|This document is available in three formats: this web page (for browsing content), PDF (comparable to original document formatting), and Word. To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.|
Statement of Warren S. Grimes
Irving D. and Florence Rosenberg Professor of Law
Southwestern Law School
Los Angeles, California
Loyalty Discounts on Strongly Branded Goods
November 20, 2006
Department of Justice and Federal Trade Commission
Single-Firm Conduct and Antitrust Law
FTC: Comments Regarding Section 2 Hearings, Project No. P062106
Loyalty Discounts on Strongly Branded Goods
A seller with market power may use that market power strategically to disadvantage rivals. One form of strategic conduct is the loyalty discount. A loyalty discount is a reduction in price contingent on the buyer purchasing a minimum percentage of the buyer's needs from the seller. In the United States, such discounts have been challenged under the Sherman Act with varying results and modes of analysis.
The seller's market power may come in the form of a unique product that, with or without patent protection, constitutes a product market by itself. But the power that allows a seller to gain a strategic advantage may simply be, in whole or in part, the power of a strong brand. In this paper, I assess the anticompetitive potential of a loyalty discount that is anchored with a strong brand and is employed to gain advantage over sellers of a weakly branded product.(1)
I. Competitive Analysis of a Loyalty Discount Encompassing a Strongly and Weakly Branded Product
A seller may offer a five percent discount to a buyer agreeing to purchase 80% of its needs for a particular product from the seller. The Eighth Circuit found this sort of loyalty discount unobjectionable in Concord Boat Corp. v. Brunswick Corp., where the court indicated that such discounts should be a Sherman Act violation only if they meet the requirements for predatory pricing (below cost pricing and a reasonable probability of recoupment).(2) In the Third Circuit's 2003 en banc decision in Le Page's, Inc. v. 3 M Corp.,(3) the Third Circuit held that the 3 M Corporation's loyalty discount involving Scotch brand tape and other products violated the Sherman Act. The en banc Third Circuit declined to require a showing of predatory pricing, focusing instead on whether 3M's loyalty discount operated to prevent an equally efficient rival from matching its terms.
Loyalty discounts may be used in a variety circumstances. An anticompetitive result is probable when a seller offers a loyalty discount on a strong brand (sold at a high margin) and a weak brand (sold at a low margin), and seeks to use the loyalty discount to disadvantage rival sellers of weak brands. Under such circumstances, application of a predatory pricing standard would tolerate conduct that is unambiguously anticompetitive.
Suppose that Acme Corporation has 80% of the market for widgets and a strong Premium brand that accounts for almost all of these sales. Acme, however, also sells a weak or private label brand in direct competition with one or more rival manufacturers of weak brands. Acme offers a 20% discount on a retailer's purchase of both Premium brand and its private label widgets if the retailer purchases 90% of its requirements for widgets from Acme.
This offer should be attractive to retailers. The seller of a strong brand enjoys vertical market power that allows a large factory mark up. Borrowing from Robert Steiner's insights, a supplier's vertical market power is a product of a strong consumer demand for a particular product or brand.(4) This vertical power usually exists along side the traditional horizontal market power, but not in every case. When vertical power exists, a consumer will switch stores in order to obtain the strong brand at the desired price.
Vertical market power is measured by a ratio of retailer and producer margins.(5) Typically, as the retailer's margin goes up, the producer's margin goes down, and vice versa. As Steiner concedes, this two stage vertical market power measure is an oversimplification of real markets that can have multiple players involved in distribution. However, measuring the margins of both producers and retailers offers a much more principled, accurate, and instructive model than the current single stage analysis which looks only at horizontal market power.(6) In the widget example, the retailer is pressured to carry the Premium widget brand because customers will look for this brand in the store and may shop elsewhere if they do not find it. But the retailer has a strong incentive to also carry the weak brand because of the higher retail mark up on such brands. Acme's loyalty discount will be attractive because the retailer can simply switch its purchasing from a rival's weak brand to Acme's weak brand and still enjoy an equivalent high retail markup. The retailer also receives a discount on purchases that it already makes for the Premium brand of widgets (the bulk of its widget purchases).
Why would Acme offer this bundled discount package rather than simply discounting the price of its weak brand of widgets? Part of the answer is that by including a discount on Premium brand widgets, Acme is offering a package that its rivals cannot match. The combined horizontal and vertical market power of Premium brand widgets cannot be matched by Acme rivals that lack a strong brand of their own.
From Acme's point of view, there is another advantage to this loyalty discount. Large retailers may prefer not to become dependent on a single supplier for an important product. Such buyers would have a strategic preference to support at least one rival of Acme so that the buyer has choices in future purchases of widgets. This strategic preference for choice means that the buyer will continue doing business with Acme's rival, even if Acme has matched the rival's price. In competing for the weak brand business, Acme will likely lose a price matching game unless it can come up with an offer that the rival cannot match. If Acme prices its weak brand widgets below cost, it could win a price-matching game, but subject itself to predatory pricing claims. To avoid this risk, Acme employs a loyalty discount that includes a rebate on purchases of high volume Premium brand widgets. Lacking a strong brand of its own, the rival cannot match the offer because its weak brand has a lower sales volume and sells at a much lower margin.
The conclusions drawn from this example would appear to apply to the loyalty discount at issue in the 3M case. The purpose of 3M's inventory-forcing discount was apparently to increase its sales of private label tape, not to increase sales of its dominant Scotch brand. 3M apparently also hoped to increase sales of other (probably weakly branded) products that were included in the bundle of goods subject to the loyalty discount.
One can find parallel cases such as SmithKline Corp. v. Eli Lilly & Co.(7) that fit the facts of the Acme example. SmithKline involved at least one strong brand or unique product that the purchaser needed in its inventory. The seller appeared to use this "must-buy" item to anchor its loyalty discount and force purchasers to also purchase one or more weaker brands. The buyers were not retailers but hospitals. This may alter a dual stage analysis, but the fundamental anticompetitive effect--that equally efficient rival sellers lacking equivalent strong brands cannot match the offer--remains the same.
II. Efficiencies as a Defense for Loyalty Discounts
In its amicus brief filed with the Supreme Court, the Department of Justice suggested that efficiencies could explain and justify some loyalty discounts.(8) As in other Sherman Act bundling or foreclosure cases, a defendant should be allowed to introduce evidence of efficiencies that might tip the balance in its favor. Because price discounts allow a seller to increase sales, there is a possibility that scale efficiencies will be enhanced. Scale efficiencies, however, do not appear to have played a significant role in any of the litigated cases.
Some commentators have pushed the argument that bundling conduct, such as might be fostered by a loyalty discount, is a procompetitive way to finance research and development when the tying product is patented.(9) This argument, if accepted, would result in a troubling and potentially far reaching erosion of the Sherman Act. Many tying products will involve a patent in at least a peripheral way. Many of these patents may have little or no market value. Even if the tying product is directly covered by a valuable patent, the additional gain from tying will depend on many external variables (for example, the state of competition in the tied product market) that have nothing to do with the underlying value of the patented product.
Of course, any profit stemming from an exercise of market power, even if no patent is involved, could theoretically be used to finance R & D. If this line of logic were accepted, it would follow that all exercises of monopoly power could be justified as an incentive for, and a subsidy to, innovation.
III. What Rules Should Govern Loyalty Discounts?
To summarize, in a loyalty-discount contest between a seller of both a strong brand and weak brand and a rival seller of a weak brand, the rival will have a two-fold handicap: (1) the rival lacking a strong brand is in all likelihood selling a small volume of goods to a given retailer, so that any loyalty discount must be larger in relative terms to match a loyalty discount offered by the strong brand seller; and (2) the rival is, even before a discounting war begins, likely to be selling its product at a substantially lower margin than the strong brand seller.
The two stage Steiner analysis demonstrates that a loyalty discount is suspect when employed by a manufacturer with a strong brand. At least for this fact pattern, a rule that loyalty discounts be condemned only when predatory pricing is demonstrated is bad antitrust policy. It has little to do with market realities. A seller possessing both horizontal and vertical market power will be selling a strong brand at a price nowhere near marginal cost (or average variable cost), a widely accepted threshold for predatory pricing. Such a seller can employ a loyalty discount to raise rivals costs or drive them from the market while still maintaining as much as possible of the high margin that the combined market power allows.
In the context of cases such as 3M and SmithKline, where the defendant's conduct is of the inventory-forcing nature, there is no reason that the rules governing this conduct should be different than those governing tying or exclusive dealing. The loyalty discount just becomes the tool for enforcing tying or exclusive dealing. Courts should examine whether the defendant has market power (horizontal and vertical) and whether that market power is being abused through forcing behavior. Because the "forcing" is in the form of a discounted price, the plaintiff cannot win by simply showing that the discount makes it more difficult for a rival to compete. The plaintiff must show that the discount is designed so that it could not be matched by an equally efficient rival.
1. The analysis presented here is drawn from a chapter entitled Making Antitrust/Intellectual Property Policy in the United States: Requirements Tie-Ins and Loyalty Discounts, in Handbook on Intellectual Property and Competition Lawto be published by Edward Elgar Publishing.
2. 207 F.3d 1029 (8th Cir. 2001).
3. 324 F.3d 141 (3d Cir. 2003)(en banc), cert. denied, 540 U.S. 807 (2003).
4. Robert L. Steiner, The Evolution and Applications of Dual-Stage Thinking, 49 Antitrust Bull. 877, 890-92 (2004).
6. As Steiner also acknowledges, there are industry structures in which the margins of retailers and manufacturers are positively related -- in which total market power in the category may be very low to the benefit of consumers or very high to their detriment. For example, Steiner has written about a case in which collusion between a manufacturer of a strong brand and its retailers resulted in high margins for both the manufacturer and the retailers. Robert L.Steiner, Exclusive Dealing + Resale Price Maintenance: A Powerful Anticompetitive Combination, 33 Sw. U. L. Rev. 447 (2004).
7. 575 F.2d 1056 (3d Cir. 1978).
8. Le Page's, Inc. v. 3 M Corp., Amicus Brief for the United States in the Supreme Court of the United States, at 12, available at .
9. Benjamin Klein & John Wiley, Competitive Price Discrimination As An Antitrust Justification for Intellectual Property Refusals to Deal, 70 Antitrust L. J. 599 (2004)(arguing that licensing restrictions and tie-ins involving intellectual property should be treated leniently because of the innovation gains that they bring).