32, 216
90. Virgin/British Airways, supra note 2, 29-30.
91. See, inter alia, Manufacture française des pneumatiques Michelin v Commission, supra note 32, 239, where the CFI stated that, to establish an abuse under Article 82 EC, it is sufficient to "show that the abusive conduct of the undertaking in a dominant position tends to restrict competition or, in other words, that the conduct is capable of having that effect."
92. See also British Airways plc v. Commission, supra note 32, 250
93. Irish Sugar v. Commission, supra note 32, 114.
94. In some cases, the CFI stated the equivalence of the anticompetitive object and effect for the purposes of assessing discount systems (see in particular, Manufacture française des pneumatiques Michelin v Commission, supra note 32
95. Irish Sugar v Commission, supra note 32, 170
96. BPB Industries plc and British Gypsum Ltd v Commission, above n 32), notwithstanding that, under Article 81 EC, only very serious restraints are unlawful for their object, whereas the competitive impact of loyalty discounts is at least ambiguous.
97. See, in particular, NV Nederlandsche Banden-Industrie Michelin v Commission, supra note 21, 81
98. Commission decision, Michelin, supra note 32, 287.
99. In NV Nederlandsche Banden-Industrie Michelin v Commission, above n 32, the ECJ seemed to suggest that only a relatively long reference period-in that case, one year-may trigger the fidelity-enhancing effect of a dominant firm's discount scheme. In San Pellegrino, the Commission decided not to act against a discount system in the market for colas, subject to the condition that the reference period did not exceed three months.
100. See press release IP (90)7 of 9 January 1990. In the 1999 press release on the treatment of discount systems in the air transport sector, following the decision Virgin/British Airways, supra note 2, the Commission indicated that the maximum admissible duration of the reference period was six months.
101. See J. Finnegan, "Commission sets out its policy on commissions paid by airlines to travel agents," 3 Competition Policy Newsletter 23 (1999).
102. Finally, in Michelin, supra note 32, the Commission criticized the significant duration of the reference period, equal to one year, because it increased the pressure to purchase exclusively from the dominant company towards the end of the period.
103. On the contrary, the analysis of the practice seems to be unaffected by the duration of the reference period in the decisions issued by the Commission in Irish Sugar plc, supra note 32 (where the Commission stated that discounts based on single order are lawful, unlike those calculated on the basis of weekly, monthly, or yearly individual targets) and Virgin/British Airways, supra note 2 (where the Commission remained silent on the question of the duration of the reference period, which varied from one month to one year). However, it has been noted that the incongruities arising in antitrust practice are not necessarily negative, because they may reflect, to a certain extent, the peculiarities of different sectors.
104. See Brian Sher, Price Discounts and Michelin 2: What Goes Around, Comes Around, 23 EUR. COMPETITION L. REV. 482 (2002).
105. NV Nederlandsche Banden-Industrie Michelin v Commission, supra note 32, 82.
106. See Commission decision, Michelin, supra note 32.
107. See NV Nederlandsche Banden-Industrie Michelin v Commission, supra note 32, 83-84
108. Commission decision, Bandengroothandel Frieschebrug BV/NV Nederlandsche Banden-Industrie Michelin, supra note 32, 47.
109. In the antitrust literature, see Luc Gyselen, Rebates: Competition on the Merits or Exclusionary Practice?, in EUROPEAN COMPETITION ANNUAL 2003: WHAT IS AN ABUSE OF A DOMINANT POSITION? 287 (Claus-Dieter Ehlermann & Isabela Atanasiu eds., Oxford: Hart Publishing 2006).
110. See, inter alia, Commission decision, Soda ash - Solvay, supra note 32, 16-18.
111. See Commission decision, European sugar industry, supra note 32
112. Virgin/British Airways, supra note 2.
113. The Commission's approach has been fiercely criticized by legal commentators and antitrust scholars. See, inter alia, John Temple Lang, "Abuse under Article 82 EC: Fundamental Issues and Standard Cases," paper presented at the 2006 St Gallen Conference, 19 April 2006, according to whom the fidelity inducing test developed by the Commission "is hopelessly ambiguous" and "provides no way of distinguishing between legitimate and unlawful competition."
114. See also Denis Waelbroeck, Michelin II: A Per Se Rule Against Rebates by Dominant Companies?, 1 J. COMPETITION L. ECON. 149 (2005).
115. See, in particular, British Airways plc v Commission, supra note 32, 293-98
116. Manufacture française des pneumatiques Michelin v Commission, supra note 32, 239-45.
117. See, e.g., Commission decision, Deutsche Post AG, supra note 2.
118. See, in particular, Manufacture française des pneumatiques Michelin v Commission, supra note 32, 239-45.
119. NV Nederlandsche Banden-Industrie Michelin v Commission, supra note 32.
120. See, in particular, Hoffman-La Roche v Commission, supra note 32, 90
121. Cooperatieve Vereniging Suiker Unie U.A. and others v Commission, supra note 32
122. Commission decision, Soda ash-Solvay, supra note 32, 62-64
123. Commission decision, Irish Sugar plc, supra note 32, 151-54
124. British Airways plc v Commission, supra note 32, 228-39.
125. See Commission decision, Bandengroothandel Frieschebrug BV/NV Nederlandsche Banden-Industrie Michelin, supra note 32, 45.
126. In particular, consumer welfare increases if, with a uniform price, only customers whose demand is inelastic would buy the product, whereas price discrimination allows the dominant undertaking to sell additional units to more price-sensitive buyers, without lowering the price charged to the former. See Jeffrey Church & Roger Ware, INDUSTRIAL ORGANIZATION - A STRATEGIC APPROACH 165 (Boston: Irwin McGraw-Hill 2000).
127. On doubts and uncertainties surrounding the antitrust treatment of price discrimination, see, among others, Richard A. Posner, Vertical Restraints and Antitrust Policy, 72 U. Chi. L. Rev. 229 (2005).
128. The beneficial effects of price discrimination may be particularly significant in the case of products covered by intellectual property rights: On the one side, the discrimination might allow innovating firms to secure a higher share of the surplus generated by the invention, thus increasing their incentives to invest in research and development; on the other side, the practice might reduce the allocative inefficiencies generated by the adoption of a single monopolistic price.
129. It has also been noted that loyalty discounts unconnected with the customer dimension and linked, to a certain extent, to distributors' sale efforts might reduce the possible gap between small and large dealers, through the grant of lump sums which may be allocated for promotional activities. See Heimler, supra note 9.
130. See Temple Lang & O'Donoghue, supra note 23.
131. See also Damien Geradin & Nicolas Petit, Price Discrimination under EC Competition Law: Another Antitrust Doctrine in search of Limiting Principles? 2 J. COMPETITION L. ECON. 479 (2006).
132. Manufacture française des pneumatiques Michelin v Commission, supra note 32, 136-50
133. Commission decision, Michelin, supra note 32, 250-53 & 265-66.
134. See, for instance, Guidelines on the application of Article 81(3) of the Treaty (2004/C 101/08), 13 & 33. In Adalat, the CFI made it clear that "the aim of [Article 81 EC] is not to eliminate obstacles to intra-Community trade altogether," so that the Commission is not allowed to attempt to achieve a result, such as the harmonization of prices in the common market, "by enlarging or straining the scope" of antitrust rules.
135. See CFI. Case 41/96 Bayer v Commission [2000] ECR II-3383, 174 & 179
136. on the CFI's judgment see Roberto Pardolesi, "C'era una volta in Europa: Di antitrust comunitario, importazioni parallele ed idoli infranti," Il Foro italiano IV, 78 (2001). The reasoning of the CFI was endorsed by ECJ, Cases 2 and 3/01 P, Commission v Bayer [2004] ECR 1-23.
137. See, inter alia, Commission decision, Soda ash - Solvay, supra note 32, 54
138. Soda ash - ICI, supra note 32, 56 & 62.
139. NV Nederlandsche Banden-Industrie Michelin v Commission, supra note 32, 85.
140. BPB Industries plc and British Gypsum Ltd v Commission, supra note 32, 58.
141. Id., 71.
142. Hoffmann-La Roche v Commission, supra note 32, 96
143. See Sher, supra note 37.
144. See Heimler, supra note 9.
145. The Discussion Paper distinguishes between conditional discounts (that is, schemes where the price charged is linked to a customer condition, such as meeting a sales threshold) and unconditional discounts (that is, a straightforward price reduction with no conditions attached) (171). With regard to the treatment of conditional rebates, the Discussion Paper further distinguishes between conditional discounts on incremental purchases above a given threshold and conditional discounts granted on all purchases in the reference period (para 151). Loyalty discounts discussed in this paper are included in the latter category (retroactive conditional discounts). Regarding unconditional discounts, the Discussion Paper proposes to apply the usual rules on predatory pricing (151). For incremental conditional discounts, the Discussion Paper employs a predation standard, with the important difference that the cost threshold proposed for the assessment of predatory pricing is the average avoidable cost, whereas in the analysis of this form of discounts the Commission makes reference to the average total cost (168).
146. See Discussion Paper, 153.
147. Id. 156.
148. Id. However, in the Discussion Paper the Commission clarifies that it may exceptionally conclude that a market-distorting foreclosure effect results even if the effective price is above ATC. In particular, this may be the case if the dominant company operates in a market where it has certain nonreplicable advantages.
149. See 165.
150. Id., 162.
151. Id.
152. Id., 164.
153. The Commission identified two types of objective justifications: The first type is where the dominant company is able to show that the otherwise abusive conduct is actually necessary on the basis of objective factors ("objective necessity defense"); the second type is where the dominant company is able to show that the otherwise abusive conduct is actually a loss-minimizing reaction to competition from others ("meeting competition defense"). Id. 77-83.
154. Id., 84-92.
155. Id., 173-74.
156. Id., 175.
157. The lack of legal certainty may lead dominant firms wishing to comply with Article 82 EC not to make use of conditional rebate schemes, even where they enhance consumer welfare. This is particularly troublesome in innovative and dynamic industries. See Damien Geradin, Christian Ahlborn, Vincenzo Denicolò & A. Jorge Padilla, "DG Comp's Discussion Paper on Article 82: Implications of the Proposed Framework and Antitrust Rules for Dynamically Competitive Industries" (2006), available at http://ssrn.com/abstract=894466 (visited 30 January 2008).
158. See supra II.A.
159. According to the Discussion Paper, 157, if the estimated commercially viable share was smaller than the required share, the Commission would continue to investigate whether this first estimate was not too low, for instance because competitors' entry or expansion was likely to occur at a larger scale than the minimum efficient scale, or because entrants were likely to concentrate sales on a limited number of clients to whom they can sell more per customer.
160. See Discussion Paper, 163.
161. Commission decision of 29 March 2006, Case COMP/E-1/38.113, Prokent-Tomra, not yet published. On the decision, see Frank P. Maier-Rigaud & Dovile Vaigauskaite, "Prokent/Tomra, a Textbook Case? Abuse of Dominance under Perfect Information," (2) DG COMP Competition Newsletter 19 (2006).
162. For a critical view, see RBB, "Tomra: Rolling Back Form-based Analysis of Rebates?," Brief 21 (2007), available at http://www.rbbecon.com/publications/downloads/rbb_brief21.pdf (visited 30 January 2008).
163. Reverse vending machines are used by supermarkets to collect empty beverage containers from, and return deposits to, consumers.
164. According to the Commission, in the market concerned, it was unlikely that a customer would immediately buy large quantities from a company that had just entered the market, because it was normal practice for customers to test the product first and, subject to a satisfactory result, decide whether to purchase more units. Although this characteristic of the market might have delayed competitors' access or expansion to a certain extent, the Commission's analysis does not demonstrate that rivals did not have the chance to establish themselves in the market by offering better products or lower prices. As the thresholds did not exactly correspond to the entire requirements of individual customers, competitors could in any case sell some units to clients willing to differentiate their supply sources, to make them test competing products.
165. Indeed, the margins on additional sales may well outweigh the cost of having to compensate a customer for failing to meet the threshold. Interestingly, in the Discussion Paper the Commission itself noted that "[t]he suction effect in principle is strongest on the last purchased unit of the product before the threshold is exceeded. However,
166. See Maier-Rigaud & Vaigauskaite, supra note 78.
167. ECJ, Case C-95/04 P British Airways v Commission [2007] ECR I-2331.
168. As already noted, in the United States a result directly contrary to the European one was reached on very similar facts, in Virgin Atlantic Airways Ltd. v British Airways PLC, supra note 26.
169. British Airways v Commission, supra note 83, 68.
170. Id., 73-74.
171. Id., 75-76.
172. See Commission decision, Virgin/British Airways, supra note 2, 29-30.
173. British Airways v Commission, supra note 83, 77, where the ECJ makes reference to "a fidelity-building effect capable of producing an exclusionary effect."
174. See Frank P. Maier-Rigaud, Article 82 Rebates: Four Common Fallacies, 2006, 2 EURO. COMPETITION J. 85 (2006), according to whom "[g]iven the limited use and scarcity of the literature one has to wonder whether there is a reason to alter the policy on rebates under Article 82."
175. See supra, II.A.
176. Reference to actual competitors' cost level would not be adequate, as the prohibition of the practice could protect inefficient firms. In addition, the dominant company might not be aware of rivals' costs, so that it could be inequitable to fine a company for having offered prices lower than competitors' cost level. On the reasons suggesting not to intervene in cases of above-cost exclusion, see, inter alia, Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
177. Contra, see Maier-Rigaud, supra note 89, on the grounds that even an inefficient competitor constitutes a competitive constraint on the dominant firm and thereby is likely to increase consumer welfare.
178. On above-cost predation, see Aaron Edlin, Stopping Above-Cost Predatory Pricing, 111 YALE L.J. 941 (2002)
179. and Einer Elhauge, Why Above-Cost Price Cuts to Drive Out Entrants are not Predatory - and the Implications for Defining Costs and Market Power, 112 YALE L.J. 681 (2003).
180. See supra Part II.A. This test is consistent with the one suggested by G. Federico, see supra note 8. The author proposes to estimate the average price that a competitor would have to practice for any given quantity to make customers indifferent between the rival's and dominant company's offers. Then, such price schedule should be compared with an efficient entrant's average variable and average total cost schedules, over the quantity range that a rival seller could credibly contest from the dominant firm.
181. Contra, on the basis of the same arguments on which the Tomra decision is grounded, see Maier-Rigaud, above n 89.
182. In the economic literature, some further predation-style cost-tests have been suggested. In particular, it has been proposed to estimate the discount that should be offered by a hypothetical rival wishing to realize a small but significant increase of its sales, equal to five percent of the customer's demand. See Heimler, supra note 9.
183. However, there is no reason to limit the analysis of the possibility to match the dominant undertaking's offer to a small but significant increase in sales, if there are no additional factors-such as brand loyalty and capacity constraints-preventing rivals from competing for larger portions of the customer's demand. Other antitrust scholars have proposed to define a kind of "battlefield" by aggregating the contestable portion of all customers' demand, to apply a price test in this share of the market. The practice would be unlawful when the average incremental price was higher than the average incremental cost. See Amelia Fletcher, "Loyalty Rebates Schemes: Towards an Effects-Based Approach?," Second conference of the Association of Competition Economics, Siena, 2-3 December 2004.
184. However, this test would not prohibit the practice when the incremental price, on average, was not predatory, even if the system prevented access to the most efficient distribution channels. Others have underlined that, in the case of rollback discounts, above a certain threshold the incremental sales necessary to achieve the target set by the dominant undertaking do not contribute to increasing its total profits. If the assured base of sales exceeds this threshold, the dominant company would earn more profits without the discount system and, therefore, the incremental price should be deemed predatory. To the contrary, if the assured base of sales does not exceed the above-mentioned threshold, it should be presumed that the practice is lawful, because the price of the contestable portion of the customer's demand would be higher than an appropriate cost measure. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
185. The application of predation-style tests to bundled (and loyalty) rebates has been criticized by Brennan, supra note 28, according to whom rebate programs should be assessed on the basis of whether they raise the price of a complement, such as retailing or distribution.
186. It should be noted that, under conditions of stochastic demand, a progressive discount system may be more effective and harmful than a scheme providing for a single turnover threshold. The latter might turn out to be too high or too low in comparison with the customer's requirements, whereas a grid of progressive thresholds linked to increasing discount rates would provide incentives to purchase from the dominant company regardless of possible demand fluctuations.
187. In the case of phenomena such as brand loyalty and product differentiation, a high and stable share of purchases from the dominant firm in past years may be indicative of the existence of a significant assured base of sales. To get more detailed information, buyers may be asked to specify whether they typically require a minimum amount of a certain good, their maximum willingness to pay to secure that amount, and which factors specifically give rise to this particular need. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10, at 117-20.
188. At the retail level, loyalty programs (such as, for instance, the so-called frequent flyer programs in the airline industry, or those offered by grocery stores, book stores, and coffee shops) are ubiquitous and commonly used both in competitive and concentrated industries. Although such programs may benefit both firms and final consumers - see David Bell & Rajiv Lal, "The Impact of Frequent Shopper Programs in Grocery Retailing," Review of Marketing Science Working Paper (2002), available at www.bepress.com/roms/vol2/iss1/paper1 (visited 30 January 2008 -economic analysis has shown that they may create switching costs and make the demand for a firm's product more inelastic, thus resulting in higher equilibrium prices and lower consumer welfare.
189. See, e.g., Klemperer, supra note 13
190. see also Byung-Do Kim, Mengze Shi & Kannan Srinivasan, Reward Programs and Tacit Collusion, 20 MARKETING SCI. 99 (2001).
191. However, even though such programs may in theory conflict with competition rules - in particular in Europe: See Adrian Emch, Frequent Flyer Programmes Under Article 82 EC - Is the Sky the Only Limit?, 30 World Competition 645 (2007) - they generally have not raised antitrust concerns, as shown by the analysis of U.S. and EC case law. In any case, most of these programs are not based on customer-specific thresholds, so that they do not constitute loyalty discounts as defined in this paper.
192. These measures of cost are influenced by the time period over which the cost structure is observed. In the long term, many fixed costs should be considered variable. Similarly, costs included among the avoidable and incremental ones tend to increase in the long term.
193. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
194. For a comparison between the two approaches, see Amelia Fletcher, "The Reform of Article 82: Recommendations on Key Policy Objectives," speech at the Competition Law Forum, Brussels, 15 March 2005, available at www.oft.gov.uk/news/speeches/2005/0205 (visited 30 January 2008).
195. See Discussion Paper, 154.
196. As already noted, the European Courts' case law does not provide clear indications as to the critical duration of the reference period. The legal and economic literature is characterized by a similar uncertainty: according to some commentators, there is no reason to grant rebates after the expiration of the terms for the payment. See Christian Roques, CFI Judgment, case T-203/01, Manufacture Francaise des Pneumatiques Michelin v. Commission, 25 EUR. COMPETITION L. REV. 688 (2004). Others have held that the critical duration of reference period cannot be established ex ante. It should be determined case-by-case on basis of the frequency of orders in the sector concerned, to verify whether system determines the bundling of a significant number of purchases.
197. See Gyselen, supra note 40.
198. See Maier-Rigaud, supra note 6.
199. Indeed, the amounts of purchases made during a particular time period vary across industries. In addition, in some markets purchases are concentrated in particular periods of the year.
200. However, the relationship between the duration of the reference period and the potential anticompetitive effects of a loyalty discount system is not clear. If demand is volatile and fluctuates over time, a short reference period may allow the supplier to adapt the threshold level to market changes frequently, so as to preserve the incentive function of the system. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
201. In this respect, it should be noted that, pursuant to the horizontal mergers guidelines, as a general rule entry would be a competitive constraint sufficient to prevent the exercise of market power if it occurs within two years. See Guidelines on the Assessment of Horizontal Mergers under the Council Regulation on the Control of Concentrations between Undertakings (2004/C 31/03), 74
202. see also Guidelines on Vertical Restraints (2000/C 291/01), 126, according to which "[i]n as far as effective entry, which would prevent or erode the supra-normal profits, is likely to occur within one or two years, entry barriers can be said to be low."
203. This part of the test should not necessarily follow the analysis of the possibility to match the dominant company's offer. If it appears clear from the outset that only a limited share of market demand is covered by the contested discount system, the case should be dismissed.
204. Under this perspective, discount systems providing for thresholds progressively rising year by year may be particularly harmful to competition, because they tend to saturate the buyers' distribution capacity.
205. Direct sales may be a realistic option if buyers are large, scope economies and number of sales points are not very important, and distribution activities provide little added value. See Majumdar, Bishop, Ridyard, Williams & Akgun, supra note 10.
206. See Tom, Balto & Averitt, supra note 11
207. Einer Elhauge, Defining Better Monopolization Standards, 56 STAN. L. REV. 253 (2003).
208. "The Exclusion of Competition for Hospital Sales Through Group Purchasing Organizations" (2002), available at http://www.law.harvard.edu/faculty/elhauge/pdf/gpo_report_june_02.pdf (visited 30 January 2008)
209. Einer Elhauge, "Antitrust Analysis of GPO Exclusionary Agreements," Comments regarding hearings on health care and competition law and policy - Statement for DOJ/FTC Hearing on GPOs, 26 September 2003, available at http://www.law.harvard.edu/faculty/ elhauge/pdf/statement_ftcdoj.pdf (visited 30 January 2008).
210. See Spector, supra note 16
211. Elhauge, "Antitrust Analysis of GPO Exclusionary Agreements," supra note 109.
212. The evolution of market shares of the dominant undertaking and its rivals before and after the implementation of the practice, as well as the entry of new operators or the exit of existing competitors, could provide important - although not decisive - indications as to the actual extent of barriers to entry and expansion raised by the discount system.
213. As underlined by the OECD, "disadvantaging or even excluding competitors through the use of fidelity discounts need not harm consumers." See OECD, supra note 9, at 8
214. see also Kallaugher & Sher, supra note 17, proposing a three-stage standard aimed at verifying whether: (i) the discount system generates appreciable switching costs; (ii) these costs constitute a significant barrier to entry or expansion of competitors; and (iii) these barriers to entry or expansion are likely to cause anticompetitive effects in terms of higher prices or a limitation of the choices available to consumers.
215. This brings us a step forward. The main utility of the dominance test in EC competition law lies in the fact that a screening based on a legal concept similar to the economic notion of market power avoids the prohibition of many forms of procompetitive unilateral conduct, thus reducing the number of false positives. The intent to prohibit certain unilateral practices by nondominant companies led in some cases to an overtly broad interpretation of the notion of dominance or vertical agreement: In this last regard, see Roberto Pardolesi, "Intese restrittive della libertà di concorrenza," and Gianluca Faella, "Le intese verticali," in Frignani & Pardolesi (eds), supra note 3, at 25, 105. If we were confident in the analysis of the actual or likely impact of a practice on the market, there would be no need for a separate dominance test. Antitrust authorities and courts could focus their assessment on the economic effects of a practice, with a view to applying competition rules to all genuinely anticompetitive unilateral conduct, without having to analyze first whether a dominant position exists.
216. See supra Part III.E.
217. See Kallaugher & Sher, supra note 17
218. Ekaterina Rousseva, The Concept of "Objective Justification" of an Abuse of a Dominant Position: Can it Help to Modernise the Analysis under Article 82 EC?, 2(2) THE COMPETITION L. REV. 27 (2006).
219. See supra Part IV.
220. See supra Part II.C.
221. Maier-Rigaud, supra note 89.
222. See Frank H. Easterbrook, When Is it Worthwhile to Use Courts to Search for Exclusionary Conduct?, 2003 COLUM. BUS. L. REV. 345 (2003).