As reported here, Johnson & Johnson scored a major victory last week in a case challenging some of its discounting practices. The jury concluded that J&J had not engaged in monopolization of the market for “trocars,” which are sharp cylindrical devices used in endoscopic surgery. Plaintiff Applied Medical Resources Corp., which sells trocars that compete with J&J’s, had complained that J&J’s discounting practices excluded it from the trocars market. Specifically, Applied complained about J&J’s “bundled discounts,” which provided a discount on J&J’s sutures (surgical stitches) to those hospital purchasers that also bought a certain percentage of their trocars from J&J.
Discounts, of course, are usually not illegal unless (1) they result in below-cost prices and (2) there is a likelihood that the discounter will be able to recoup its losses by charging supracompetitive prices in the future (when other firms have been driven out of the market by its discounting). This lenient rule makes sense. Any single-product discount resulting in an above-cost price could be matched by a firm that was as efficient as the discounter, so any firm excluded by an above-cost discount must be a less efficient rival. Because antitrust law favors lower prices and generally doesn’t (and shouldn’t) seek to protect less efficient rivals, single-product discounts resulting in above-cost prices are OK.
Bundled discounts — package discounts conditioned upon purchasing products from multiple product markets — are a different competitive animal. At least in theory, a bundled discount that results in above-cost pricing (for the bundle) may exclude equally efficient rivals that sell a narrower line of products. Consider, for example, a manufacturer (A) that sells both shampoo and conditioner and competes against another manufacturer (B) that sells only shampoo. B, the more efficient shampoo manufacturer, can produce a bottle of shampoo for $1.25. It costs A $1.50 to produce a bottle of shampoo and $2.50 to produce a bottle of conditioner. If purchased separately, A charges $2.00 for shampoo and $4.00 for conditioner ($6.00 total), but if the consumer purchases both products at once, A will sell the combination for $5.00. That $1.00 bundled discount results in a price that is $1.00 greater than A’s cost for the two products ($4.00). Nonetheless, the above-cost bundled discount could exclude B. B could stay in the market only if it charged no more than $1.00 for shampoo (so that a consumer’s total price of B’s shampoo and A’s conditioner would not exceed $5.00, A’s package price), but B’s marginal cost of producing shampoo is $1.25. Accordingly, A’s bundled discount could eliminate B as a competitor even though B is the more efficient producer and even though A’s discounted price is above its cost of producing the bundle.
[NOTE: This example is admittedly somewhat artificial. I use it because it appears (using slightly different numbers) in one of the leading bundled discounts decisions.]
Applied’s original claim was that it was excluded by J&J’s bundled discount even though that discount resulted in an above-cost price for the sutures/trocars bundle. Because Applied does not make sutures (an item on which J&J enjoys significant market power and, consequently, high profits), it could compete with J&J only by offering the full amount of J&J’s trocars/sutures discount on Applied’s narrower (trocars-only) product line. Doing so would require it to price below cost and would drive it out of business.
Unfortunately for Applied, there were some inconvenient facts. First, J&J’s bundled discounts were requested by the hospitals themselves (acting through buying groups called “Group Purchasing Organizations”) during a competitive bidding process. Second, J&J’s bundled discounts were primarily designed to compete with its chief rival, Tyco Corp. (formerly U.S. Surgical), which sells a complete line of trocars and sutures and was offering its own bundled discounts in competition with J&J’s. This “bundle-to-bundle” competition was fierce and consumer-friendly. Finally, J&J responded to Applied’s complaints about the bundles by “carving out” Applied’s products, so that hospitals could count purchases of Applied trocars — but not Tyco trocars — toward the purchases required to earn the discount on J&J sutures. Without doubt, J&J’s bundled discounts were aimed at competing with its chief rival, Tyco, and resulted in lower prices for hospitals.
While this case seems like a no-brainer, that didn’t stop Applied from jumping on the LePage’s bandwagon and going after $54 million in purported (pre-trebled) damages. Fortunately, this jury was smart enough to recognize that the discounts at issue were ultimately good for consumers. Who knows what will happen next time.
It’s high time for the courts to develop some clear guidelines — including manageable safe harbors — for evaluating bundled discounts. I have suggested an easily administrable, balanced approach that would approve most such discounts but would impose liability for truly exclusionary practices. As Josh has noted, Herbert Hovenkamp’s new book (The Antitrust Enterprise) suggests another workable approach. Dan Crane has proposed another. Hopefully, courts will soon latch on to one of these proposals. The very fact that the Applied Medical v. J&J case got as far as it did suggests that the LePage’s-driven status quo is intolerable.
Thanks for raising the tying point, Keith. I would echo Josh’s points and note an additional hurdle a tying plaintiff would have to establish — that there was, in fact, a de facto tie-in.
Whereas tying involves a refusal to sell one product (the tying product “A”) without another (the tied product “B”), bundled discounts merely “encourage” — not require — purchases of both products. Now, the degree of “encouragement” could be so great as to constitute a de facto tie. For example, if a lamp monopolist charges $100 for a lamp but will sell a lamp/lampshade combo for $5, he has effectively tied the two products, even though the lamp is technically available separately. It thus makes sense to conclude that bundled discounts may, in certain circumstances, amount to de facto tie-ins.
But in what circumstances? Presumably, the plaintiff would have to show that the bundled discount is so “effective” that it has an effect similar to an outright refusal to sell the packaged products separately. The Areeda-Hovenkamp treatise suggests that courts should focus on the “proportion of separate purchasesâ€? –i.e., the percentage of purchases of “tiedâ€? product B, by purchasers who buy both the defendant’s “tyingâ€? product A and any seller’s product B, that are outside the defendant’s package. The treatise suggests that if separate purchases exceed 10 percent, there should be no illegal tie.
{In addition, the treatise posits three “safe harborsâ€? where a tie should not be found even if separate purchases are less than 10 percent. The three safe harbors are where: (1) the separate price of tying product A is less than or equal to the market price of A; (2) the “package price” of tied product B (i.e., the price of the package less the separate price of tying product A) exceeds or equals the market price of B; or (3) the package price of B exceeds or equals the marginal cost of rivals’ B. In the first two situations, there can be no legitimate concern that the defendant has jacked up the separate price of tying product A in order to induce buyers to purchase the package. In the third, there is no injury to competition because the defendants’ rivals could compete by reducing their prices to competitive levels.}
Of course, the Areeda-Hovenkamp proposal is not the law; it’s just a suggestion for courts. But courts analyzing bundled discounts as tie-ins are going to have to analyze whether a de facto tie-in has occurred, and they’re likely to follow some test similar to the Areeda-Hovenkamp suggestion. My guess it that most bundled discounts plaintiffs could not make the showing necessary to establish a de facto tie-in.
As Josh observes, cases like LePage’s give them an easier option.
Keith hints at what I believe to be one of the most problematic aspects of cases like LePage’s. The alternative tying and exclusionary theories of this type of bundling or other vertical contracts (like slotting!) require substantial foreclosure and proof of anticompetitive effect. These are though hurdles to overcome, and should be, such that we do not chill pro-competitive competition for distribution.
But hey! Why not just re-tool the complaint as a bundled discounts/LePage’s style case and get past summary judgment without evidence of either? I won’t comment on the JJ case specifically, but it does make sense to bring a case on a bundled discount theory if you cannot show the requisite foreclosure or competitive effects. In the case of bundle on bundle competition, these requirements are not likely to be met.
As Keith points out, the exclusionary theories do not require proof of below cost pricing. But this is precisely why it is so essential to sensible antitrust policy that plaintiff’s demonstrate substantial foreclosure when alleging exclusionary distribution contracts facilitate monopolization. My post, which Thom kindly links to above, reviewing Hovenkamp’s treatment of this issue touches on this issue in a bit more detail.
Excellent post, Thom. You address the predation arguments against bundled discounts. What about the theory that sometimes (perhaps not here) they are a de facto tie (a theory of liability that doesn’t depend on below cost pricing)?