Cascade Health Solutions v. PeaceHealth

502 F.3d 895, 2007 U.S. App. LEXIS 21075, 2007 WL 2473229

Opinion

GOULD, Circuit Judge:

McKenzie-Willamette Hospital (“McKenzie”) filed a complaint in the district court against PeaceHealth asserting seven claims for relief. Five of the claims arose under the federal antitrust laws: monopolization, attempted monopolization, conspiracy to monopolize, tying, and exclusive dealing. The other two claims arose under Oregon state law: price discrimination and intentional interference with prospective economic advantage.

Before trial, the district court granted summary judgment in favor of Peace-Health on McKenzie’s tying claim. After a two-and-a-half-week trial, the jury rendered a verdict in favor of PeaceHealth on McKenzie’s claims of monopolization, conspiracy to monopolize, and exclusive dealing. However, the jury found in favor of McKenzie on McKenzie’s claims of attempted monopolization, price discrimination, and tortious interference. The jury awarded McKenzie $5.4 million in damages, which the district court trebled for a final award of $16.2 million. The district court also awarded McKenzie $1,583,185.57 in attorneys’ fees, costs, and expenses.

We vacate the jury’s verdict in favor of McKenzie on the attempted monopolization, price discrimination, and tortious interference claims, and we vacate the district court’s summary judgment in favor of PeaceHealth on the tying claim. We also vacate the district court’s award of attorneys’ fees, costs, and expenses. We remand for further proceedings.

I

A

McKenzie and PeaceHealth are the only two providers of hospital care in Lane County, Oregon. The jury found and, for the purposes of this appeal, the parties do not dispute, that the relevant market in this case is the market for primary and secondary acute care hospital services in Lane County. Primary and secondary acute care hospital services are common medical services like setting a broken bone and performing a tonsillectomy. Some hospitals also provide what the parties call “tertiary care,” which includes more complex services like invasive cardiovascular surgery and intensive neonatal care.

In Lane County, PeaceHealth operates three hospitals while McKenzie operates one. McKenzie’s sole endeavor is McKenzie-Willamette Hospital, a 114-bed hospital that offers primary and secondary acute care in Springfield, Oregon. McKenzie does not provide tertiary care. In the time period leading up to and including this litigation, McKenzie had been suffering financial losses, and, as a result, merged with Triad Hospitals, Inc. 1 so that it could add tertiary services to its menu of care.

The largest of PeaceHealth’s three facilities is Sacred Heart Hospital, a 432-bed operation that offers primary, secondary, and tertiary care in Eugene, Oregon. PeaceHealth also operates Peace Harbor Hospital, a 21-bed hospital in Florence, Oregon and Cottage Grove Hospital, an 11-bed hospital in Cottage Grove, Oregon. In Lane County, PeaceHealth has a 90% market share of tertiary neonatal services, a 93% market share of tertiary cardiovascular services, and a roughly 75% market share of primary and secondary care services.

To understand the antitrust issues in this case, it is necessary to appreciate the structure of the market in which this case arises. The market for hospital services and medical care is complex. However, based on the record, there appear to be three major participants in the market for hospital services: hospitals, insurers, and patients. Hospitals, like those operated by PeaceHealth and McKenzie, provide services to patients and sell services to insurers. Insurers are usually commercial health insurance companies that seek to buy medical services from hospitals on the best terms possible. The insurers in turn sell insurance services to patients and employers. Patients buy health insurance from insurers (often through their employers) and sometimes buy services from hospitals.

In the transaction between a hospital that sells care services and an insurer that buys care services, the price agreed upon is often referred to as a “reimbursement rate.” For example, in a hospital-insurer contract, the agreed upon price might be “a 90% reimbursement rate.” A 90% reimbursement rate price means that, when the insurer must purchase services from the hospital, the insurer gets a 10% discount off the hospital’s regular price, also called the charge master or list price. It follows that hospitals prefer high reimbursement rates and insurers prefer low reimbursement rates, as each group pursues its own economic interest.

B

Before trial, the district court granted summary judgment to PeaceHealth on McKenzie’s tying claim, concluding that McKenzie had not presented any evidence that PeaceHealth “coerced” insurers into purchasing primary and secondary services from it in order for the insurers to obtain tertiary services. The district court let the remainder of McKenzie’s claims proceed to trial before a jury. On McKenzie’s monopolization and attempted monopolization claims, McKenzie’s primary theo ry was that PeaeeHealth engaged in anticompetitive conduct by offering insurers “bundled” or “package” discounts. McKenzie asserted that Peace-Health offered insurers discounts of 35% to 40% on tertiary services if the insurers made PeaeeHealth their sole preferred provider for all services— primary, secondary, and tertiary. McKenzie introduced evidence of a few specific instances of PeaceHealth’s bundled discounting practices.

For example, in 2001, PeaeeHealth was the only preferred provider of hospital care under the preferred provider plan (“PPP”) of Regence BlueCross BlueShield of Oregon (“Regence”). 2 At that time, Re-gence was paying PeaeeHealth a 76% reimbursement rate for all of PeaceHealth’s medical services, including primary, secondary, and tertiary services. Around that time, pursuant to McKenzie’s request, Regence considered adding McKenzie to the PPP as a preferred provider of primary and secondary services. When Re-gence’s contract with PeaeeHealth came up for its annual renewal, Regence solicited two proposals from PeaeeHealth. Under one proposal, PeaeeHealth would remain the only preferred provider. Under the other proposal, McKenzie would be added as a preferred provider. Peace-Health offered an 85% reimbursement rate for all services if it remained Regence’s sole preferred provider of primary, secondary, and tertiary services, and a 90% reimbursement rate if McKenzie was added as a preferred provider of primary and secondary services. Regence thereafter declined to • include McKenzie as a preferred provider.

That same year, McKenzie sought and received admission as a preferred provider of primary and secondary sendees under the preferred plan offered by Providence Health Plan (“Providence”). Until then, PeaeeHealth was the only preferred provider of primary, secondary, and tertiary services in the Providence preferred plan. Upon McKenzie’s admission as a preferred provider, PeaeeHealth increased its reimbursement rate with Providence from 90% to 93%. The evidence showed that insurers who made PeaeeHealth their exclusive preferred provider across all services, thus purchasing from PeaeeHealth a full complement of primary, secondary, and tertiary services, paid lower reimbursement rates than insurers who purchased tertiary services from PeaeeHealth, but at least some primary and secondary services from McKenzie.

The jury rejected McKenzie’s claims of monopolization,. conspiracy to monopolize, and exclusive dealing in its verdict for PeaeeHealth on those issues. However, the jury found in favor of McKenzie on its claims of attempted monopolization, price discrimination, and tortious interference. The jury awarded damages of $5.4 million on each claim. McKenzie elected to pursue its remedy under federal law on the attempted monopolization claim, so the district court, pursuant to § 4(a) of the Clayton Act, 15 U.S.C. § 15(a), trebled the jury’s $5.4 million award on the attempted monopolization claim for a final damage award of $16.2 million. The district court denied PeaceHealth’s motion for judgment as a matter of law on the claims the jury decided in McKenzie’s favor, and also awarded McKenzie $1,583,185.57 in attorneys’ fees and costs.

PeaeeHealth appeals the judgment entered pursuant to the jury verdict in McKenzie’s favor. McKenzie cross-appeals the district court’s grant of summary judgment to PeaceHealth on McKenzie’s tying claim. Both parties appeal the district court’s award of attorneys’ fees and costs.

II

We first address PeaceHealth’s appeal of the jury verdict in McKenzie’s favor on McKenzie’s claims of attempted monopolization, price discrimination, and tortious interference.

We address initially the attempted monopolization claim. Section 2 of the Sherman Act makes it illegal to “attempt to monopolize ... any part of the trade or commerce among the several States, or with foreign nations.” 15 U.S.C. § 2. “[T]o demonstrate attempted monopolization a plaintiff must prove (1) that the defendant has engaged in predatory or anticompeti-tive conduct with (2) a specific intent to monopolize and (3) a dangerous probability of achieving monopoly power.” Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 456, 113 S.Ct. 884, 122 L.Ed.2d 247 (1993); Amarel v. Connell, 102 F.3d 1494, 1521 (9th Cir.1996). 3

PeaceHealth’s appeal centers on the first element of the Spectrum Sports test, the conduct element. Anticompeti-tive conduct is behavior that tends to impair the opportunities of rivals and either does not further competition on the merits or does so in an unnecessarily restrictive way. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 605 n. 32, 105 S.Ct. 2847, 86 L.Ed.2d 467 (1985). PeaceHealth contends that we should vacate the jury’s verdict because the district court incorrectly instructed the jury about when bundled discounting can amount to anticompetitive conduct. This leads us to consider at some length the phenomena of bundles and bundled discounts.

Bundling is the practice of offering, for a single price, two or more goods or services that could be sold separately. A bundled discount occurs when a firm sells a bundle of goods or services for a lower price than the seller charges for the goods or services purchased individually. See Daniel A. Crane, Mixed Bundling, Profit Sacrifice, and Consumer Welfare, 55 Emory L.J. 423, 425 (2006); David S. Evans. & Michael Salinger, Why Do Firms Bundle and Tie?, 22 Yale J. on Reg. 37, 41 (2005); Thomas A. Lambert, Evaluating Bundled Discounts, 89 Minn. L.Rev. 1688, 1693 (2005). As discussed above, PeaceHealth offered bundled discounts to Regence and other insurers in this case. Specifically, PeaceHealth offered insurers discounts if the insurers made PeaceHealth their exclusive preferred provider for primary, secondary, and tertiary care.

Bundled discounts are pervasive, and examples abound. Season tickets, fast food value meals, all-in-one home theater systems — all are bundled discounts. Like individual consumers, institutional purchasers seek and obtain bundled discounts, too. See, e.g., LePage’s Inc. v. 3M, 324 F.3d 141, 154 (3d Cir.2003) (en banc) (involving rebates offered by 3M tó retailers who purchased 3M’s full line of health care, home care, home improvement, stationary, retail auto, and “Leisure Time” products); Invacare Corp. v. Respironics, Inc., No. 1:04-CV-1580, 2006 WL 3022968, at *1 (N.D.Ohio Oct.23, 2006) (involving a medical device manufacturer who bundled the masks worn 'by persons with obstructive sleep apnea with the devices that blow air into the masks); Masimo Corp. v. Tyco Health Care Group, L.P., No. CV 02-4770, 2006 WL 1236666, at *9 (C.D.Cal. Mar.22, 2006) (involving rebates offered by Tyco to hospitals that purchased both Tyco’s oxi-metry and non-oximetry products together); J.B.D.L. Corp. v. Wyeth-Ayerst Labs., Inc., No. 1:01-CV-704, 2005 WL 1396940, at *3 (S.D.Ohio June 13, 2005) (involving, rebates offered by Wyeth to pharmacy benefit managers based on combined purchases of estrogen-replacement drugs, oral contraceptives, an antidepressant, an antibiotic, a calcium channel blocker, and a beta blocker), aff'd, 485 F.3d 880 (6th Cir.2007). The varied and pervasive nature of bundled discounts illustrates that such discounts transcend market boundaries. On the one hand, the world’s largest corporations offer bundled discounts as their product lines expand with the convergence of industries. 4 On the other hand, a street-corner vendor with a food cart — a merchant with limited capital — might offer a discount to a customer who -buys a drink and potato chips to complement a hot dog. The fact that such diverse sellers offer bundled discounts shows that such discounts are a fundamental option for both buyers and sellers. 5

Bundled discounts generally benefit buyers because the discounts allow the buyer to get more for less. 6 Lambert, supra, 89 Minn. L.Rev. at 1726 (suggesting that bundled discounts always provide some immediate consumer benefit in the form of lower prices); 3 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law ¶ 749b at 324 (Supp.2006) (explaining that “[t]he great majority of discounting practices are procompetitive” and “reflect hard bargaining”). Bundling can also result in savings to the seller because it usually costs a firm less to sell multiple products to one customer at the same time than it does to sell the products individually. United States v. Microsoft Corp., 253 F.3d 34, 87 (D.C.Cir.2001) (per curiam) (noting that “[bjundling obviously saves distribution and consumer transaction costs” and allows firms to “capitalize on certain economies of scope”); Crane, supra, 55 Emory L.J. at 430-33 (discussing how package discounts can create economies of scope and transaction costs savings). 7

Not surprisingly, the Supreme Court has instructed that, because of the benefits that flow to consumers from discounted prices, price cutting is a practice the antitrust laws aim to promote. See Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986) (“[CJutting prices in order to increase business often is the very essence of competition.”). Consistent with that principle, we should not be too quick to condemn price-reducing bundled discounts as anticompetitive, lest we end up with a rule that discourages legitimate price competition. See Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 234 (1st Cir.1983) (Breyer, J.).

However, it is possible, at least in theory, for a firm to use a bundled discount to exclude an equally or more efficient competitor and thereby reduce consumer welfare in the long run. See Richard A. Posner, Antitmst Law 236 (2d ed.2001); Bariy Nalebuff, Exclusionary Bundling, 50 Antitrust Bull. 321, 321 (2005). For example, a competitor who sells only a single product in the bundle (and who produces that single product at a lower cost than the defendant) might not be able to match profitably the price created by the multi-product bundled discount. See Ortho Diagnostic Sys., Inc. v. Abbott Labs., Inc., 920 F.Supp. 455, 467 (S.D.N.Y.1996). This is true even if the post-discount prices for both the entire bundle and each product in the bundle are above the seller’s cost. See Ortho, 920 F.Supp. at 467 (noting that “a firm that enjoys a monopoly on one or more of a group of complementary products, but which faces competition on others, can price all of its products above average variable cost and yet still drive an equally efficient competitor out of the market”). Judge Kaplan’s opinion in Ortho provides an example of such a situation:

Assume for the sake of simplicity that the case involved the sale of two hair products, shampoo and conditioner, the latter made only by A and the former by both A and B. Assume as well that both must be used to wash one’s hair. Assume further that A’s average variable cost for conditioner is $2.50, that its average variable cost for shampoo is $1.50, and that B’s average variable cost for shampoo is $1.25. B therefore is the more efficient producer of shampoo. Finally, assume that A prices conditioner and shampoo at $5 and $3, respectively, if bought separately but at $3 and $2.25 if bought as part of a package. Absent the package pricing, A’s price for both products is $8. B therefore must price its shampoo at or below $3 in order to compete effectively with A, given that the customer will be paying A $5 for conditioner irrespective of which shampoo supplier it chooses. With the package pricing, the customer can purchase both products from A for $5.25, a price above the sum of A’s average variable cost for both products. In order for B to compete, however, it must persuade the customer to buy B’s shampoo while purchasing its conditioner from A for $5. In order to do that, B cannot charge more than $0.25 for shampoo, as the customer otherwise will find A’s package cheaper than buying conditioner from A and shampoo from B. On these assumptions, A would force B out of the shampoo market, notwithstanding that B is the more efficient producer of shampoo, without pricing either of A’s products below average variable cost.

Id.; see also 3 Areeda & Hovenkamp, supra, ¶ 749a at 318-19 (Supp.2006) (providing a similar example). It is worth reiterating that, as the example above shows, a bundled discounter can exclude rivals who do not sell as great a number of product lines without pricing its products' below its cost to produce them. Thus, a bundled discounter can achieve exclusion without sacrificing any short-run profits. See Nalebuff, supra, 50 Antitrust Bull, at 339 (providing an example of exclusion accomplished with an increase in profits).

In this case, McKenzie asserts it could provide primary and secondary services at a lower cost than PeaceHealth. Thus, the principal anticompetitive danger of the bundled discounts offered by PeaceHealth is that the discounts could freeze McKenzie out of the market for primary and secondary services because McKenzie, like seller B in Judge Kaplan’s example, does not provide the same array of services as PeaceHealth and therefore could possibly not be able to match the discount Peace-Health offers insurers.

From our discussion above, it is evident that bundled discounts, while potentially procompetitive by offering bargains to consumers, can also pose the threat of anticompetitive impact by excluding less diversified but more efficient producers. These considerations put into focus this problem: How are we to discern where antitrust law draws the line between bundled discounts that are procom-petitive and part, of the normal rough-and-tumble of our competitive economy and bundled discounts, offered by firms holding- or on the verge of gaining monopoly power in. the-relevant market, that harm competition and are thus proscribed by § 2 of the Sherman-Act?

2

In this case, the district court based its jury instruction regarding the anticompeti-tive effect of bundled discounting on the Third Circuit’s en banc decision in LePage’s Inc. v. 3M, 324 F.3d 141 (3d Cir.2003) (en banc). In that case, the plaintiff, LePage’s, was the market leader in sales of “private label” (i.e., store brand) transparent tape. See id. at 144. As LePage’s market share fell and its profitability declined, it brought suit asserting that 3M, who manufactured Scotch tape, some private label tape, and many other products that LePage’s did not produce (like healthcare products and retail automotive products), leveraged its monopoly over Scotch brand tape to monopolize the private label tape market. Id. at 145, 154. Specifically, LePage’s alleged that 3M’s multi-tiered bundled rebate structure was anticompeti-tive. Id. at 145. The bundled rebate structure offered progressively higher rebates when customers increased purchases across 3M’s different product lines — discounts LePage’s could not offer because it did not sell the .same diverse array of products as 3M. See id. A jury found that 3M’s conduct violated § 2 of the Sherman Act and 3M appealed. Id.

The primary issue before the Third Circuit was whether 3M unlawfully maintained its monopoly power through the bundled discount program. See id. at 146-47. 3M argued that its bundled rebate structure was legal as a matter of law because it never priced below cost. Id. at 147. 3M relied heavily on the United States Supreme Court’s decision in Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 113 S.Ct. 2578, 125 L.Ed.2d 168 (1993). In Brooke Group, a primary-line price discrimination case brought under the Robinson-Patman Act, the Supreme Court held that, in a single product predatory pricing case, a plaintiff must prove (1) that its rival’s low prices were below an appropriate measure of its rival’s costs and (2) that its rival “had a reasonable prospect, or, under § 2 of the Sherman Act, a dangerous probability, of recouping its investment in below-cost prices.” Id. at 222, 224, 113 S.Ct. 2578. In LePage’s, the Third Circuit, in a 7-3 en banc decision, refused to apply Brooke Group’s below-cost pricing requirement to bundled discounting.

The Third Circuit first distinguished Brooke Group by noting that the defendant in that case was an oligopolist while 3M was a monopolist. LePage’s, 324 F.3d at 151-52. The court reasoned that while Brooke Group’s requirement of below-cost pricing with a probability of recoupment is appropriate when the defendant is an oli-gopolist who still faces competition when it tries to recoup the losses it suffered during the predation period, below-cost pricing and a probability of recoupment should not be required when the defendant is a monopolist whose behavior will be unconstrained by the market after it eliminates its lone rival. See id. The court in Le-Page’s also noted that the plaintiff in Brooke Group simply challenged the defendant’s pricing practices, not bundling accomplished through discounting. See id. at 151. The court reasoned that Brooke Group did not require below-cost pricing for any pricing practice,to be deemed exclusionary. See id.

The court noted that “[t]he principal anticompetitive effect of bundled rebates as offered by 3M is that when offered by a monopolist they may foreclose portions of the market to a potential competitor who does not manufacture an equally diverse group of products and who therefore cannot make a comparable offer.” Id. at 155. The Third Circuit concluded that the jury could reasonably have found that 3M used its monopoly in transparent tape along with its extensive catalog of other products to exclude LePage’s from the market and that 3M did not present any adequate business justification for its bundled discounting program. Id. at 164, 169. The court thus affirmed the jury verdict in LePage’s favor, id. at 169, even though LePage’s economist testified that LePage’s was not as efficient a tape producer as 3M, see id. at 177 (Greenberg, J., dissenting). 8

In this case, the district court used LePage’s to formulate its jury instruction. Specifically, the district court instructed the jury that

plaintiff ... contends that defendant has bundled price discounts for its primary, secondary, and tertiary acute care products and that doing so is anticompetitive. Bundled pricing occurs when price discounts are offered for purchasing an entire line of services exclusively from one supplier. Bundled price discounts may be anti-competitive if they are offered by a monopolist and substantially foreclose portions of the market to a competitor who does not provide an equally diverse group of services and who therefore cannot make a comparable offer.

As 3M did in LePage’s, PeaeeHealth argues that the jury instruction incorrectly stated the law because it allowed the jury to find that a defendant with monopoly power (or, in the case of an attempted monopolization claim, a dangerous probability of achieving monopoly power) engaged in exclusionary conduct by simply offering a bundled discount that its competitor could not match. The instruction did not require the jury to consider whether the defendant priced below cost. Le-Page’s, PeaeeHealth asserts, was wrongly decided because it allows the jury to conclude, from the structure of the market alone, that a competitor has been anticom-petitively excluded from the market. 9 We generally review jury instructions for abuse of discretion, but we review de novo whether jury instructions correctly stated the law. Voohries-Larson v. Cessna Aircraft Co., 241 F.3d 707, 713 (9th Cir.2001).

As the bipartisan Antitrust Modernization Commission (“AMC”) 10 recently noted, the fundamental problem with the Le- Page’s standard is that it does not consider whether the bundled discounts constitute competition on the merits, but simply concludes that all bundled discounts offered by a monopolist are anticompetitive with respect to its competitors who do not manufacture an equally diverse product line. Anti-trust Modernization Comm’n, Report and Recommendations 97 (2007) [hereinafter AMC Report], The LePage’s standard, the AMC noted, asks the jury to consider whether the plaintiff has been excluded from the market, but does not require the jury to consider whether the plaintiff was at least as efficient of a producer as the defendant. Id.; see also LePage’s, 324 F.3d at 175 (Greenberg, J., dissenting) (noting that “LePage’s did not even attempt to show that it could not compete by calculating the discount that it would have had to provide in order to match the discounts offered by 3M through its bundled rebates”). Thus, the LePage’s standard could protect a less efficient competitor at the expense of consumer welfare. As Judge Greenberg explained in his LePage’s dissent, the Third Circuit’s standard “risks curtailing price competition and a method of pricing beneficial to customers because the bundled rebates effectively lowered [the seller’s] costs.” LePage’s, 324 F.3d at 179 (Greenberg, J., dissenting).

The AMC also lamented that LePage’s “offers no clear standards by which firms can assess whether their bundled rebates are likely to pass antitrust muster.” AMC Report, supra, at 94. The Commission noted that efficiencies, and not schemes to acquire or maintain monopoly power, likely explain the use of bundled discounts because many firms without market power offer them. Id. at 95. The AMC thus proposed a three-part test that it believed would protect pro-competitive bundled discounts from antitrust scrutiny. The AMC proposed that:

Courts should adopt a three-part test to determine whether bundled discounts or rebates violate Section 2 of the Sherman Act. To prove a violation of Section 2, a plaintiff should be required to show each one of the following elements (as well as other elements of a Section 2 claim): (1) after allocating all discounts and rebates attributable to the entire bundle of products to the competitive product, the defendant sold the competitive product below its incremental cost for the competitive product; (2) the defendant is likely to recoup these short-term losses; and (3) the bundled discount or rebate program has had or is likely to have an adverse effect on competition.

Id. at 99. The AMC reasoned that the first element would (1) subject bundled discounts to antitrust scrutiny only if they could exclude a hypothetical equally efficient competitor and (2) provide sufficient clarity for businesses to determine whether their bundled discounting practices run afoul of § 2. Id. at 100. The AMC concluded that the three-part test would, as a whole, bring the law on bundled discounting in line with the Supreme Court’s reasoning in Brooke Group. Id.

We must decide whether we should follow LePage’s or whether we should part ways with the Third Circuit by adopting a cost-based standard to apply in bundled discounting cases.

Observers have commented that, in some respects, bundled discounts are similar to both predatory pricing and tying. See Nalebuff, supra, 50 Antitrust Bull, at 365; Daniel L. Rubinfeld, SM’s Bundled Rebates: An Economic Perspective, 72 U. Chi. L.Rev. 243, 252-56 (2005). As the Supreme Court explained in Brooke Group, a plaintiff in a single product predatory pricing case must establish that the defendant priced below cost and that there was a probability the defendant could recoup the losses it suffered during the predation period. See Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578. In a normal tying case, however, while a plaintiff must prove that it was “coerced” into buying the tied products from the defendant, a plaintiff does not need to prove that the defendant priced the products below cost, and therefore the plaintiff also does not need to prove any recoupment of losses. See Datagate, Inc. v. Hewlett-Packard Co., 60 F.3d 1421, 1423-24 (9th Cir.1995).

However, “[o]ne difference between traditional tying by contract and tying via package discounts is that the traditional tying contract typically forces the buyer to accept both products, as well as the cost savings.” 3 Areeda & Hovenkamp, supra, ¶ 749b2 at 332 (Supp.2006). Conversely, “the package discount gives the buyer the choice of accepting the cost savings by purchasing the package, or foregoing the savings by purchasing thé products separately.” Id. The package discount thus does not constrain the buyer’s choice as much as the traditional tie. For that reason, the late-Professor Areeda and Professor Hovenkamp suggest that “[a] variation of the requirement that prices be ‘below cost’ is essential for the plaintiff to establish one particular element of unlawful bundled discounting — namely, that there was actually ‘tying’ — that is, that the purchaser was actually ‘coerced’ (in this case, by lower prices) into taking the tied-up package.” Id. at 331.

In addition, the Supreme Court has forcefully suggested that we should not condemn prices that are above some measure of incremental cost. See id. ¶ 737a at 393 (2d ed.2002) (quoting Brooke Group, 509 U.S. at 223, 113 S.Ct. 2578). In Brooke Group, the Court held that “a plaintiff seeking to establish competitive injury resulting from a rival’s low prices must prove that the prices complained of are below an appropriate measure of its rival’s costs.” Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578. In the course of rejecting the plaintiffs argument that a predatory pricing plaintiff need not prove below-cost pricing, the Court wrote that it has “rejected ... the notion that above-cost prices that are below general market levels or the costs of a firm’s competitors inflict injury to competition cognizable under the antitrust laws.” Id. at 223, 113 S.Ct. 2578 (citing Atl. Richfield Co. v. USA Petroleum Co., 495 U.S. 328, 340, 110 S.Ct. 1884, 109 L.Ed.2d 333 (1990)). The Court went on to emphasize that “[flow prices benefit consumers regardless of how those prices are set, and so long as they are above predatory levels, they do not threaten competition.” Id. (internal quotation omitted). The Court also noted the broad application of the principle that only below-cost prices are anticompetitive, stating that “[w]e have adhered to this principle regardless of the type of antitrust claim involved.” Id. (internal quotation omitted). “As a general rule,” the Court concluded, “the exclusionary effect of prices above a relevant measure of cost either reflects the lower cost structure of the alleged predator, and so represents competition on the merits, or is beyond the practical ability of a judicial tribunal to control without courting intolerable risks of chilling legitimate price-cutting.” Id.; accord Matsushita, 475 U.S. at 594, 106 S.Ct. 1348.

The Court recently reemphasized these principles in Weyerhaeuser Co. v. Ross-Simmons Hardwood Lumber Co., — U.S. -, 127 S.Ct. 1069, 1078, 166 L.Ed.2d 911 (2007), a case in which the Court held that Brooke Group’s below-cost pricing requirement applies in cases in which the plaintiff alleges that the defendant engaged in predatory bidding — the practice of bidding up input costs to drive rivals out of business. Specifically, the Court held that a predatory bidding “plaintiff must prove that the alleged predatory bidding led to below-cost pricing of the predator’s outputs. That is, the predator’s bidding on the [input] side must have caused the cost of the relevant output to rise above the revenues generated in the sale of those outputs.” Weyerhaeuser, 127 S.Ct. at 1078.

Of course, in neither Brooke Group nor Weyerhaeuser did the Court go so far as to hold that in every case in which a plaintiff challenges low prices as exclusionary conduct the plaintiff must prove that those prices were below cost. But the Court’s opinions strongly suggest that, in the normal case, above-cost pricing will not be considered exclusionary conduct for antitrust purposes, and the Court’s reasoning poses a strong caution against condemning bundled discounts that result in prices above a relevant measure of costs.

The Supreme Court’s long and consistent adherence to the principle that the antitrust laws protect the process of competition, and not the pursuits of any particular competitor, reinforce our conclusion of caution concerning bundled discounts that result in prices above an appropriate measure of costs. The Court voiced this principle most notably in Brunswick Corp. v. Pueblo Bowl-O-Mat, 429 U.S. 477, 97 S.Ct. 690, 50 L.Ed.2d 701 (1977). In that case, the plaintiffs challenged, under § 7 of the Clayton Act, 11 Brunswick’s acquisition of unprofitable bowling centers. Id. at 479-80, 97 S.Ct. 690. The plaintiffs, like McKenzie, sought treble damages under § 4 of the Clayton Act. The Court considered the “narrow” issue of “whether antitrust damages are available where the sole injury alleged is that competitors were continued in business, thereby denying [the plaintiffs] an anticipated increase in market shares.” Id. at 484, 97 S.Ct. 690. The Court observed that the damages the plaintiffs sought were “designed to provide them with the profits they would have realized had competition been reduced.” Id. at 488, 97 S.Ct. 690. Noting that “[t]he antitrust laws, however, were enacted for ‘the protection of competition not competitors,’ ” the Court reasoned that “[i]t is inimical to the purposes of these laws to award damages for the type of injury claimed here.” Id. (quoting Brown Shoe Co. v. United States, 370 U.S., 294, 320, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962)). The Court concluded:

We therefore hold that for plaintiffs to recover treble damages on account of § 7 violations, they must prove more than injury causally linked to an illegal presence in the market. Plaintiffs must prove anti-trust injury, which is to say injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful. The injury should reflect the anticompetitive effect either of the violation or of anticompetitive acts made possible by the violation.

Id. at 489, 97 S.Ct. 690.

Subsequent to Brunswick, the Court has often reinforced the principle that the antitrust laws’ prohibitions focus on protecting the competitive process and not on the success or failure of individual competitors. See, e.g., Volvo Trucks N. Am., Inc. v. Reeder-Simco GMC, Inc., 546 U.S. 164, 179-80 126 S.Ct. 860, 872, 163 L.Ed.2d 663 (2006) (“Interbrand competition, our opinions affirm, is the primary concern of antitrust law.” (internal quotation omitted)); Spectrum Sports, 506 U.S. at 458, 113 S.Ct. 884 (“The purpose of the [Sherman] Act is not to protect businesses from the working of the market; it is to protect the public from the failure of the market. The law directs itself not against conduct which is competitive, even severely so, but against conduct which unfairly tends to destroy competition itself.”); Atl. Richfield, 495 U.S. at 331, 110 S.Ct. 1884 (holding that a firm does not incur an antitrust injury when it loses sales to a competitor charging nonpredatory prices pursuant to a vertical, maximum-price-fixing scheme); Cargill, Inc. v. Monfort of Colo., Inc., 479 U.S. 104, 113, 107 S.Ct. 484, 93 L.Ed.2d 427 (1986) (extending antitrust injury requirement to suits for in-junctive relief under § 16 of the Clayton Act, 15 U.S.C. § 26); J. Truett Payne Co. v. Chrysler Motors Corp., 451 U.S. 557, 562, 101 S.Ct. 1923, 68 L.Ed.2d 442 (1981) (extending antitrust injury requirement to price discrimination suits arising under § 2 of the Clayton Act). The Court’s reasoning and conclusions in Brooke Group, as reaffirmed recently in Weyerhaeuser, accordingly show a measured concern to leave unhampered pricing practices that might benefit consumers, absent the clearest showing that an injury to the competitive process will result. Microsoft, 253 F.3d at 58; Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039, 1060-61 (8th Cir.2000).

One of the challenges of interpreting and enforcing the amorphous prohibitions of §§ 1 and 2 of the Sherman Act is ensuring that the antitrust laws do not punish economic behavior that benefits consumers and will not cause long-run injury to the competitive process. A bundled discount, however else it might be viewed, is a price discount on a collection of goods. The Supreme Court has undoubtedly shown a solicitude for price competition. In Wey-erhaeuser, Justice Thomas, writing for the Court, reminded us that, in Brooke Group, the Court had cautioned that “the costs of erroneous findings of predatory-pricing liability were quite high because [t]he mechanism by which a firm engages in predatory pricing — lowering prices — is the same mechanism by which a firm stimulates competition, and therefore, mistaken findings of liability would chill the very conduct the antitrust laws are designed to protect.” Weyerhaeuser, 127 S.Ct. at 1075 (internal quotations omitted, alteration in original).

Given the endemic nature of bundled discounts in many spheres of normal economic activity, we decline to endorse the Third Circuit’s definition of when bundled discounts constitute the exclusionary conduct proscribed by § 2 of the Sherman Act. Instead, we think the course safer for consumers and our competitive economy to hold that bundled discounts may not be considered exclusionary conduct within the meaning of § 2 of the Sherman Act unless the discounts resemble the behavior that the Supreme Court in Brooke Group identified as predatory. 12 Accordingly, we hold that the exclusionary conduct element of a claim arising under § 2 of the Sherman Act cannot be satisfied by reference to bundled discounts unless the discounts result in prices that are below an appropriate measure of the defendant’s costs. 13

The next question we must address is how we define the appropriate measure of the defendant’s costs in bundled discounting cases and how we determine whether discounted prices fall below that mark. Defining the appropriate measure of costs in a bundled discounting case is more complex than in a single product case. In a single product case, we may simply ask whether the defendant has priced its product below its incremental cost of producing that product because a rival that produces the same product as efficiently as the defendant should be able to match any price at or above the defendant’s cost. However, as we discussed above, a defendant offering a bundled discount, without pricing below cost either the individual products in the bundle or the bundle as a whole, can, in some cases, exclude a rival who produces one of the products in the bundle equally or more efficiently than the defendant. Thus, simply asking whether the defendant’s prices are below its incremental costs might fail to alert us to bundled discounts that threaten the exclusion of equally efficient rivals. Nonetheless, we are mindful that, in single product pricing cases, the Supreme Court has not adopted rules condemning prices above a seller’s incremental costs. With these considerations in mind, we assess the rules the parties and amici propose for us to use in bundled discounting cases to determine the appropriate measure of a defendant’s costs and whether a defendant has priced below that level.

PeaceHealth and some amici urge us to adopt a rule they term the “aggregate discount” rule. This rule condemns bundled discounts as anticompetitive only in the narrow cases in which the discounted price of the entire bundle does not exceed the bundling firm’s incremental cost to produce the entire bundle. Peace-Health and amici argue that support for such a rule can be found in the Supreme Court’s single product predation cases— Brooke Group and Weyerhaeuser.

We are not persuaded that those cases require us to adopt an aggregate discount rule in multi-product discounting cases. As we discussed above, bundled discounts present one potential threat to consumer welfare that single product discounts do not: A competitor who produces fewer products than the defendant but produces the competitive product at or below the defendant’s cost to produce that product may nevertheless be excluded from the market because the competitor cannot match the discount the defendant offers over its numerous product lines. This possibility exists even when the defendant’s prices are above cost for each individual product and for the bundle as a whole. See Ortho, 920 F.Supp. at 467; Nalebuff, supra, 50 Antitrust Bull. at 359 (“Whether or not a collection of goods is sold at a profit does not reveal whether one-good rivals were foreclosed.”). Under a discount aggregation rule, anticompetitive bundled discounting schemes that harm competition may too easily escape liability.

Additionally, as commentators have pointed out, Brooke Group’s safe harbor for above-cost discounting in the single product discount context is not based on a theory that above-cost pricing strategies can never be anticompetitive, but rather on a cost-benefit rejection of a more nuanced rule. 3 Areeda & Hovenkamp, supra, ¶ 749b at 324 (Supp.2006); Lambert, supra, 89 Minn. L.Rev. at 1704; see also Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 414, 124 S.Ct. 872, 157 L.Ed.2d 823 (2004) (explaining that while above-cost predatory pricing schemes may exist, they are “ ‘beyond the practical ability of a judicial tribunal to control’ ” (quoting Brooke Group, 509 U.S. at 223, 113 S.Ct. 2578)). That is, the safe harbor rests on the premise that “any consumer benefit created by a rule that permits inquiry into above-cost, single-product discounts, but allows judicial condemnation of those deemed legitimately exclusionary, would likely be outweighed by the consumer harm occasioned by over-deterring nonexclusionary discounts.” Lambert, supra, 89 Minn. L.Rev. at 1705; see Weyerhaeuser, 127 S.Ct. at 1075 (noting the high costs of erroneous findings of predatory-pricing liability because “[t]he mechanism by which a firm engages in predatory pricing — lowering prices — is the same mechanism by which a firm stimulates competition” (alteration in original, internal quotations omitted)); 3 Areeda & Hovenkamp, supra, ¶ 749b at 324 (Supp.2006) (noting that “our measurement tools are too imprecise to evaluate [above-cost discounting] strategies without creating an intolerable risk of chilling competitive behavior”). So, in adopting an appropriate cost-based test for bundled discounting cases, we should not adopt an aggregate discount rule without inquiring whether a rule exists that is more likely to identify anticompetitive bundled discounting practices while at the same time resulting in little harm to competition.

The first potential alternative cost-based standard we consider derives from the district court’s opinion in Ortho. This standard deems a bundled discount exclusionary if the plaintiff can show that it was an equally efficient producer of the competitive product, but the defendant’s bundled discount made it impossible for the plaintiff to continue to produce profitably the competitive product. As the district court in Ortho phrased the standard: a plaintiff basing a § 2 claim on an anti-competitive bundled discount “must allege and prove either that (a) the monopolist has priced below its average variable cost or (b) the plaintiff is at least as efficient a producer of the competitive product as the defendant, but that the defendant’s pricing makes it unprofitable for the plaintiff to continue to produce.” Ortho, 920 F.Supp. at 469. Under this standard, above-cost prices are not per se legal. Cf. Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578. Instead, this standard treats below-cost prices as simply one beacon for identifying discounts that create the risk of excluding firms that are as efficient as the defendant — the unique anticompetitive risk posed by bundled discounts. See Ortho, 920 F.Supp. at 466-67. Under Ortho’s standard, an above-cost discount can still be anticompetitive if a plaintiff proves it is as efficient a producer as the defendant, but is excluded because the defendant sells in more product markets than the plaintiff and can “spread the total discount over all those product lines and ... force competitors to provide the entire dollar amount of the discount on a smaller collection of products.” Lambert, supra, 89 Minn. L.Rev. at 1728. As compared-to the discount aggregation rule, Ortho’s approach does a better job of identifying bundled discounts that threaten harm to competition.

However, one downside of Ortho’s standard . is that it does .not provide adequate gúidanee to sellers who wish to offer pro- competitive, bundled discounts because the standard looks to the costs of the actual plaintiff. A potential defendant who is considering offering a bundled discount will likely not have access to information about its competitors’ costs, thus making it hard for that potential discounter, under the Ortho standard, to determine whether the discount it wishes to offer complies with the antitrust laws. Also, the Ortho standard, which asks whether the actual plaintiff is as efficient a producer as the defendant, could require multiple suits to determine the legality of a single bundled discount. While it might turn out that the plaintiff in one particular case is not as efficient a producer of the competitive product as the defendant, another rival might be. This second rival would have to bring another suit under the Ortho approach. We decline to adopt a rule that might encourage more antitrust litigation than is reasonably necessary to ferret out anticompetitive practices. See Bell Atl. Corp. v. Twombly, — U.S. -, 127 S.Ct. 1955, 1966-67, 167 L.Ed.2d 929 (2007); see also Manual for Complex Litigation (Fourth) § 30 (2004) (“Antitrust litigation can ... involve voluminous documentary and testimonial evidence, extensive discovery, complicated legal, factual, and technical (particularly economic) questions, numerous parties and attorneys, and substantial sums of money.... Antitrust trials usually are long, and there often are controversies over settlements and attorney fees.”). Accordingly, we do not adopt Ortho’s approach, which we believe would be unduly cumbersome for sellers to assess and thus might chill procompetitive bundled discounting.

Instead, as our cost-based rule, we adopt what amici refer to as a “discount attribution” standard. 14 Under this standard, the full amount of the discounts given by the defendant on the bundle are allocated to the competitive product or products. If the resulting price of the competitive product or products is below the defendant’s incremental cost to produce them, the trier of fact may find that the bundled discount is exclusionary for the purpose of § 2. This standard makes the defendant’s bundled discounts legal unless the discounts have the potential to exclude a hypothetical equally efficient producer of the competitive product. 15 Cf. Ortho, 920 F.Supp. at 469 (deeming bundled discounts anticompetitive if the actual plaintiff is excluded but equally efficient).

In their leading treatise on antitrust law, Professors Areeda and Hovenkamp support an approach that focuses on whether a bundled discount excludes a hypothetical equally efficient rival. Rejecting Ortho’s “actual plaintiff’ standard, they explain:

[W]e would not require a showing that the actual plaintiff be equally efficient. The relevant question is not necessarily whether a particular plaintiff was equally efficient, but whether the challenged bundling practices would have excluded an equally efficient rival, without reasonable justification. This rule is preferable on grounds of both administrability and principle. On the first, proving whether a hypothetical equally efficient rival is excluded by a multiproduct discount is typically quite manageable. By contrast, proof that the plaintiff is equally efficient can be quite difficult, particularly in cases where the defendant produces a larger product line than the plaintiff and there are joint costs.

A requirement that the bundling practice be sufficiently severe so as to exclude an equally efficient single-product rival, and without an adequate business justification, seems to strike about the right balance between permitting aggressive pricing while prohibiting conduct that can only be characterized as anticompetitive. Requiring the defendant’s pricing policies to protect the trade of higher cost rivals is overly solicitous of small firms and denies customers the benefits of the defendant’s lower costs. Further, if the practice will exclude an equally efficient rival, then it will exclude whether or not the rival is equally efficient in fact.

3 Areeda & Hovenkamp, supra, ¶ 749a at 322-23 (Supp.2006) (footnotes omitted); accord Nalebuff, supra, 50 Anti-trust Bull. at 328-29. Judge Posner’s work on antitrust law also supports an approach that asks whether a bundled discount excludes a hypothetical equally efficient rival, stating that the acts of a monopolist should be deemed exclusionary if “the challenged practice is likely in the circumstances to exclude from the defendant’s market an equally or more efficient competitor.” Posner, supra, at 194-95.

Areeda and Hovenkamp also support using a discount attribution approach to determine if a bundled discount is exclusionary. They state:

To see whether a package price is “exclusionary” ... one simply attributes the entire discount on all products in the package to the product for which exclusion is claimed. If the resulting price is less than the defendant’s cost, then the package discount is exclusionary as against a rival who makes only one of the two goods in the package.

3 Areeda & Hovenkamp, supra, ¶ 749b2 at 335-36 (Supp.2006) (footnotes omitted); accord Nalebuff, supra, 50 Anti-trust Bull. at 328. The discount attribution standard has also been used by two of the district courts in the small number of published opinions dealing with allegedly exclusionary bundled discounts. See Info. Res., Inc. v. Dun & Bradstreet Corp., 359 F.Supp.2d 307, 307 (S.D.N.Y.2004) (“When price discounts in one market are bundled with the price charged in a second market, the discounts must be applied to the price in the second market in determining whether that price is below that product’s average variable cost.”); Virgin Atl. Airways Ltd. v. British Airways PLC, 69 F.Supp.2d 571, 580 n. 8 (S.D.N.Y.1999) (holding that a bundled discount is exclusionary “if the competitive product in the bundle[was] sold for a price below average variable cost after the discounts on the monopoly items in the bundle were subtracted from the price of that competitive product”), aff'd, 257 F.3d 256 (2d Cir.2001). The discount attribution standard is also the standard endorsed by the AMC. AMC Report, supra, at 99 (requiring plaintiff to prove that “after allocating all discounts and rebates attributable to the entire bundle of products to the competitive product, the defendant sold the competitive product below its incremental cost for the competitive product”).

The discount attribution standard provides clear guidance for sellers that engage in bundled discounting practices. A seller can easily ascertain its own prices and costs of production and calculate whether its discounting practices run afoul of the rule we have outlined. See Nalebuff, supra, 50 Antitrust Bull, at 330. Unlike under the Ortho standard, under the discount attribution standard a bundled discounter need not fret over and predict or determine its rivals’ cost structure. 16

We are aware that liability under the discount attribution standard has the potential to sweep more broadly than under the aggregate discount rule or the Ortho standard. However, there is limited judicial experience with bundled discounts, and academic inquiry into the competitive effects of bundled discounts is only beginning. 17 By comparison, the Supreme Court’s decision in Brooke Group (prefaced by the Court’s discussion of predatory pricing in Matsushita, 475 U.S. at 588-91, 106 S.Ct. 1348) marked the culmination of nearly twenty years of scholarly and judicial analysis of the feasibility and competitive effects of single product predatory pricing schemes. 18 Cf. 3 Areeda & Hovenkamp, supra, ¶ 749b at 323 (Supp.2006) (“[T]he theory of anticompetitive discounting is in much the same position as the theory of predatory pricing was in the 1970s: no shortage of theories, but a frightening inability of courts to assess them.”). The cost-based standard we adopt will allow courts the experience they need to divine the prevalence and competitive effects of bundled discounts and will allow these difficult issues to further percolate in the lower courts. As the Solicitor General noted in his amicus brief urging the denial of certiorari in LePage’s:

There is insufficient experience with bundled discounts to this point to make a firm judgment about the relative prevalence of exclusionary versus procom-petitive bundled discounts. Relative to the practice of predatory pricing analyzed in Brooke Group, there is less knowledge on which to assess whether, or to what extent, the legal approach to a monopolist’s allegedly exclusionary bundled discounts should be driven by a strong concern for false positives and low risk of false negatives. Further empirical development may shed light on that question.

Brief for United States as Amicus Curiae at 14, 3M Co. v. LePage’s Inc., 542 U.S. 953, 124 S.Ct. 2932, 159 L.Ed.2d 835 (2004) (No. 02-1865), 2004 WL 1205191 (citation omitted). Pending further judicial and academic inquiry into the prevalence of anti-competitive bundled discounts, we think it preferable to allow plaintiffs to challenge bundled discounts if those plaintiffs can prove a defendant’s bundled discounts would have excluded an equally efficient competitor.

To summarize, the primary anticompeti-tive danger posed by a multi-product bundled discount is that such a discount can exclude a rival is who is equally efficient at producing the competitive product simply because the rival does not sell as many products as the bundled discounter. Thus, a plaintiff who challenges a package discount as anticompetitive must prove that, when the full amount of the discounts given by the defendant is allocated to the competitive product or products, the resulting price of the competitive product or products is below the defendant’s incremental cost to produce them. This requirement ensures that the only bundled discounts condemned as exclusionary are those that would exclude an equally efficient producer of the competitive product or products.

The next issue before us is the appropriate measure of incremental costs in a bundled discounting case. In single product predatory pricing cases, the appropriate measure of incremental costs is an open question in this circuit. See Rebel Oil Co. v. Atl. Richfield Co., 146 F.3d 1088, 1092 (9th Cir.1998). The Supreme Court has likewise refused to decide the matter. See Brooke Group, 509 U.S. at 222 n. 1, 113 S.Ct. 2578; Cargill, 479 U.S. at 118 n. 12, 107 S.Ct. 484.

As our cases and the relevant academic literature thoroughly discuss, firms face both fixed costs' — costs that a firm must bear regardless of the amount of output— and variable costs — costs that change with the amount of output. The sum of fixed and variable costs is a firm’s total cost. Marginal cost is the increase to total cost that occurs as a result of producing one additional unit of output. Average cost is the sum of fixed costs and total variable costs, divided by the amount of output. In their oft-cited 1975 law review article, Professors Areeda and Turner concluded that the optimal measure of a firm’s cost in a predatory pricing case is marginal cost — the cost to produce one additional unit and the price that would obtain in the market under conditions of perfect competition. See Phillip Areeda & Donald F. Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L.Rev. 697, 712, 716 (1975). However, Professors Areeda and Turner also recognized, that “[t]he incremental cost of making and selling the last unit cannot' readily be inferred from conventional business • accounts, which typically go no further than showing observed average variable cost.” Id. at 716. Thus, the professors adopted average variable cost as a surrogate for marginal cost. Id. A number of circuits have adopted the Aree-da-Turner formulation and concluded that prices below average variable cost can indicate predation. See, e.g., Stearns Airport Equip. Co. v. FMC Corp., 170 F.3d 518, 532 (5th Cir.1999); Morgan v. Ponder, 892 F.2d 1355, 1360 (8th Cir.1989); Barry Wright, 724 F.2d at 236; Ne. Tel. Co. v. AT & T Co., 651 F.2d 76, 87-88 (2d Cir.1981). 19

Likewise, “we have approved the use of marginal or average variable cost statistics in proving predation.” See William Inglis & Sons Baking Co. v. ITT Cont’l Baking Co., 668 F.2d 1014, 1033 (9th Cir.1981). 20 We have also held that a plaintiff can establish a prima facie case of predatory pricing by proving that the defendant’s prices were below average variable cost. Id. at 1036. We see no reason to depart from these principles in the bundled discounting context, and we hold that the appropriate measure of costs for our cost-based standard is average variable cost.

In summary, we hold the following: To prove that a bundled discount was exclusionary or predatory for the purposes of a monopolization or attempted monopolization claim under § 2 of the Sherman Act, the plaintiff must establish that, after allocating the discount given by the defendant on the entire bundle of products.to the competitive product or products, the defendant sold the competitive product or products below its average variable cost of producing them. The district court’s jury instruction on the attempted monopolization claim, which built on the holding of LePage’s that we have rejected, thus contained an error of law. 21

McKenzie argues that we may nevertheless affirm the jury’s verdict on the principle that flawed jury instructions are a harmless error when the facts which needed to be proven are strongly supported by the evidence presented at trial. See Harmsen v. Smith, 693 F.2d 932, 945 (9th Cir.1982); Cancellier v. Federated Dep't Stores, 672 F.2d 1312, 1316 (9th Cir.1982). In support of this argument, McKenzie points out that there was “un-' disputed evidence of [PeaceHealth’s] higher prices and the need for [McKenzie] to sell beneath variable cost to hold Regence harmless from [PeaceHealth’s] threatened price increases.” However, as we have held, the relevant inquiry is not whether PeaceHealth’s pricing practices forced McKenzie to price below cost, but whether PeaceHealth priced its own services below an appropriate measure of its cost, as we have defined that concept using the discount attribution rule. In this case, we cannot conclude that .the error in,the jury instructions was harmless. We vacate the judgment entered in McKenzie’s favor and remand for further proceedings consistent with our opinion. 22

B

After trial, the jury also returned a verdict in favor of McKenzie on its claim of primary-line price discrimination 23 under Oregon state law. McKenzie’s theory was that PeaceHealth discriminated in price as between Regence and Providence. Specifically, PeaceHealth, who was the exclusive preferred provider in Regence’s PPP, charged Regence an 85% reimbursement rate while PeaceHealth charged Providence, an insurer with whom PeaceHealth had no exclusive arrangement, a 93% reimbursement rate. McKenzie alleged that PeaceHealth’s price discrimination injured McKenzie because its pricing scheme was the cause of McKenzie’s inability to obtain preferred status with Regence.

The Oregon price discrimination statute, 24 Or.Rev.Stat. § 646.040, is nearly identical to the federal price discrimination provisions in § 2 of the Robinson-Patman Act, 25 15 U.S.C. § 13. Because section 646.040 is essentially a verbatim reproduction of § 2 of the Robinson-Patman Act, Oregon courts Took to federal price discrimination law for guidance in interpreting section 646.040. Yamaha Store of Bend, Or., Inc. v. Yamaha Motor Corp., 310 Or. 333, 798 P.2d 656, 659 n. 6 (1990), modified on reconsideration on other grounds, 311 Or. 88, 806 P.2d 123 (1991); Redmond Ready-Mix, Inc. v. Coats, 283 Or. 101, 582 P.2d 1340, 1346 (1978). The seminal federal case dealing with primary-line price discrimination under the Robinson-Patman Act is Brooke Group. In Brooke Group, as we. discussed above, the United States Supreme Court held that there are two prerequisites to a claim of primary-line price discrimination under the Robinson-Patman Act. First, “a plaintiff seeking to establish competitive injury resulting from a rival’s low prices must prove that the prices complained of are below an appropriate measure of its rival’s costs.” Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578. Second, the plaintiff must prove “that the competitor had a reasonable prospect ... of recouping its investment in below-cost prices.” Id. at 224, 113 S.Ct. 2578.

The district court in this case, however, did not instruct the jury to consider whether PeaceHealth priced its services below cost in deciding whether McKenzie established a claim of primary-line price discrimination under Oregon law. The district court instead instructed the jury as follows:

[I]n order for the plaintiff to establish a violation of the price discrimination statute, it has the burden of proving each and every one of the following elements by a preponderance of the evidence: (1) That there were contemporaneous sales by a defendant to other insurers in the relevant market; (2) that defendant has discriminated in price between insurers in the contemporaneous sale of hospital services; and (3) that the effect of defendant’s price discrimination was to substantially lessen competition or create a monopoly in the sale of hospital services in the relevant market, or to injure, destroy or prevent competition’ between plaintiff and defendant.

This instruction is inconsistent with Brooke Group, which requires a plaintiff to prove (1) below-cost pricing and (2) likelihood of recoupment regardless of “whether the claim alleges predatory pricing under § 2 of the Sherman Act or primary-line price discrimination under the Robinson-Patman Act.” Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578.

Of course, we are not deciding a Robinson-Patman Act case, but a case arising under Oregon’s price discrimination statute. Because our decision rests upon a question of state law, we must divine how the Oregon Supreme Court would decide the issue before us. Burlington Ins. Co. v. Oceanic Design & Constr., Inc., 383 F.3d 940, 945 (9th Cir.2004). McKenzie asserts that the Oregon Supreme Court would not require a plaintiff to prove below-cost pricing as a prerequisite to a claim of primary-line price discrimination. McKenzie relies primarily on the Oregon Supreme Court’s 1978 decision in Redmond Ready-Mix for this argument.

In Redmond Ready-Mix, the plaintiff brought a primary-line price discrimination suit under section 646.040 against the husband and wife who were the plaintiffs competitor in the market to sell retail premixed concrete. 582 P.2d at 1342. The plaintiff asserted that the defendants sold concrete for a lower price in the geographic areas in which the plaintiff competed with the defendants than in the geographic areas in which the plaintiff did not compete. See id. at 1342-45. The trial court found that the defendants did not violate section 646.040. See id. at 1342. The trial court reasoned that although the defendants sold concrete at lower prices in the competitive geographic area than in other areas, the plaintiff did not establish that the lower prices lessened competition as required by the section 646.040 because the “defendants were not selling below their average cost.” Id. The Oregon Supreme Court affirmed. Id. at 1352.

In accordance with the prevailing federal case law at the time, the Oregon Supreme Court noted that a claim of primary-line price discrimination could be sustained upon a showing that the defendant price discriminated with a “predatory intent.” See id. at 1348. The court pointed out that predatory intent could be implied from below-cost selling. Id. at 1350. The court held, however, that there was no evidence the defendants priced below cost. See id.

The court then noted that, alternatively, the plaintiffs could establish a claim of primary-line price discrimination by show: ing that the defendants’ price discrimination resulted in a substantial impairment of competition. Id. The court noted a number of factors it looked to in order to determine whether the defendants’ price discrimination impaired competition. See id. These factors included: (a) monopoly or overpowering position of the seller in the market; (b) aggressive objectives towards the seller’s smaller rivals; (c) deep, sustained undercutting of rivals’ prices; (d) persistent sales below the seller’s cost; and (e) actual or impending demise of a seller’s sole rival in a market. Id. The court held that the evidence did not demonstrate any impairment of competition and affirmed the judgment of the trial court. See id. at 1352.

McKenzie argues that under Oregon law, unlike under Brooke Group, a plaintiff can establish a claim of primary-line price discrimination by simply showing predatory intent or a substantial impact on competition as defined by the five factors outlined in Redmond Ready-Mix. However, the Oregon Supreme Court decided Redmond Ready-Mix in 1978 — fifteen years before the United States Supreme Court’s decision in Brooke Group and at a time when “[m]uch of the ‘federal law’ on [primary-line price discrimination was] in a recognized state of confusion.” Id. at 1346. Because of Brooke Group federal law is clearer today (at least on one issue) — a plaintiff in a primary-line price discrimination cáse must prove that its rival priced below cost. Since Brooke Group, no Oregon court has published an opinion interpreting section 646.040. We believe the Oregon Supreme Court, if given the opportunity, would follow Brooke Group.

First, the Oregon Supreme Court has held that federal price discrimination cases are important persuasive authority in interpreting Oregon’s price discrimination law. See Yamaha, 798 P.2d at 659 n. 6. Second, even though the Oregon Supreme Court in Redmond Ready-Mix did not have the benefit of Brooke Group, the court emphasized that below-cost pricing was an important factor in its decision that the defendants had not engaged in primary-line price discrimination. See Redmond, 582 P.2d at 1349 (emphasizing that predation can be implied from “below-cost selling” (internal quotation omitted)). Provided the benefit of nearly thirty years of evolution of predatory pricing theory and federal predatory pricing law, we believe the Oregon Supreme Court, if presented with the opportunity, would follow Brooke Group and require a plaintiff alleging primary-line price discrimination to prove that its rival priced below cost.

As we noted above, the district court’s jury instructions in this case did not require the jury to find that, as a pre-requi-site to McKenzie’s claim of primary-line price discrimination, PeaceHealth priced below cost. Because the district court’s jury instructions were erroneous, we vacate the jury’s verdict in favor of McKenzie on McKenzie’s claim of primary-line price discrimination.

C

Finally, the jury found in favor of McKenzie on its Oregon tort law claim of intentional interference with prospective economic advantage. The parties agree that a claim of tortious interference under Oregon law requires a complementary finding of a violation of the antitrust laws. See Kovac v. Crooked River Ranch Club & Maint. Ass’n, 186 Or.App. 545, 63 P.3d 1197, 1201 (2003); Willamette Dental Group, P.C. v. Or. Dental Serv. Corp., 130 Or.App. 487, 882 P.2d 637, 644 (1994). Because we have vacated the jury’s verdict in favor of McKenzie on McKenzie’s antitrust claims, we also vacate the jury’s verdict in favor of McKenzie on McKenzie’s claim of intentional interference with prospective economic advantage.

Ill

We next address McKenzie’s cross-appeal.

Before trial, the district court granted PeaceHealth summary judgment on McKenzie’s claim that PeaceHealth illegally tied primary and secondary services to its provision of tertiary services in violation of § 1 of the Sherman Act, 15 U.S.C. § 1. The district court granted summary judgment because McKenzie presented no evidence that the insurers were coerced into taking the tied product.

We review de novo the district court’s grant of summary judgment. Welles v. Turner Entm’t Co., 488 F.3d 1178, 1183 (9th Cir.2007). Federal Rule of Civil Procedure 56(c) entitles a party to summary judgment “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” In deciding a motion for summary judgment, we view the evidence in the light most favorable to the non-moving party, and draw all justifiable inferences in favor of the non-moving party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986); Betz v. Trainer Wortham & Co., 486 F.3d 590, 591 (9th Cir.2007).

A tying arrangement is a device used by a seller with market power in one product market to extend its market power to a distinct product market. Paladin Assocs., Inc. v. Mont. Power Co., 328 F.3d 1145, 1159 (9th Cir.2003). To accomplish this objective, the seller conditions the sale of one product (the tying product) on the buyer’s purchase of a second product (the tied product). 26 See Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451, 461, 112 S.Ct. 2072, 119 L.Ed.2d 265 (1992); Posner, supra, at 197. Tying arrangements are forbidden on the theory that, if the seller has market power over the tying product, the seller can leverage this market power through tying arrangements to exclude other sellers of the tied product. 27 See Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 14, 104 S.Ct. 1551, 80 L.Ed.2d 2 (1984); Fortner Enters., Inc. v. U.S. Steel Corp., 394 U.S. 495, 517-18, 89 S.Ct. 1252, 22 L.Ed.2d 495 (1969) [hereinafter Fortner I] (White, J., dissenting).

The Supreme Court has developed a unique per se rule for illegal tying arrangements. For a tying claim to suffer per se condemnation, a plaintiff must prove: (1) that the defendant tied together the sale of two distinct products or services; (2) that the defendant possesses enough economic power in the tying product market to coerce its customers into purchasing the tied product; and (3) that the tying arrangement affects a “not insubstantial volume of commerce” in the tied product market. See Paladin Assocs., Inc., 328 F.3d at 1159 (citing Eastman Kodak, 504 U.S. at 461-62, 112 S.Ct. 2072).

As to the first element, McKenzie alleged two distinct products: tertiary services (the tying or desired product) and primary and secondary services (the tied or forced product). As to the third element, PeaceHealth does not dispute that the tying arrangement affected a substantial volume of commerce in the market for primary and secondary services. See Fortner I, 394 U.S. at 501, 89 S.Ct. 1252. Thus, the only issue we must decide is whether PeaceHealth coerced purchases of primary and secondary services.

McKenzie first argues that, -in this particular case, it need not demonstrate coercion because it was a third party to the tying arrangements between PeaceHealth and the insurers, or, at the very least, that the standard for coercion is lower in cases brought by a third-party plaintiff. For the premise that the standard of coercion is lower or nonexistent for plaintiffs who are not parties to the tying arrangement, McKenzie relies heavily on the Fifth Circuit’s opinion in Heatransfer Corp. v. Volkswagenwerk, A.G., 553 F.2d 964 (5th Cir.1977). In that case, the Fifth Circuit suggested that, in cases brought by third parties, “[t]he fact of coercion appears less important ... [than] the fact of foreclosure.” Id. at 978. But the Fifth Circuit did not abandon the coercion requirement in third-party suits. Instead, the court concluded that if the purchaser under the tying arrangement is “coerced or ‘persuaded’ to buy goods which they otherwise would not buy, with the result being tremendous lessening of the market in which a competitor sells his product, such a showing is sufficient to submit the question of a Section 1 antitrust violation to the jury.” Id.

Additionally, the suggestion that a lower (or no) coercion standard must be satisfied in third party suits, and in particular the dictum in Heatransfer, has been criticized by commentators:

A few dicta have suggested that standards for proving a tying condition — • often expressed as “coercion” — should be lower for defendant’s rival than for its customers. This distinction was rightly rejected by the Eleventh Circuit, which correctly pointed out that every plaintiff must prove the tying condition, although of course the competitor need not show that it was itself was subjected to any such condition. Moreover, Supreme Court discussions about the existence of a tie have not varied according to the status of the plaintiff.

10 Areeda & Hovenkamp, supra, ¶ 1752d at 264 (2d ed.2004) (footnotes omitted) (citing Tic-X-Press v. Omni Promotions Co., 815 F.2d 1407, 1415 n. 15 (11th Cir.1987); Heatransfer, 553 F.2d at 978). Indeed, the Supreme Court has emphasized that the coerced purchase of the tied product is the key aspect of an illegal tie:

[T]he essential characteristic of an invalid tying arrangement lies in the seller’s exploitation of its control over the tying product to force the buyer into the pur chase of a tied product that the buyer either did not want at all, or might have preferred to purchase elsewhere on different terms.

Jefferson Parish, 466 U.S. at 12, 104 S.Ct. 1551 (emphasis added). Thus, because coercion is often the touchstone issue in assessing a claim of illegal tying, we reject McKenzie’s argument that, because it was not a party to the tying arrangement, it does not need to demonstrate coercion as part of its tying claim.

McKenzie next argues that, even if coercion must be shown in tying cases brought by third parties, there was at least a disputed factual issue regarding coercion in this case. As evidence that no coercion was present in this case, the district court, in granting summary judgment to Peace-Health, relied heavily on the deposition testimony of Farzenah Whyte, Regence’s contract negotiator, who testified that Re-gence voluntarily entered into its contracts with PeaceHealth. PeaceHealth also points out that some insurers contracted to purchase PeaceHealth’s services without exclusivity, indicating that PeaceHealth did not force those who wanted tertiary services to purchase primary and secondary services from PeaceHealth also. 28 Cf. Moore v. Jas. H. Matthews & Co., 550 F.2d 1207, 1217 (9th Cir.1977) (noting that “coercion may be implied from a showing that an appreciable number of buyers, have accepted burdensome terms, such as a tie-in”). However, when all justifiable factual inferences are drawn in McKenzie’s favor, there is no doubt that PeaceHealth’s practice of giving a larger discount to insurers who dealt with it as an exclusive preferred provider may have coerced some insurers to purchase primary and secondary services from PeaceHealth rather than from McKenzie. We conclude that, as a whole, the evidence shows genuine factual disputes about whether PeaceHealth forced insurers either as an implied condition of dealing or as a matter of economic imperative through its bundled discounting, ■ to take its primary and secondary services if the insurers wanted tertiary services.

First, while Whyte testified that Re-gence was not explicitly forced to deal exclusively with PeaceHealth, Whyte also testified that the higher prices Peace-Health would have charged Regence had McKenzie been admitted to Regence’s PPP would have had a “large impact” on Regence. Also, Whyte stated that she had been “held hostage” by ■ PeaceHealth’s pricing practices.

Standing alone, the fact that a customer would end up paying higher prices to purchase the tied products separately does not necessarily create a fact issue on coercion. Paladin, 328 F.3d at 1162; Robert’s Waikiki U-Drive, Inc. v. Budget Rent-a-Car Sys., Inc., 732 F.2d 1403, 1407 (9th Cir.1984). However, the record contains additional evidence of economic coercion. For example, while Pea-ceHealth emphasizes that four insurers in Lane County purchased PeaceHealth’s services separately, “a trivial proportion of separate sales shows that the package discount is as effective as an outright refusal to sell [the tying product] separately.” 10 Areeda & Hovenkamp, supra, ¶ 1758b at 327 (2d ed.2004). In this case, there are twenty-eight insurers operating in Lane County. The fact that only four of them, or about 14% percent, made a separate purchase may indicate some degree of coercion, placing this issue in the realm of disputed facts that must be tendered to a jury. See id. at 328 (suggesting that a less than 10% proportion of separate sales indicates an illegal tie). Additionally, McKenzie provided some evidence that its prices on primary and secondary services were lower than PeaceHealth’s prices on those services. Again, while not dispositive evidence of an illegal tie, it is a permissible inference that a rational customer would not purchase PeaceHealth’s allegedly overpriced product in the absence of a tie. See Data Gen. Corp. v. Grumman Sys. Support Corp., 36 F.3d 1147, 1181 (1st Cir.1994); 10 Areeda & Hovenkamp, supra, ¶ 1756b3 at 301 (2d ed.2004); cf. Amerinet, Inc. v. Xerox Corp., 972 F.2d 1483, 1501 (8th Cir.1992) (refusing to find an illegal tie in part because the plaintiff did not demonstrate that defendant used its market power in the tying product market “to shelter an inferior or overpriced product from competition”). McKenzie also offered expert testimony that Regence’s exclusive relationship with PeaceHealth made no economic sense, evidencing coercion.

Finally, the Supreme Court has condemned tying arrangements when the seller has the market power to force a purchaser to do something that he would not do in a competitive market. Jefferson Parish, 466 U.S. at 17, 104 S.Ct. 1551. PeaceHealth was the only provider of tertiary services in the relevant geographic market. The substantial market power PeaceHealth possessed as a result of being the exclusive provider of tertiary services in Lane County creates a possibility that PeaceHealth was able to force unwanted purchases of primary and secondary services. In light of the evidence adduced by McKenzie at summary judgment, whether PeaceHealth in fact used its market power to effectively coerce purchases of primary and secondary services is a question that can be answered only through further factual development. The need for further factual development renders summary judgment on McKenzie’s tying claim inappropriate. 29 Because a trier of fact might reasonably determine McKenzie established a claim of illegal tying based on the evidence in the record, we vacate the district court’s order granting summary judgment to PeaceHealth and .remand for further proceedings. 30

McKenzie raises two other issues on its cross-appeal related to rulings made by the district court during the course of the trial.

First, McKenzie contends that the district court erred by not admitting into evidence what McKenzie considered to be anti-competitive' conduct of PeaceHealth in petitioning the Oregon attorney general to stop the McKenzie-Triad merger or condition approval of the merger on McKenzie taking certain actions. McKenzie maintains that the district court erred in holding that PeaceHealth’s activity was protected from anti-trust scrutiny under the Noerr-Pennington doctrine, which protects an antitrust defendant’s right to petition the government. See United Mine Workers v. Pennington, 381 U.S. 657, 670, 85 S.Ct. 1585, 14 L.Ed.2d 626 (1965); E.R.R. Presidents Conference v. Noerr Motor Freight, Inc., 365 U.S. 127, 143-44, 81 S.Ct. 523, 5 L.Ed.2d 464 (1961). Because we vacate the jury’s verdict in Part II of our opinion, McKenzie’s argument, which challenges evidentiary rulings the district court made at trial, is moot, and we decline to address it.

The second issue McKenzie raises on its cross-appeal is whether the jury instruction on “combination or conspiracy” to monopolize was correct. As we discussed ábove, the jury found for PeaceHealth on McKenzie’s conspiracy to monopolize claim. The district court refused to give the following instruction proposed by McKenzie: “The involuntary nature of one’s participation in a conspiracy to monopolize is no defense. An antitrust conspirator can be liable for damages even though he participates only under duress.” McKenzie culled this language from our opinion in Calnetics Corp. v. Volkswagen of America, Inc., 532 F.2d 674, 682 (9th Cir.1976) (per curiam), in which we wrote, “[t]he involuntary nature of one’s partic ipation in a conspiracy to monopolize is no defense. An antitrust conspirator can be liable for damages even though he participates only under coercion.” McKenzie argues that, because this instruction was not given, the jury may have found in Peace-Health’s favor because it viewed Regence’s participation in PeaceHealth’s alleged conspiracy to monopolize as involuntary.

As we noted above, the general rule is that we “ ‘review [ ] jury instructions to determine whether, taken as a whole, they mislead the jury or state the law -incorrectly to. the prejudice of the objecting party. . So long as they do not, we review the formulation of the instructions and the choice of language for abuse of discretion.’ ” City of Long Beach v. Standard Oil Co. of Cal., 46 F.3d 929, 933 (9th Cir.1995) (quoting Reed v. Hoy, 909 F.2d 324, 326 (9th Cir.1989)).

While McKenzie’s proposed instruction is a correct statement of the law of conspiracy as we explained it in Calnet-ics Corp., it was within the district court’s discretion to refuse to give the requested instruction because the instruction could have confused the jury. It is true that an antitrust conspirator “can be liable .for damages” even though he participates in the conspiracy only under coercion. See Flintkote Co. v. Lysfjord, 246 F.2d 368, 375 (9th Cir.1957). But the only conspirator who could have been “liable for damages” in this case was PeaeeHealth, the sole defendant. Conversely, if anyone participated in the conspiracy under coercion, it was Regence. “A district court has substantial latitude in tailoring jury instructions .... ” Kendall-Jackson Winery, Ltd. v. E. & J. Gallo Winery, 150 F.3d 1042, 1051 (9th Cir.1998). We conclude that the district court was within its discretion in refusing to give McKenzie’s proposed conspiracy instruction.

IV

The final issue before us is the appeal and cross-appeal of the district court’s award of attorneys’ fees and costs to McKenzie. Because we have vacated the district court’s judgment in favor of McKenzie on the merits of McKenzie’s attempted monopolization, price discrimination, and tortious interference claims, McKenzie is no longer a prevailing party for. the purposes of Federal Rule of Civil Procedure 54(d)(1) and § 4(a) of the Clayton Act, 15 U.S.C. § 15(a). McKenzie is thus not entitled to attorneys’ fees, costs, and expenses, and we vacate the district court’s order awarding fees, costs, and expenses to McKenzie. If McKenzie prevails on remand, it may renew its request for attorneys’ fees and costs. We dismiss McKenzie’s cross-appeal on attorneys’ fees and costs as moot.

V

To .summarize: In No. 05N55640,' we VACATE the judgment in favor of McKenzie and REMAND for further proceedings. In No. 05-35627, we VACATE the summary judgment in favor of Peace-Health and REMAND for further proceedings. 31 In No. 05-36153, we VACATE the district court’s order awarding attorneys’ fees and costs to McKenzie. In No. 05-36202, we DISMISS the appeal as moot. Each party shall bear its own costs on appeal.

1 . As a result of the merger, McKenzie's name changed to Cascade Health Solutions. For the purposes of this opinion, we, like the parties, continue to refer to Cascade Health Solutions as McKenzie.

2 . In a preferred provider plan, health care providers contract with an insurer to provide health care to the insurer's customers. The insurer’s customers pay much higher prices if they obtain services from providers other than those with whom their insurer has contracted.

3 . The focus in attempted monopolization cases on a defendant's "specific intent” to monopolize and on the "dangerous probability” that monopoly will result traces its roots to the Supreme Court’s earliest pronouncements on the Sherman Act. Over one hundred years ago, Justice Holmes explained that the Sherman Act permits claims

against combinations in restraint of commerce among the states and against attempts to monopolize the same. Intent is almost essential to such a combination, and is essential to such an attempt. Where acts are not sufficient in themselves to produce a result which the law seeks to prevent — for instance, the monopoly, — but require further acts in addition to the mere forces of nature to bring that result to pass, an intent to bring it to pass is necessary in order to produce a dangerous probability that it will happen. But when that intent and the consequent dangerous probability exist, this statute, like many others, and like the common law in some cases, directs itself against that dangerous probability as well as against the completed result.

Swift & Co. v. United States, 196 U.S. 375, 396, 25 S.Ct. 276, 49 L.Ed. 518 (1905) (citation omitted). By contrast, in monopolization cases, monopolistic intent 'can be inferred from the exclusionary conduct of a firm with monopoly power. United States v. Aluminum Co. of Am., 148 F.2d 416, 432 (2d Cir.1945) (Hand, J.) (noting that, in a monopolization case, "no intent is relevant except that which is relevant to any liability, criminal or civil: i.e. an intent to bring about the forbidden act”); United States v. United Shoe Mach. Corp., 110 F.Supp. 295, 346 (D.Mass.1953) ("Defendant intended to engage in the leasing practices and pricing policies which maintained its market power. That is all the intent which the law requires when both the complaint and the judgment rest on a charge of ‘monopolizing’, not merely ‘attempting to monopolize’. Defendant having willed the means, has willed the end."), aff'd, 347 U.S. 521, 74 S.Ct. 699, 98 L.Ed. 910 (1954). In attempted monopolization cases, though, the defendant firm "has not yet achieved a position of power in the market but is trying to build up such a position. Being without power to exploit or exclude, such a firm must be shown to have a specific intent to achieve these results.” A.D. Neale & D.G. Goyder, The Antitrust Laws of the United States of America 93 (3d ed.1980).

4 . For example, in the telecommunications field, it is common for companies to offer not only phone service, but also Internet access and television service, and many of these companies offer bundled discounts to customers who purchase their entire package. See Ken Belson, Dial M for Merger, N.Y. Times, Jan. 28, 2005, at Cl; Ken Belson, Cable’s Rivals Lure Customers with Packages, N.Y. Times, Nov. 22, 2004, at Cl.

5 . That bundled discounts are a common feature of our current economic system is relevant to our analysis of allegedly anticompeti-tive conduct under § 2 of the Sherman Act. The Supreme Court, in assessing the stare decisis effect of its prior precedents under §' 1 of the Sherman Act, recently noted that "[fjrom the beginning the Court has treated the Sherman Act as a common-law statute,” and that "[j]ust as the common law adapts to modern understanding and greater experi ence, so too does the Sherman Act’s prohibition on 'restraint[s] of trade’ evolve to meet the dynamics of present economic conditions." Leegin Creative Leather Prods., Inc. v. PSKS, Inc., - U.S. -, 127 S.Ct. 2705, 2720, 168 L.Ed.2d 623 (2007) (third alteration in original). The frequency with which we see bundled discounts in varied contexts does not insulate such discounts from antitrust review, but it heightens the need to ensure that the rule adopted does not expose inventive and legitimate forms of price competition to an overbroad liability standard.

6 . The Supreme Court has recognized the principle that package pricing is usually pro-competitive, noting that "[bjuyers often find package sales attractive; a seller's decision to offer such packages can merely be an attempt to compete effectively — conduct that is entirely consistent with the Sherman Act.” Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 12, 104 S.Ct. 1551, 80 L.Ed.2d 2 (1984).

7 . The academic literature provides other examples of ways in which sellers benefit from bundling. See, e.g., Crane, supra, 55 Emory L.J. at 430-43 (suggesting sellers can use bundles to instill customer loyalty, lower net prices to consumers by eliminating multiple monopoly-price markups on complementary goods, and price discrimination); see also Antitrust Modernization Comm’n, Report and Recommendations 95 (2007) (suggesting sellers can use bundled discounts to increase demand in lieu of advertising,-encourage use of a new product, or enter a new market).

8 . Judge Scirica and then-Judge Alito joined Judge Greenberg's dissent from the majority opinion. The Third Circuit reaffirmed the rule of LePage’s in United States v. Dentsply Int’l, Inc., 399 F.3d 181, 187 (3d Cir.2005).

9 . After oral argument, we issued an order inviting amicus briefing on the issue of whether a plaintiff seeking to establish the anticompetitive conduct element of an attempted monopolization claim by showing that the defendant offered bundled discounts must prove that the defendant's prices were below the defendant’s costs. Cascade Health Solutions v. Peacehealth, 479 F.3d 726, 727 (9th Cir.2007). We also sought input on the appropriate measure of costs if a plaintiff must prove below-cost pricing. Id. Finally, we asked amici who were arguing that a plaintiff should not be required to prove below-cost pricing to suggest alternative standards for the trier of fact to use in determining whether bundled discounts are anticompetitive. Id. We thank the many amici who accepted our invitation for their thoughtful briefs.

10 .Congress created the AMC in the Antitrust Modernization Commission Act of 2002, Pub.L. No. 107-273, §§ 11051-60, 116 Stat. 1758, 1856-59. The Act entrusted the AMC with four tasks: (1) soliciting the views of all parties concerned with the federal antitrust laws; (2) examining whether the antitrust laws needed modernization; (3) evaluating proposals to modernize the antitrust laws; and (4) submitting a report to the President and Congress containing a statement of the AMC's findings and conclusions and recommending any legislative or administrative action the AMC considered appropriate. See id. §§ 11053,11058. The procedure for appointing the twelve commissioners ensured that both major political parties were equally represented on the AMC. See id. § 11054.

11 . Section 7 of the Clayton Act forbids acquisitions that "substantially ... lessen competition [] or tend to create a monopoly.” 15 U.S.C. § 18.

12 . McKenzie contends that Brooke Group is not persuasive in this case because Brooke Group dealt with liability for primary-line price discrimination in violation of § 2(a) of the Robinson-Patman Act, whereas this case arises under § 2 of the Sherman Act. However, the Court made clear in Brooke Group that, whether a predatory pricing claim arises under § 2(a) of the Robinson-Patman Act or § 2 of the Sherman Act, the concerns are essentially the same, noting that:

There are, to be sure, differences between the two statutes. For example, we interpret § 2 of the Sherman Act to condemn predatory pricing when it poses "a dangerous probability of actual monopolization,” whereas the Robinson-Patman Act requires only that there be "a reasonable possibility” of substantial injury to competition before its protections are triggered. But whatever additional flexibility the Robinson-Patman Act standard may imply, the essence of the claim under either statute is the same: A business rival has priced its products in an unfair manner with an object to eliminate or retard competition and thereby gain and exercise control over prices in the relevant market.

Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578 (citations omitted).

13 . Of course, even if the exclusionary conduct element is satisfied by bundled discounts at price levels that yield a conclusion of below-cost sales, under the appropriate measure, there cannot be Sherman Act § 2 liability for attempted monopolization unless the other elements of a specific intent to monopolize and dangerous probability of success are satisfied.

14 . In the academic literature, this standard is sometimes referred to as a "discount allocation” or "discount reallocation” standard. See e.g., Daniel A. Crane, Multiproduct Discounting: A Myth of Nonprice Predation, 72 U. Chi. L.Rev. 27, 28 (2005).

15 . A variation of the example from Ortho illustrates how the discount attribution standard condemns discounts that could not be matched by an equally or more efficient producer of the competitive product. Recall that the example involves A, a firm that makes both shampoo and conditioner. A’s incremental cost of shampoo is $1.50 and A's incremental cost of conditioner is $2.50. A prices shampoo at $3 and conditioner at $5, if purchased separately. However, if purchased as a bundle, A prices shampoo at $2.25 and conditioner at $3. Purchased separately from A, the total price of one unit of shampoo and one unit of conditioner is $8. However, with the bundled discount, a customer can purchase both products from A for $5.25, a discount of $2.75 off the separate prices, but at a price that is still above A's variable cost of producing the bundle. Applying the discount attribution rule to the example; we subtract the entire discount on the package of products, $2.75, from the separate per unit price of the competitive product, shampoo, $3. The resulting effective price Of shampoo is thus $0.25, meaning that, if a customer must purchase conditioner from A at the separate price of $5, a rival who produces only shampoo must sell the shampoo for $0.25 to make customers indifferent between A's bundle and the separate purchase of conditioner from A and shampoo from the hypothetical rival. A’s pricing scheme thus has the effect of excluding any potential rival who would produce only shampoo, and would produce it at an incremental cost above $0.25. However, as we noted above, A’s incremental cost of pro ducing shampoo is $1.50. Thus, A's pricing practices exclude potential competitors that could produce shampoo more efficiently than A (i.e., at an incremental cost of less than $1.50). A’s discount could thus be considered exclusionary under our rule, supporting Sherman Act § 2 liability if the other elements were proved.

16 . Professor Nalebuff identifies the practical problem of calculating a rival firm's costs as a compelling argument in favor of a standard that focuses on whether bundled discounts would exclude a hypothetical equally efficient competitor:

There is ... a practical problem in determining if a rival firm is equally efficient or not. The problem is compounded for the monopolist who is looking for a bright line test to know whether its bundled pricing might be exclusionary or not. The solution to both these problems is to pick the monopolist itself as the equally efficient rival.

Nalebuff, supra, 50 Antitrust Bull. at 330.

17 . Although the volume of case law dealing with bundled discounting is small, one thirty-year-old case shows that antitrust claims based on bundled discounting practices are nothing new under the sun. See SmithKline Corp. v. Eli Lilly & Co., 427 F.Supp. 1089, 1124 (E.D.Pa.1976), aff’d, 575 F.2d 1056 (3d Cir.1978).

18 . A 1975 article by Professors Areeda and Turner ignited the modern debate about predatory pricing, see Phillip Areeda & Donald F. Turner, Predatoiy Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L.Rev. 697 (1975), and many prominent antitrust scholars weighed in on the topic in the following decade-and-a-half. See, e.g., Robert H. Bork, The Antitrust Paradox 154-55 (1978); George A. Hay, Predatory Pricing, 58 Antitrust L.J. 913 (1989); Frank H. Easterbrook, Predatoiy Sti-ategies and Counterstrategies, 48 U. Chi. L.Rev. 263 (1981); Paul L. Joskow & Alvin K. Klevorick, A Framework for Analyzing Predatoiy Pricing Policy, 89 Yale L.J. 213 (1979); William J. Baumol, Quasi-Permanence of Price Reductions: A Policy for Prevention of Predatory Pricing, 89 Yale L.J. 1 (1979); Oliver E. Williamson, Predatoiy Pricing: A Strategic and Welfare Analysis, 87 Yale L.J. 284 (1977); see also Wesley J. Leibeler, Whither Predatory Pricing? From Areeda and Turner to Matsu-shita, 61 Notre Dame L.Rev. 1052 (1986) (discussing the history of predatoiy pricing theory in the courts and academic literature).

19 . At least one circuit has held that average total cost, not average variable cost, is the appropriate baseline for determining predation. See McGahee v. N. Propane Gas Co., 858 F.2d 1487, 1500 (11th Cir.1988). However, such an approach is inconsistent with the Supreme Court’s instruction in Brooke Group that predatory prices are those below " 'some measure of incremental cost.’ ” Brooke Group, 509 U.S. at 223, 113 S.Ct. 2578 (quoting Cargill, 479 U.S. at 117-18 n. 12, 107 S.Ct. 484) (emphasis added). As the Antitrust Law treatise explains:

In the ordinary case a predator increases output out of existing facilities, cutting the price to predatory levels. For this reason the Supreme Court has emphasized that predators must have excess capacity from which to produce the increased output. But in that case, thfe only “incremental” cost of the predation is variable costs.

3 Areeda & Hovenkamp, supra, ¶ 741c at 444 .(2d ed.2002) (footnote omitted).

20 . In a number of cases decided before Brooke Group, we held that pricing below marginal cost or average variable cost provided evidence that a pricing scheme was predatory, but also held that that mode of proof was not exclusive. See Transamerica Computer Co. v. IBM Corp., 698 F.2d 1377, 1385 (9th Cir.1983); Inglis, 668 F.2d at 1033. We suggested that an above-cost pricing policy could be predatory if accompanied by evidence of predatory intent, market power, or “long-run behavior.” See Transamerica, 698 F.2d at 1387. Other circuits rejected the notion that predation could be proved through evidence of intent alone, see, e.g., Barry Wright, 724 F.2d at 232, and Brooke Group's holding that “a plaintiff seeking to establish competitive injury resulting from a rival’s low prices must prove that the prices complained of are below an appropriate measure of its rival's costs” put to rest any notion that predation can be proven through evidence of intent alone, Brooke Group, 509 U.S. at 222, 113 S.Ct. 2578.

21 . As we noted above, the AMC’s proposed standard in bundled discounting cases, in addition to requiring below-cost pricing, also contains two further proposed elements.

The second element proposed by the AMC is that there is a dangerous probability that the defendant will recoup its investment in the bundled discounting program. AMC Report, supra, at 99. This requirement, adopted from Brooke Group, is imported from the single product predatory pricing context, but we think imported incorrectly. We do not believe that the recoupment requirement from single product cases translates to multi-prod-uct discounting cases. Single-product predatory pricing, unlike bundling, necessarily involves a loss for the defendant. For a period of time, the defendant must sell below its cost, with the intent to eliminate its competitors so that, when its competition is eliminated, the defendant can charge - supracompetitive prices, recouping its losses and potentially more. By contrast, as discussed above, exclusionary bundling does not necessarily involve any loss of profits for the bundled discounter. See Nalebuff, supra, 50 Antitrust Bull. at 327. As the example from Ortho illustrates, a bundled discounter can exclude its rivals who do not sell as many product lines even when the bundle as a whole, and the individual products within it, are priced above the discounter’s incremental cost to produce them. The trier of fact can identify cases that present this possibility for anticompetitive exclusion by applying the discount attribution standard outlined above. Under that standard, the ultimate question is whether the bundled discount would exclude an equally efficient rival. But because discounts on all products in the bundle have been allocated to the competitive product in issue, a conclusion of below-cost sales under the discount attribution standard may occur in some cases even where there is not an actual loss because the bundle is sold at a price exceeding incremental cost. In such a case, we do not think it is analytically helpful to think in terms of recoupment of a loss that did not occur.

The third element proposed by the AMC is that “the bundled discount or rebate program has had or is likely to have an adverse effect on competition.’’ AMC Report, supra, at 99. We view this final element as redundant because it is no different than the general requirement of “antitrust injury” that a plaintiff must prove in any private antitrust action. See Brunswick, 429 U.S. at 489, 97 S.Ct. 690 (defining antitrust injury as "injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants’ acts unlawful” and noting that "[t]he injury should reflect the anticompetitive effect either of the violation or of anticompetitive acts made possible by the violation”).

For these reasons, while adopting the AMC’s proposal to require below-cost sales to prove exclusionary conduct, we, do not. adopt the element of recoupment, which we think may be inapplicable in some cases, and we do not adopt the element of “adverse effect on competition” as we think that is superfluous in light of the general and pre-existing requirement of antitrust injury under Brunswick.

22 . PeaceHealth also argues that because the jury found in its favor on McKenzie’s claim of exclusive dealing under § 1 of the Sherman Act, it is entitled to judgment as a matter of law on McKenzie’s claim of attempted monopolization under § 2 of the Sherman Act. Our vacatur of the jury’s verdict on the attempted monopolization claim makes it unnecessary for us to fully address that argument. However, we previously have held that "[t]he 'predatory or anticompetitive conduct’ element of § 2 attempt, like the conduct element of monopolization, encompasses more than violations of § 1.” Cal. Computer Prods., Inc. v. IBM Corp., 613 F.2d 727, 737 (9th Cir.1979). Specifically, § 1 is limited to concerted activity, while § 2 reaches unilateral exclusive conduct. See id.; Microsoft, 253 F.3d at 70.

23 . Price discrimination claims can take three forms: primary line, secondary line, or tertiary line. Primary-line price discrimination includes conduct, like predatory pricing, that injures the direct competitors of the discriminating seller. See Volvo Trucks, 546 U.S. at 176-77, 126 S.Ct. at 870. Secondary-line price discrimination includes conduct that injures competition between the customers of a discriminating seller, as, for example, when one seller charges two customers different prices for the same good. See id. Tertiary-line price discrimination includes conduct that injures competition between the customers of the direct purchasers of a discriminating seller. See id.

24 . The Oregon statute provides:

It is unlawful for any person engaged in commerce or food commerce, or both, in the course of such commerce, either directly or indirectly, to discriminate in price between different purchasers of commodities, or services or output of a service trade, of like grade and quality or to discriminate in price between different sections, communities or cities or portions thereof or between different locations in sections, communities, cities or portions thereof in this state, where the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce, or to injure, destroy or prevent competition with any person who either grants or knowingly receives the benefit of such discrimination, or with customers of either of them.

Or.Rev.Stat. § 646.040(1).

25 .The Robinson-Patman Act provides:

It shall be unlawful for any person engaged in commerce, in the course of such commerce, either directly or indirectly, to discriminate in price between different purchasers of commodities of like grade and quality, where either or. any of the purchases involved in such discrimination are in commerce, where such commodities are sold for use, consumption, or resale within the United - States or any Territory thereof or the District of Columbia or any insular possession or other place under the jurisdiction of the United States, and where the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce, or to injure, destroy, or prevent competition with any person who either grants or knowingly receives the benefit of such discrimination, or with customers of either of them....

15 U.S.C. § 13(a).

26 . A § 1 violation can also occur when the customer promises not to take the tied product from the defendant's competitor, but courts "rarely encounterf]'' such a situation. 10 Areeda & Hovenkamp, supra, ¶ 1752c n. 8 at 263 (2d ed.2004).

27 . For criticism of the leverage theory, see Bork, supra, at 372. See also Christopher R. Leslie, Cutting Through Tying Theory With Occam’s Razor: A Simple Explanation of Tying Arrangements, 78 Tul. L.Rev. 727, 731-41 (2004) (summarizing the conflict between leverage theorists and the Chicago School).

28 . McKenzie has filed a motion to strike portions of PeaceHealth’s brief citing to evidence that insurers had alternatives to taking all services from PeaceHealth because the evidence PeaceHealth cites was not in the portion of the record designated to the district court on summary judgment. Specifically, McKenzie argues that we should not permit PeaceHealth on appeal to refer to portions of exhibits that, while submitted to the district court in support of PeaceHealth's motion for summary judgment, did not have their relevant portions highlighted for the district court.

Under the district court's Local Rule 56.1(c)(3), the moving party is required to highlight relevant portions of documents presented to the court, and under Local Rule 56.1(e), "the Court has no independent duty to search and consider any part of the court record not otherwise referenced in the separate concise statements of the parties.” D. Or. R. 56.1. However, all of the documents cited by PeaceHealth on appeal were before the district court, even if not highlighted, ’ Moreover, the principal policy underlying local rules like Rule 56.1 is to obviate the need for the district court to search the record for facts relevant to summary judgment. Delange v. Dutra Constr., Co., 183 F.3d 916, 919 n. 2 (9th Cir.1999) (per curiam). Such a policy has no impact on the scope of our appellate review. See Fed. R.App. P. 10(a) (stating that the record on appeal consists of all papers and exhibits filed in the district court). We therefore deny McKenzie’s motion to strike portions of PeaceHealth’s combined brief.

29 . The Supreme Court has also held that the unique character of the tying product can provide a basis for holding that a defendant has sufficient economic power in the tying product, market to coerce acceptance of the tied product. See U.S. Steel Corp. v. Fortner Enters., Inc., 429 U.S. 610, 619, 97 S.Ct. 861, 51 L.Ed.2d 80 (1977) [hereinafter Fortner II ] (citing N. Pac. Ry. Co. v. United States, 356 U.S. 1, 78 S.Ct. 514, 2 L.Ed.2d 545 (1958); Int’l Salt Co. v. United States, 332 U.S. 392, 68 S.Ct. 12, 92 L.Ed. 20 (1947)). The Court in Fortner II recognized that the key question in establishing sufficient market power is whether the seller has some cost advantage not shared by its competitors which makes its competitors unable to provide the tying product and that a mere showing that its competitors did not want to provide the tying product is insufficient .to establish an illegal tie. Id. at 621-22, 97 S.Ct. 861. At the summary judgment stage, the evidence presented by McKenzie was sufficient to create a factual issue about whether McKenzie could not provide tertiary services or whether it was simply unwilling, as a matter of business strategy, to provide tertiary services.

30 . If, on remand, McKenzie stakes its tying claim not on a theory that PeaceHealth explicitly (e.g., by contract) or implicitly coerced insurers to purchase primary and secondary services from PeaceHealth as a condition to obtaining tertiary1 services, but on a theory that PeaceHealth's bundled discounts effectively left insurers with no rational economic choice other than purchasing tertiary services from PeaceHealth, such a claim might raise the question of whether, to establish the coercion element of a tying claim through a bundled discount, McKenzie must prove that Pea-ceHealth priced below a relevant measure of its costs. Some commentators would require a plaintiff alleging that a bundled discount amounts to an illegal tie to prove below-cost prices. See, e.g., 3 Areeda & Hovenkamp, supra, ¶ 749b2 at 334 (Supp.2006). It is unclear whether the AMC intended its three:part test to apply when a plaintiff alleging an illegal tying arrangement asserts that the defendant's pricing practices coerced unwanted purchases of the tied product. See AMC Report, supra, at 114 n. 157 ("The recommended three-part test is proposed here for challenges to bundled pricing practices, and its purpose, as the text explains, is to avoid deterring procompetitive price reductions. The Commission is not recommending application of this test outside the bundled pricing context, for example in tying or exclusive dealing cases. The Commission did not undertake to study tying and exclusive dealing issues more generally.”). The parties have not briefed this issue to us, and the parties did not raise the issue before the district court. We therefore leave it to the district court, if necessary, to decide the issue in the first instance on remand. Singleton v. Wulff, 428 U.S. 106, 121, 96 S.Ct. 2868, 49 L.Ed.2d 826 (1976).

31 . In No. 05-35627, we also decline to address McKenzie's Noeir-Pennington arguments because these related to an evidentiary ruling and the issue may not arise on a retrial. Further, we hold that the district court’s jury instruction on combination or conspiracy was not an abuse of discretion.