The provision of health care financing services has become increasingly competitive. Hospitals and physicians (the providers of service) have begun to offer financing packages, much as automobile manufacturers sometimes finance their own products. In health care, where the need for service often depends on events beyond anyone’s control, financing often is combined with insurance to spread the risks.
Sometimes the financing and insurance package is part of a new method of supplying the service; the health maintenance organization (HMO) is both a method of joining physicians in a firm and a method of financing their service by selling memberships for stated monthly prices. The physicians at HMOs are paid salaries rather than fees for each service they render. Sometimes the financing is independent of the method of supplying the service. Several hospitals in Indiana that use traditional organization (most physicians are independent contractors rather than employees), and pay each provider per service, also have begun to offer financing to patients.
One package is the preferred provider organization (PPO). In exchange for a *1330stated monthly payment, the hospital promises to pay the costs of patients who use particular providers. Patients who use providers other than the “preferred” ones must pay part or all of the fees themselves. These copayments are meant to induce patients to stick with the preferred providers — perhaps more often to request their physicians to stick with the preferred providers.1 This may send extra business to these providers, who in exchange may agree to take less for each case. The assembler of the PPO plan may pass the savings along to the purchasers of the coverage. A PPO plan specifies in advance the fee it will pay a provider for any given medical service.
The purchasers of the service are not necessarily the patients. Employers often supply health care coverage for employees, and they are intensely interested in reducing the price of any given level of care. These employers may shop among different plans assembled by different HMOs and hospitals. They also consider traditional insurance packages. Blue Gross and Blue Shield of Indiana (the Blues) offer a service benefit plan that has attracted a substantial following from both employers and individual purchasers of insurance. Providers sign up as members of the Blues’ plan, and they promise to accept as payment the “usual, customary, and reasonable” fee for a service as the Blues determine that fee to be, case by case. Other sellers of insurance agree to pay stated amounts per type of service rendered or to pay a stated percentage of the provider’s bill. These plans have attracted fewer subscribers. Some large employers simply hire insurance companies to administer the employers’ own plans; the administrator receives the bills, the employer determines what it will pay, and the administrator sends the money on to the providers. Administered self-insurance has been growing at the expense of other plans.
Patients and employers must choose among these plans. Once they have chosen a plan, they may have little control over their care. Choosing a PPO plan or HMO may lock a person into a particular provider. On the other hand, the choice among plans is relatively free in advance, and patients may shop among plans that are compatible with their needs and with their physicians’ limitations (each physician will have privileges at a subset of local hospitals). This case is about the choice among financing packages.
I
The plaintiffs in this case are 80 acute-care hospitals (the Hospitals). All 80 provide care on a fee for service basis, and all 80 receive payments from many insurance plans and administered self-insurance plans as well as from patients. Some of the 80 also offer PPO plans; others are preparing to do so. Forty of the 80 plaintiffs have appealed.
The Blues have been losing market share in Indiana for some years. In 1980 the Blues insured almost two million of Indiana's 5.5 million population. By 1984 they insured only about 1.45 million people. This is still a large share; at some of the Hospitals more than 80% of all patients are covered by the Blues, and throughout Indiana about 50% of all hospitals’ revenues come from payments made by the Blues. This may be a misleading figure because it includes payments the Blues made as administrators of self-insurance plans and of Indiana’s Medicare plan. The Blues say that they are much smaller — they insure only about 27% of all patients in Indiana and distributed in 1982 only about $450 million in Indiana on behalf of privately-insured patients, while Indiana’s hospitals re*1331ceived some $2.2 billion from all sources. By all accounts the Blues are large in relation to the next-largest private supplier of health insurance in Indiana, which underwrites about 3% of all private insurance in the state. Just how “large” the Blues are turns out not to matter, so we do not pursue the question.
All agree that however large the Blues may be, they are losing business. Concerned about this, the Blues decided to offer a PPO of their own, in addition to their traditional service benefit plans. The Blues also decided to merge, eliminating the longstanding practice of one plan’s underwriting hospitals’ services and another’s underwriting physicians’ services. The Blues asked for bids from all acute-care hospitals in Indiana and invited each to bid a percentage discount from its regular fees. That PPO plan and merger precipitated this case. The hospitals that offer PPO plans saw the Blues’ decision as a threat to their success. All hospitals saw a PPO plan as a threat to revenues — those who participated in the plan might collect less per service rendered, and those outside the plan might lose volume.
Ninety-one of Indiana’s 115 acute-care hospitals submitted bids, and the Blues signed up 61 of the 91. Forty-two of the 80 plaintiffs are among the 61. Eleven plaintiffs did not bid, and 27 bid but were not selected. All 80 remain eligible to participate in the regular service benefit plan offered by the Blues, which is the Blues’ most popular product. All hospitals in Indiana also may provide services to patients covered by the Blues’ PPO, but the Blues will reimburse only 75% of the hospitals’ fees; the patients must pay the rest. The Blues will reimburse 100% of the agreed charges when insureds use hospitals within the PPO.
The Blues wanted to put their PPO into effect early in 1985. The Hospitals began this suit on November 14, 1984, seeking injunctive relief against the Blues’ proposed PPO under sections 1 and 2 of the Sherman Act, 15 U.S.C. §§ 1 and 2, and provisions of Indiana law.
The district court set the case for a hearing on the Hospitals’ request for a preliminary injunction. The court told the Hospitals they could present as much evidence as they wanted, and the hearing lasted 11 days in February 1985. More than 30 witnesses testified; more than 400 exhibits were introduced. On March 1, 1985, the district court denied the request for a preliminary injunction. The Blues’ PPO immediately went into effect. The court later entered a partial final judgment disposing of the Hospitals’ claims under state law, while reserving their antitrust claims; it issued a certificate under Fed.R.Civ.P. 54(b) permitting immediate appeal of the state law issues. We therefore have before us the denial of preliminary relief under the Sherman Act and the final judgment under state law.
The district court made extensive findings of fact. 603 F.Supp. 1077 (S.D.Ind. 1985). The most important of these concern the Hospitals’ claim that the Blues have (and abused) “market power,” the ability to raise price significantly higher than the competitive level by restricting output. See NCAA v. University of Oklahoma, 468 U.S. 85, 104 S.Ct. 2948, 2965-67 & n. 38, 82 L.Ed.2d 70 (1984); United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 391, 76 S.Ct. 994, 1005, 100 L.Ed. 1264 (1956); William M. Landes & Richard A. Posner, Market Power in Antitrust Cases, 94 Harv.L.Rev. 937 (1981). The court found that the Blues do not have the power to restrict output in the market or to raise price because they furnish a fungible product that other people can and do supply easily.
The court treated the product as “health care financing.” The Blues, other insurance companies, hospitals offering PPOs, HMOs, and self-insuring employers all offer methods of financing health care. Employers and individual prospective patients easily may switch from one financing package to another; nothing binds an employer or patient to one plan. To put it differently, even though everyone wants medical insurance, and the demand for this service *1332as a whole may be inelastic (meaning that purchases fall less than 1% in response to a 1% increase in price, which makes the increase profitable), when customers are not tied to particular sellers each seller may perceive the demand as highly elastic (meaning that customers will quickly switch if any one supplier raises price, which makes the increase unprofitable). The court concluded: “Consumers are extremely price sensitive and will readily switch on the basis of price from one company or form of financing to another. Consequently, no competitor ... has the power to control prices____” Finding 8, 603 F.Supp. at 1080.
The market in health care financing is competitive, the court concluded, not only because customers can switch readily but also because new suppliers can enter quickly and existing ones can expand their sales quickly. More than 1000 firms are licensed to sell health insurance in Indiana, and more than 500 sell this insurance currently. According to the district court, all can expand on a moment’s notice. “Entry barriers into the market for health care financing are extremely low. All that is needed to compete in Indiana, for example, is sufficient capital to underwrite the policies and a license from the Indiana Insurance Commissioner.” Finding 11, 603 F.Supp. at 1080. Of the 500 firms now selling insurance, many operate nationwide and have (or can attract) plenty of capital against which to write policies — if the price is right. The court listed “Prudential, Aetna, Metropolitan and Equitable, each of which [has] premium income and assets in the tens of billions and operates nationally.” Finding 12, 60S F.Supp. at 1080. The court also observed that firms may elect self-insurance, and HMOs may expand, in response to an increase in the price of insurance.
Buyers’ willingness to switch and sellers’ ability to enter and expand rapidly, the district court concluded, means that “a firm’s share of premium revenues reflects no more than its ability to compete successfully in meeting consumer demands.” Finding 14, 603 F.Supp. at 1081. The Blues cannot exclude competitors, cannot raise prices without losing business quickly; the Blues’ size therefore indicates only their success in offering the package of price and service that customers prefer, not any market power.
The district court also found that PPO plans “contain cost by promoting price competition among hospitals” (Finding 18, 603 F.Supp. at 1081) and that many of the large national insurers, as well as the larger hospitals in Indiana, are offering or planning to offer PPO plans. “By thus offering financing arrangements to consumers to pay for hospital services, these hospitals [offering PPOs] are vertically integrating into the health care financing market” (Finding 20, 603 F.Supp. 1082).
The PPO program will enable the Blues to offer lower premiums, the court found; it estimated the savings at 10-20% for 1985. Finding 32, 603 F.Supp. at 1083. These savings come from the increased utilization of hospital services made possible by the patients’ incentives to use the selected providers, and from “utilization controls and procedures ... intended to generate savings by eliminating needless in-patient admissions and unnecessary operations.” Finding 39, 603 F.Supp. at 1084. Insurance creates “moral hazard.” Once a person has insurance, he wants the best care regardless of cost — for someone else bears the cost. When, as happens often, the physician rather than the patient makes the important choices, the physician may be inclined to provide all the service for which insurance will pay, knowing that his patient will not resist this recommendation (at least not on account of expense). Yet if every physician supplies more or better care, price must rise, which patients as a group must pay in higher insurance bills. The “utilization controls” to which the district court referred are a method of counteracting moral hazard by limiting each insured’s access to care. These limits may restore the appropriate level of care and save cost. Moving patients from one hospital to another also may save cost. It may be cheaper to operate one hospital at 95% *1333of capacity and another at 55%, rather than each at 75%; the less-used hospital can close a wing and reduce its staff.
In light of its conclusions on the benefits of PPOs, the district court thought that a preliminary injunction would injure rather than promote the public interest. “The public ... would forfeit the benefits of competition- and the opportunity for decreased health care costs. Competition would be restrained in the health care financing market because Blue Cross/Blue Shield cannot offer its product to the public____ Competition would also be restrained in the hospital services market because hospitals would not have to compete against one another on the basis of price for Blue Cross/Blue Shield insureds. In addition, the public would lose the improved hospital utilization control which the PPO would sponsor.” Finding 43, 603 F.Supp. at 1084. Because the Blues lack market power, too, the district court thought the Hospitals had no reasonable chance of success on the merits of their antitrust claims. With the public interest and the law coinciding, the court concluded, the Blues must prevail.
The court also made findings pertinent to the Hospitals’ claim that the PPO violates state law. We postpone recitation of these to the discussion of state law, and we turn to the antitrust argument.
II
The Hospitals insist that the public interest is on their side and that they are likely to prevail on the merits. These are related considerations. See Lawson Products, Inc. v. Avnet, Inc., 782 F.2d 1429, 1433-34 (7th Cir.1986); American Hospital Supply Corp. v. Hospital Products Ltd., 780 F.2d 589, 593-94 (7th Cir.1986); Roland Machinery Co. v. Dresser Industries, Inc., 749 F.2d 380, 387-88 (7th Cir.1984). The district court, when exercising its discretion in evaluating and weighing the factors of traditional injunction analysis, seeks to hold to a minimum the sum of two potential costs: the cost of denying an injunction if the plaintiff is ultimately determined to be entitled to relief, and the cost of granting an injunction if the defendant is ultimately determined to have violated no legal command.
These costs usually fall on the parties. They are also likely to be short term costs, so that a mistake is not catastrophic. In antitrust litigation, by contrast, third parties may feel substantial effects. This suit is between hospitals and an insurer, but the principal effect of the PPO plan is on the price patients pay for insurance. A mistaken grant of an injunction may elevate this price, harming the consumers that antitrust laws are designed to protect. If there is a mistake, it may last a long while. Antitrust cases are notoriously extended. Sometimes preliminary injunctions in antitrust cases condemn the proposed action, and the defendant abandons it rather than face the costs and uncertainties of lengthy litigation. The Third Circuit took judicial notice of this effect of preliminary injunctions in holding that targets of tender offers may not seek relief on antitrust grounds. H.H. Robertson Co. v. Guardian Industries Corp., No. 85-3232 (3d Cir.1986).
As a rule a court need not dwell on the costs to consumers of mistaken injunctions, because plaintiffs have no reason to seek remedies that will injure the beneficiaries of the antitrust laws. The plaintiffs may be consumers themselves or may have interests identical to those of consumers. At other times, however, the plaintiff’s interests do not coincide with those of consumers. For example, in Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 97 S.Ct. 690, 50 L.Ed.2d 701 (1977), the plaintiff operated bowling establishments in competition with the defendant Brunswick, which acquired additional establishments in violation of § 7 of the Clayton Act. The plaintiff suffered injury in fact because the price bowling establishments could charge to consumers fell after the acquisition; had Brunswick not acquired the lanes, they would have gone bankrupt and disappeared. Pueblo therefore had been injured as a result of a violation — but *1334not because the violation had raised prices to consumers. The Supreme Court held that Pueblo had not suffered “antitrust injury,” which means injury from higher prices or lower output, the principal vices proscribed by the antitrust laws. Whenever the plaintiff and consumers have divergent rather than congruent interests, there is a potential problem in finding “antitrust injury.” If, as in Brunswick itself, the plaintiff and the defendant are competitors, the. plaintiff gains from higher prices and loses from lower prices — just the opposite of the consumers’ interest. When the plaintiff is a poor champion of consumers, a court must be especially careful not to grant relief that may undercut the proper functions of antitrust.
Brunswick was a suit for damages, but an injunction would have been even worse than damages. If a court had issued an injunction compelling Brunswick to divest the establishments, which would have gone bankrupt, prices would have risen to consumers’ detriment. Because injunctions may injure consumers just as surely as damages may, the “antitrust injury” rule applies to requests for damages and injunctions alike. Midwest Communications, Inc. v. Minnesota Twins, Inc., 779 F.2d 444, 452-53 (8th Cir.1985); Schoenkopf v. Brown & Williamson Tobacco Corp., 637 F.2d 205 (3d Cir.1980).
The risk that consumers and plaintiffs may have divergent interests arises when, as in Brunswick, the plaintiff and the defendant are horizontal rivals. Then the plaintiff wants higher prices, consumers want lower prices. The books contain examples of firms that invoke the antitrust laws to obtain shelter from competition rather than to promote competition. One example from our court is ECOS Electronics Corp. v. Underwriters Laboratories, Inc., 743 F.2d 498, 501 (7th Cir.1984), cert. denied, — U.S. —, 105 S.Ct. 1178, 84 L.Ed.2d 327 (1985). See William J. Baumol & Janusz A. Ordover, Use of Antitrust to Subvert Competition, 28 J.L. & Econ. 247 (1985), for further examples.
The equitable remedy of preliminary injunctions has always stressed the importance of ensuring that preliminary injunctions not injure the public at the same time as they assist the plaintiff. Given the risk that business rivals may seek to use antitrust to stifle rather than promote competition, district courts should pay particular attention to the public interest in such litigation. Thus in attempting to weigh the equities of granting or denying a preliminary injunction in the antitrust setting, the pro or anti-competitive effects on the market at large should be an important factor in the district court’s analysis.
Some of the Hospitals offer PPO plans of their own. The district court found that Methodist Hospital has signed up more than 10,000 insureds in a PPO that includes seven hospitals and that Methodist Hospital has marketed this plan across the state (Finding 21, 603 F.Supp. at 1082). Many other plaintiff Hospitals “have developed or are developing” their own PPO plans (Finding 20, ibid.). Moreover, the district court also found that the public interest lies in the continuation of the Blues PPO, as we discussed above.
We start with a proposition established by Brillhart v. Mutual Medical Insurance, Inc., 768 F.2d 196 (7th Cir.1985): the Blues are financial intermediaries, purchasing agents for the consumers of medical services. See also Kartell v. Blue Shield of Massachusetts, Inc., 749 F.2d 922 (1st Cir.1984), cert. denied, — U.S. —, 105 S.Ct. 2049, 85 L.Ed.2d 322 (1985). The Blues, as financial intermediaries, may drive any bargains open to the consumers of services. The Rule of Reason rather than the per se rule supplies the standard of analysis.
The analysis of the adoption of the PPO plan must begin with an assessment of market power. Market power is a necessary ingredient in every case under the Rule of Reason. See Polk Bros., Inc. v. Forest City Enterprises, Inc., 776 F.2d 185, 191 (7th Cir.1985) (collecting cases). Unless the defendants possess market power, it is unnecessary to ask whether their *1335conduct may be beneficial to consumers. Firms without power bear no burden of justification. The Hospitals say that the Blues have a large share of the market for medical insurance in Indiana, and that this establishes market power.
In many cases a firm’s share of current sales does indicate power. Sales may reflect the ownership of the productive assets in the business. Market power comes from the ability to cut back the market’s total output and so raise price; consumers bid more in competing against one another to obtain the smaller quantity available. When a firm (or group of firms) controls a significant percentage of the productive assets in the market, the remaining firms may not have the capacity to increase their sales quickly to make up for any reduction by the dominant firm or group of firms.
In other cases, however, a firm’s share of current sales does not reflect an ability to reduce the total output in the market, and therefore it does not convey power over price. Other firms may be able, for example, to divert production into the market from outside. They may be able to convert other productive capacity to the product in question or import the product from out of the area. If firms are able to enter, expand, or import sufficiently quickly, that may counteract a reduction in output by existing firms. And if current sales are not based on the ownership of productive assets — so that entrants do not need to build new plants or otherwise take a long time to supply consumers’ wants— the existing firms may have no power at all to cut back the market’s output. To put these points a little differently, the lower the barriers to entry, and the shorter the lags of new entry, the less power existing firms have. When the supply is highly elastic, existing market share does not signify power. Will v. Comprehensive Accounting Corp., 776 F.2d 665, 672 n. 3 (7th Cir.1985); United States v. Waste Management Inc., 743 F.2d 976 (2d Cir.1984).
The district court found that each of the factors suggesting that market share does not imply market power is present in the market for medical insurance. New firms may enter easily. Existing firms may expand their sales quickly; the district court pointed out that insurers need only a license and capital, and that firms such as Aetna and Prudential have both. There are no barriers to entry — other firms may duplicate the Blues’ product at the same cost the Blues incur in furnishing their coverage. See George J. Stigler, The Organization of Industry 67-70 (1968) (defining barriers to entry as differentials in the long-term costs of production); cf. Harold Demsetz, Barriers to Entry, 72 Am.Econ. Rev. 47 (1982) (showing that not all barriers, as so defined, injure effective competition). The Blues and other nonprofits may have an edge because of the lower tax Indiana places on premiums paid to them, but this sort of advantage is not pertinent here. Other mutual insurance carriers (including Prudential) can get the same tax break. A PPO plan does not exploit the tax advantage as compared with any other plan the Blues could offer. The tax benefits may or may not be desirable as a matter of state policy, but this is no concern of antitrust law.
The Blues do not own any assets that block or delay entry. The insurance industry is not like the steel industry, in which a firm must take years to build a costly plant before having anything to sell. The “productive asset” of the insurance business is money, which may be supplied on a moment’s notice, plus the ability to spread risk, which many firms possess and which has no geographic boundary. Cf. Hood v. Tenneco Texas Life Insurance Co., 739 F.2d 1012, 1019 (5th Cir.1984) (insurance industry marked by ease of entry); Alabama Association of Insurance Agents v. Board of Governors, 533 F.2d 224, 250-51 (5th Cir.1976) (financial services in general are competitive because of the ease of moving money), modified, 558 F.2d 729 (1977), cert. denied, 435 U.S. 904, 98 S.Ct. 1448, 55 L.Ed.2d 494 (1978). The district court emphasized that every firm can expand its sales quickly if the price is right, that no *1336firm has captive customers, and that many firms want to serve this market. The conclusion that the Blues face vigorous and effective competition is not clearly erroneous. See also National Bancard Corp. v. VISA U.S.A., Inc., 779 F.2d 592, 604-05 (11th Cir.1986) (defining a market of “all payment devices” on basis of a conclusion that one financial service is a ready substitute for another).
Still, the Hospitals .say, the conclusion is legally irrelevant. Ease of entry and the absence of barriers do not matter if the defendant has a large market share. The Hospitals are wrong. Market share is just a way of estimating market power, which is the ultimate consideration. When there are better ways to estimate market power, the court should use them. See Waste Management, supra. Market share reflects current sales, but today’s sales do not always indicate power over sales and price tomorrow. United States v. General Dynamics Corp., 415 U.S. 486, 94 S.Ct. 1186, 39 L.Ed.2d 530 (1974), illustrates the point. The sellers of a large share of all current sales of coal in the midwest merged. The Court held, however, that share did not demonstrate power, because current deliveries of coal were largely committed under long term contracts. The pertinent competitive criterion was the ability to make future commitments of coal, and existing deliveries actually restricted the ability to make such commitments. The real “owners” of the coal currently being delivered were the recipients under the contracts, not the sellers. One of the firms in the merger had committed all of its economically-recoverable coal, and so its disappearance by merger did not remove from the market any competitive force that could be preserved by enjoining the merger. The Court concluded that because market shares did not reflect tomorrow’s ability to compete, they did not supply a reason to forbid the merger.
Other cases, many from this circuit, have said that market share is simply an indication of power and possesses no other significance. Will v. Comprehensive Accounting Corp., supra, 776 F.2d at 672 n. 3; United Airlines, Inc. v. CAB, 766 F.2d 1107, 1115 (7th Cir.1985); Kaiser Aluminum & Chemical Corp. v. FTC, 652 F.2d 1324, 1341 (7th Cir.1981); Juneau Square Corp. v. First Wisconsin National Bank of Milwaukee, 624 F.2d 798, 813 (7th Cir.), cert. denied, 449 U.S. 1013, 101 S.Ct. 571, 66 L.Ed.2d 472 (1980). Waste Management, supra, and Broadway Delivery Corp. v. United Parcel Service of America, Inc., 651 F.2d 122 (2d Cir.1981), cert. denied, 454 U.S. 968, 102 S.Ct. 512, 70 L.Ed.2d 384 (1982), express the same point in the Second Circuit, and almost every other circuit has a similar holding. Cf. du Pont, supra (sole maker of cellophane lacks market power because of competition from other flexible wrappings).
The inquiry in each case is the ability to control output and prices, an ability that depends largely on the ability of other firms to increase their own output in response to a contraction by the defendants. Indeed it is usually best to derive market share from ability to exclude other sources of supply. This is the method the Department of Justice adopted in its Merger Guidelines. Cf. Landes & Posner, supra; George J. Stigler & Robert A. Sherwin, The Extent of the Market, 28 J.L. & Econ. 555 (1985). If the definition of the market builds in a conclusion that there are no significant additional sources of supply and no substitutes from the consumers’ perspective, then the market share indicates power over price. But a calculation of the Blues’ share of current coverage in Indiana does not capture the possibility of new entry and expanded sales by rivals, and this is why the district court properly held that the geographic market “is regional, if not national” (Finding 9, 603 F.Supp. 1080). This larger market may not seem useful from the perspective of consumers in Indiana, who must obtain their insurance from firms offering it there. It is highly pertinent, however, from the perspective of the Blues’ rivals and potential rivals, and therefore from the perspective of constraints on the Blues’ ability to raise price. *1337The Blues’ rivals, whose mobility is not restricted, protect consumers, whose mobility is restricted.
The district court therefore did not commit a legal error or make a clear error in finding the facts. So far as the record stands, the Blues lack market power and are therefore entitled to adopt a PPO plan without further scrutiny under the Sherman Act. '
The merger of the two plans does not alter the analysis or independently violate the antitrust laws. The district court found that the plans “have for more than thirty years acted as one company” (Finding 2, 603 F.Supp. at 1079). It is therefore appropriate to treat them as if they had been one corporation all along. A merger, like a cartel, may “deprive[] the marketplace of the independent centers of decisionmaking that competition assumes and demands,” and the joinder of previously independent firms “suddenly increases the economic power moving in one particular direction.” Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 104 S.Ct. 2731, 2741, 81 L.Ed.2d 628 (1984). But the two Blue plans offered complementary products (insurance against hospitals’ costs and insurance against physicians’ costs); they did not compete by offering substitute products. Their merger did not change the conditions of competition in the market. The district court was entitled to treat them as a single firm under Copperweld. 7 Phillip E. Areeda, Antitrust Law ¶ 1464f (1986). Even if the two plans’ formal separation makes treatment under Copperweld inappropriate, the merger of firms that were jointly controlled does not call for close scrutiny. United States v. Citizens & Southern National Bank, 422 U.S. 86, 95 S.Ct. 2099, 45 L.Ed.2d 41 (1975).
Ill
The Hospitals try to avoid this conclusion by urging that § 2 of the Sherman Act imposes an additional test — that the Blues act without anticompetitive intent. Relying on United States v. Aluminum Co. of America, 148 F.2d 416 (2d Cir.1945), the Hospitals insist that any firm with a large market share has an obligation not to augment that share at the expense of rivals and may not drive hard bargains.
Although the district court did not make findings concerning the Blues’ intent, the Hospitals say that the record reeks of bad intent. The Blues evidently wanted to drive down the price they paid to the providers, which entails bad intent; more, the Blues engaged in calculated planning to preserve or enlarge their market share. The Hospitals quote from a report that a consulting firm rendered to the Blues in 1981: “The market is mature with little growth opportunity left. The [Blues] dominate the marketplace and recently have been losing shares____ The competition will move to increase market share at the [Blues’] expense. Market strategy for this market might be to segment the market.” In January 1983 the Blues circulated an internal report stating in part:
The Proposition: That [the Blues] use its market position and its control over substantial sums of health care dollars to negotiate lower fees for provider services ... Blue Cross plans that enjoy major discounts in their hospital reimbursement contracts do have considerable advantage over competitors. It appears that these discount arrangements reinforce what is a common perception among both political leaders and businessmen. Which is that our control of so many health care dollars should put us in a position to negotiate lower prices for provider services____
The Recommended Response: [The Blues are] in a unique position to serve as a broker between what might appear to be the conflicting interests of financially threatened providers and cost conscious group purchasers____ We also control an overwhelming share of the marketplace in key areas____ In short, the time seems right for an aggressive new stance in the financing of health care benefits____ The growth of competitive forces poses a grave threat, and will grow worse without counteraction.
*1338Res ipsa loquitur, the Hospitals say. If the Blues lack market power, how come their own planning documents talk this way? The Hospitals say that the district court should have inferred power from intent, and that given this intent to prevent competition the district court also was required to enjoin the adoption of the PPO.
We assume without deciding that a court sometimes may infer market power from a sufficiently clear demonstration that a firm believes that it possesses power. Even so, an argument about evil intent in antitrust requires us to ask: “intent to do what?” The Hospitals seem to think that intent to get the best price is a bad intent. They cite Mandeville Island Farms v. American Crystal Sugar Co., 334 U.S. 219, 68 S.Ct. 996, 92 L.Ed. 1328 (1948), to show that a monopsonistic depression of price is as bad as a monopolistic increase in price. True enough, see United States v. Capitol Service, Inc., 756 F.2d 502 (7th Cir.), cert. denied, — U.S. —, 106 S.Ct. 311, 88 L.Ed.2d 288 (1985), but Mandeville was a conspiracy to depress prices, and price-fixing cartels are unlawful independent of their efficacy. See Arizona v. Maricopa County Medical Society, 457 U.S. 332, 102 S.Ct. 2466, 73 L.Ed.2d 48 (1982). The Blues are a single firm, and the acts of single firms are judged by a different standard under § 2.
Competition is a ruthless process. A firm that reduces cost and expands sales injures rivals — sometimes fatally. The firm that slashes costs the most captures the greatest sales and inflicts the greatest injury. The deeper the injury to rivals, the greater the potential benefit. These injuries to rivals are byproducts of vigorous competition, and the antitrust laws are not balm for rivals’ wounds. The antitrust laws are for the benefit of competition, not competitors. Brunswick, supra, 429 U.S. at 488, 97 S.Ct. at 697, quoting from Brown Shoe Co. v. United States, 370 U.S. 294, 320, 82 S.Ct. 1502, 1521, 8 L.Ed.2d 510 (1962); Brunswick Corp. v. Riegel Textile Corp., 752 F.2d 261, 266 (7th Cir.1984), cert. denied, — U.S. —, 105 S.Ct. 3480, 87 L.Ed.2d 615 (1985); ECOS Electronics, supra. The antitrust laws protect efficient production for the benefit of consumers. Reiter v. Sonotone Corp., 442 U.S. 330, 342, 99 S.Ct. 2326, 2332, 60 L.Ed.2d 931 (1979); NCAA, supra, 104 S.Ct. at 2966 n. 49.
Sometimes injury to rival firms can be a presursor to injury to consumers; after knocking rivals out of the market, a firm may curtail output and raise price.' Section 2 may be used to prevent this conduct. Yet it must be used with the greatest caution. Action that injures rivals may ultimately injure consumers, but it is also perfectly consistent with competition, and to deter aggressive conduct is to deter competition. Thus the plaintiff faces a stiff burden in any § 2 litigation. “It is not enough that a single firm appears to ‘restrain trade’ unreasonably, for even a vigorous competitor may leave that impression. For instance, an efficient firm may capture unsatisfied customers from an inefficient rival, whose own ability to compete may suffer as a result. This is the rule of the marketplace and is precisely the sort of competition that promotes the consumer interests that the Sherman Act aims to foster. In part because it is sometimes difficult to distinguish robust competition from conduct with long-run anti-competitive effects, Congress authorized Sherman Act scrutiny of single firms only when they pose a danger of monopolization. Judging unilateral conduct in this manner reduces the risk that the antitrust laws will dampen the competitive zeal of a single aggressive entrepreneur.” Copperweld, supra, 104 S.Ct. at 2740.
So “intent to harm rivals” is not a useful standard in antitrust. See also Barry Wright Corp. v. ITT Grinnell Corp., 724 F.2d 227, 232 (1st Cir.1983): “ ‘[Ijntent to harm’ [rivals] without more offers too vague a standard in a world where executives may think no further than ‘Let’s get more business,’ and long-term effects on consumers depend in large measure on competitors’ responses.” Neither is “intent to do more business,” which amounts *1339to the same thing. Vigorous competitors intend to harm rivals, to do all the business if they can. To penalize this intent is to penalize competition. See also 7 Areeda, Antitrust, supra at ¶ 1506.
What of other “intents”? One intent reflected in the Blues’ documents is to buy medical care for less. Again, though, this is just another description for hard bargaining. Even a monopolist may bargain hard. See MCI Communications Corp. v. AT&T, 708 F.2d 1081 (7th Cir.), cert. denied, 464 U.S. 891, 104 S.Ct. 234, 78 L.Ed.2d 226 (1983), which holds that a dominant firm may slash prices to marginal cost in an effort to capture patronage. Other courts have held that monopolists may raise prices to customers, may charge what the traffic will bear, so long as they came by their market power lawfully. E.g., Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263, 296-98 (2d Cir.1979), cert. denied, 444 U.S. 1093, 100 S.Ct. 1061, 62 L.Ed.2d 783 (1980). Even the largest firms may engage in hard competition, knowing that this will enlarge their market shares. E.g., Foremost Pro Color, Inc. v. Eastman Kodak Co., 703 F.2d 534 (9th Cir.1983), cert. denied, 465 U.S. 1038, 104 S.Ct. 1315, 79 L.Ed.2d 712 (1984); Telex Corp. v. IBM Corp., 510 F.2d 894 (10th Cir.), cert. dismissed, 423 U.S. 802, 96 S.Ct. 8, 46 L.Ed.2d 244 (1975). Alcoa, which seemed to suggest the opposite, has been limited by the Second Circuit in Berkey, supra, 603 F.2d at 272-75, and does not now assist the Hospitals.
Bad intent in antitrust law must mean something other than the intent reflected in the Blues’ memos and the consultant’s report. We need not decide just what it means — if it means anything. The Supreme Court hinted in Aspen Skiing Co. v. Aspen Highlands Skiiing Corp., — U.S. —, 105 S.Ct. 2847, 2857-59, 86 L.Ed.2d 467 (1985), that intent “is merely relevant” to the question whether the large firm seeks to exclude competition on a “basis other than efficiency”. The focus must be on the objective basis, not the mental state. The First Circuit held in Grinnell that “intent” is a useless and confusing term best discarded in favor of analysis of objective indicators. See also Deauville Corp. v. Federated Department Stores, Inc., 756 F.2d 1183 (5th Cir.1985). The Second Circuit in Berkey reformulated intent to ask whether the large firm intended to “transfer” power from one market to another or to do something smaller firms could not do. Neither approach would assist the Hospitals — the Blues are not trying to get a monopoly in a second market, and the record establishes that “small” firms, including some of the plaintiffs, can offer PPO plans. More than a dozen PPO plans are operating or organizing in Indiana. Still another approach is to ask whether the defendant intends to (and can) raise its rivals’ costs of doing business. See Litton Systems, Inc. v. AT&T, 700 F.2d 785 (2d Cir.1983), cert. denied, 464 U.S. 1073, 104 S.Ct. 984, 79 L.Ed.2d 220 (1984). See also 3 Phillip Areeda & Donald F. Turner, Antitrust Law ¶ 626 (1978); Steven C. Salop & David T. Scheffman, Raising Rivals’ Costs, 73 Am.Econ.Rev. 267 (1983); Richard S. Markovits, The Limits to Simplifying Antitrust, 63 Tex.L.Rev. 41, 58-60 (1984). When a firm finds a way to confront its rivals with higher costs, it may raise its own prices to consumers without drawing increased output from them. See Aspen Skiing, supra. Exclusive dealing arrangements and boycotts sometimes may raise rivals’ costs. Cf. Northwest Wholesale Stationers, Inc. v. Pacific Stationery & Printing Co., — U.S. —, 105 S.Ct. 2613, 86 L.Ed.2d 202 (1985). But the Blues have not insisted that hospitals in the Blues’ PPO refrain from joining other PPOs, so rivals have access to hospitals on the same basis as the Blues. Some hospitals participate in more than three different PPO plans. Finding 34, 603 F.Supp. at 1683.
The Hospitals urge on us a version of an argument that the Blues’ PPO will raise rivals’ costs. The argument, which the parties call the “cost-shifting” argument, starts from the premise that hospitals raise just enough revenue to break even each year. The hospitals offer discounts to the Blues in order to participate in their PPO. *1340They therefore receive less revenue from patients covered by the PPO. To break even they must obtain more revenue from other patients. This - means they must “shift” to their other patients their costs of operation. When the hospitals raise their prices to other insurance plans (including the hospitals’ own PPO plans), these plans will be unable to compete with the Blues. The Blues’ PPO will have raised its rivals’ costs, in violation of § 2.
The district court did not address this argument explicitly. It apparently treated the cost-shifting argument as an attack on price discrimination, to which it correctly replied that discrimination is not forbidden. Finding 46, 603 F.Supp. at 1084. This does not dispose of a claim that PPOs raise rivals’ costs. Still, we can construct from the findings an answer sufficient for the time being. An argument that the Blues are shifting costs to rivals must start from the proposition that the Blues have market power. Why else would hospitals give the Blues a price break that injures rivals? The district court found that the Blues lack power, a finding we have held is not clearly erroneous. The cost-shifting argument also assumes that the price charged to the patients in the Blues’ PPO plan is below the appropriate measure of cost; the Hospitals contend that they must raise prices elsewhere to subsidize these patients. But the district court found that the PPO genuinely saves on the costs of care, and this too is not clearly erroneous. If the discounts given on patients covered by the PPO are justified by reductions in cost, there is nothing to “shift” to other patients.
We do not imply by this discussion that the Blues must demonstrate that the prices charged to the PPO patients are “cost-justified.” That has been a notorious quagmire in litigation under the RobinsonPatman Amendments to § 2 of the Clayton Act, 15 U.S.C. § 13. There is certainly no burden of justification in the absence of market power — no rational hospital sells below cost to a buyer without market power (the Hospitals do not say that they are themselves engaged in predatory pricing). It is also hard to see why, if the Hospitals can raise their prices to other buyers of their services, they do not do so whether or not they join the Blues’ PPO plan. We have held that large firms may drive hard bargains, and this does not imply that district courts should become little versions of the Office of Price Administration and assess the “cost-justification” for prices charged to these large customers. The Robinson-Patman amendments sometimes require this investigation, but they are limited to price discrimination affecting “commodities of like grade and quality” (15 U.S.C. § 13(a)). Medical services are not “commodities.” The Supreme Court has repeatedly stated that the control of price discrimination poses substantial risks to competition, which often works through “discriminatory” chiseling down of prices. E.g., Great Atlantic & Pacific Tea Co. v. FTC, 440 U.S. 69, 80-81, 99 S.Ct. 925, 933, 59 L.Ed.2d 153 (1979); United States v. United States Gypsum Co., 438 U.S. 422, 450-59, 98 S.Ct. 2864, 2880-84, 57 L.Ed.2d 854 (1978); Exxon Corp. v. Governor of Maryland, 437 U.S. 117, 133, 98 S.Ct. 2207, 2217, 57 L.Ed.2d 91 (1978). We have entered this bog only because here the district court did find that there were justifications for the lower price, and this is enough to demonstrate that it did not abuse its discretion in declining to issue a preliminary injunction against the PPO plan.
IV
The Blues say that once we have agreed with the district court that the Hospitals are unlikely to prevail, we should bring the litigation to an immediate conclusion. Antitrust litigation may be exceptionally expensive, and we could avoid further expense by directing the district court to grant judgment to the Blues. The Blues offer two arguments.
First, they say, the Hospitals had every opportunity to offer evidence at the hearing on their request for a preliminary injunction. They have nothing more to show, *1341and so there is no point in further proceedings. It may be that the district judge will come to this conclusion, but we do not direct him to do so.
The district court is best equipped to determine what additional evidence the Hospitals have to offer and how it might affect the findings of fact. The Hospitals may have an uphill battle to show the district judge evidence that calls his findings into question, but we do not think that we should prohibit the district judge from allowing the Hospitals that opportunity. Evidence on cost-shifting, for example, was poorly developed. Did hospitals that joined the Blues’ PPO raise prices to their non-PPO patients (by more than hospitals outside the Blues’ PPO did)? Have the hospitals that joined the Blues’ PPO lost patients as a result? Did anyone else? Did other insurers have to raise their prices (which cost-shifting implies)? Questions of this sort, which now lack answers, will be important if the district judge should be converted by further evidence to the view that the Blues have market power.
Second, the Blues argue that their PPO is protected by the “state action” doctrine of Parker v. Brown, 317 U.S. 341, 63 S.Ct. 307, 87 L.Ed. 315 (1943). That doctrine excludes from the antitrust laws closely supervised private conduct that implements clearly articulated state policies. See Southern Motor Carriers Rate Conference, Inc. v. United States, — U.S. —, 105 S.Ct. 1721, 1727, 1730, 85 L.Ed.2d 36 (1985). See also Cantor v. Detroit Edison Co., 428 U.S. 579, 96 S.Ct. 3110, 49 L.Ed.2d 1141 (1976). Indiana has a policy favoring PPOs, the Blues say, which meets this standard. This appears to say that (a) Indiana favors PPOs over other methods of organization, (b) the Blues’ PPO is the favored kind, and (c) the state supervises the PPO contracts with hospitals as well as the insurance contracts with the purchasers of service. The district court did not discuss this contention, and it did not make the findings about the nature of Indiana’s law and the Blues’ PPO that would be necessary to resolve it. We therefore do not consider this argument further. The district judge should do so if he deems resolution of the question necessary.
V
Although more proceedings may be in store on the Hospitals’ antitrust arguments, the district court entered a final judgment rejecting their arguments under state law. The principal law in question, Ind.Code § 27-8-11-3, became effective on December 31, 1984. Subsection (a) allows insurers to establish PPO plans. The other pertinent provisions state:
(b) Before entering into any agreement ... an insurer shall establish terms and conditions that must be met by providers wishing to enter into an agreement with the insurer____ These terms and conditions may not discriminate unreasonably against or among providers____ [Njeither differences in prices among hospitals or other institutional providers produced by a process of individual negotiation nor price differences among other providers in different geographical areas or different specialties constitutes unreasonable discrimination.
(c) No hospital, physician, pharmacist, or other provider ... willing to meet the terms and conditions offered to it by an insurer may be denied the right to enter into an agreement____
The Hospitals’ strongest argument is that the Blues have discriminated unreasonably among providers, contrary to the prohibition of subsection (b). Before assessing the legal contentions, however, we review the procedures by which the parties negotiated the PPO contracts.
The Blues announced the terms of their plan and invited all 115 general acute care hospitals to submit bids, expressed as percentages of their existing prices. Findings 26 and 27, 603 F.Supp. at 1082-83. In response to comments from the hospitals, the Blues revised some of the terms of the proffered contracts and asked for renewed bids. Findings 35 and 36, 603 F.Supp. at 1083-84. The terms included utilization controls and other changes designed to re*1342duce the cost of service. There followed negotiating sessions with each of the 91 hospitals that submitted a bid. Although the district court did not make findings on this question, the parties agree that the Blues gave each hospital little if any information about the bids made by other hospitals in the area and suggested that each hospital bid a greater discount. The Blues never proposed a particular discount or said what discount would be sufficient; they simply waited for hospitals to bid less.
From among the hospitals that got this far, the Blues selected those that bid the lowest prices and were “conveniently located” for the Blues’ insureds. Finding 30, 603 F.Supp. at 1083. Although the decision to exclude a hospital was almost always based on price, in two cases geography played a role. Again this is a subject on. which the district court was silent but the parties largely agree. The Blues excluded Winona Memorial Hospital, despite its acceptable price, because another hospital in the same city was better located. The Blues excluded St. Joseph’s Hospital of Ft. Wayne for two reasons — they deemed its bid of 80% of prior prices a “low-ball” that was sure to be increased, and they concluded that it was not as conveniently located as Parkview Hospital in the same city.
The Blues defend this geographic selectivity by relying on the provision in § 3(b) stating that “price differences among other providers in different geographical areas or different specialties” do not constitute “unreasonable discrimination.” The statute, say the Blues, “is not a model of drafting clarity” but is “plainly [designed] to permit the insurer to select its PPO hospitals based, not only on price, but also on the geographic accessibility needs of its insureds.” This is not persuasive.
The clause in question states an exception to the rule of § 3(b) that plans not “unreasonably discriminate” among hospitals. The exception covers “differences in prices among hospitals or other institutional providers produced by a process of individual negotiation” and “price differences among other providers in different geographical areas or different specialties”. Read as a whole, this means that specified differences in price are not “unreasonable discrimination.” The specified differences are: for hospitals, differences individually negotiated; for other providers (such as physicians), differences by location or specialty. There is no linguistic support for reading this clause to allow geographic price schedules for “hospitals or other institutional providers”, let alone to allow geographic distinctions unrelated to price. Perhaps the legislative history of the statute might supply such a reason, but Indiana does not publish the legislative history of its laws, and the parties have not provided us with any other access to that history.
The reading we have given the language makes more than linguistic sense. Hospitals do so much business that it is (relatively) easy to negotiate individually with them about price. There are thousands of other providers, however, and individual price negotiations might be prohibitively difficult. The statute therefore authorizes the insurer to promulgate price schedules and allows these to make distinctions on the basis of geography (medical care may be less costly in rural areas than in cities) and specialty (a specialist in a given procedure may charge more than a general practitioner, reflecting the specialist’s greater expertise and the lower risk of error). The language of the statute thus serves useful functions without authorizing non-price discrimination among hospitals on the basis of geography.
Still, this does not carry the day for the Hospitals. That the Blues’ reliance on the price proviso is unwarranted simply returns us to the general principle: A PPO plan’s terms must not “unreasonably discriminate” among hospitals. The Hospitals argue as if all distinctions on the basis of geography are “unreasonable,” but they do not say why. In some cases the savings in cost may depend on the ability to direct patients to a particular hospital, the better to use its facilities fully. This may require reducing the number of hospitals in the *1343PPO plan in a given city. The Hospitals reply by invoking § 3(c), which says that no provider “willing to meet the terms and conditions offered to it ... may be denied the right to enter into an agreement”. These “terms,” the Hospitals say, are the ones written in the contract, not the criteria the Blues use to choose among bidding hospitals, and therefore § 3(c) forbids geographic distinctions. The Blues rejoin that § 3(c) cannot be given so broad a reading, for then the statute would also exclude all distinctions on the basis of price, which the legislature plainly has authorized.
We need not decide whose reading of § 3(c) is correct. The Hospitals were allowed to put on all the evidence they desired, and they showed only two instances of the use of geography. We need not determine whether § 3 allows every use of geography; it is enough to determine whether the Hospitals have demonstrated a legal flaw in these two uses. They have not. Winona Memorial Hospital is not a plaintiff. The plaintiff Hospitals may not urge a wrong done to a stranger to the litigation as a ground of relief for themselves. Secretary of State of Maryland v. Joseph H. Munson Co., 467 U.S. 947, 104 S.Ct. 2839, 2846-48, 81 L.Ed.2d 786 (1984). St. Joseph’s Hospital, a plaintiff, was excluded for two reasons: price and location. It has not established that but for the consideration of location it would have been given a PPO contract. The district court did not make findings of fact on the question, but it does not appear to be a seriously disputed issue. The Hospitals insist that the Blues are not entitled to refer to price in deciding which hospitals to include in the program, but if the Blues may use price at all they may determine which prices quoted to them are bona fide.
The Hospitals do not disagree with the Blues’ contention that they determined St. Joseph’s bid to be a low-ball quote, too low to be justified by its costs (on which the Blues had data) and therefore too low to be sustained. One witness testified without contradiction that St. Joseph’s bid was well below that of any other hospital, and another testified that the Blues feared that “at the first opportune time [St. Joseph’s] would be asking for an unreasonably high increase”. Unless a court must evaluate the propriety of every pricing decision made by an insurance company — and there is no support for that in § 3 — this testimony is enough to show that location was not dispositive in St. Joseph’s case. Consequently, there is support in the record for the district court’s conclusion that the Hospitals have not established a violation of § 3.
This conclusion holds, however, only if price is a legitimate ground for excluding a hospital from a PPO. The Hospitals say that it is not, relying on § 3(c). On their reading of the statute, only a Hospital’s failure to meet the terms and conditions of the PPO plan as a whole may be a reason for exclusion. They submit that price may be a “term” only if the same price is offered to all hospitals by contract. This cannot be right. It would nullify the explicit authorization in § 3(b) for the use of individually-negotiated prices. More, it would make hash of the idea of a PPO. A “preferred” provider is one to which the insurer seeks to funnel business, partly in the search for operating efficiencies, partly in the search for lower prices. If high-price bidders had to be designated as “preferred” providers, it would be difficult to contain costs. An insurer could not cut out of its system the high-price providers, and there would be no reason for hospitals to bid against one another for inclusion. The requirement that hospitals bid in order to get the plan’s business is the principal source of the incentive to cut costs, and the prospect of getting more business by becoming a “preferred” provider is the principal benefit to a hospital. We do not think that Indiana took back in § 3(c) the authorization to establish PPO plans granted by §§ 3(a) and (b).
The Hospitals' last argument under § 3 is that the Blues may not take advantage of the price proviso of § 3(b) because they did not engage in “individual negotiation”, as the statute requires. Offi*1344ciáis of the Blues met separately with officials of each hospital, but the hospitals did all the talking about price. The Hospitals say that “negotiation” means “discussion whereby parties mutually interested seek to resolve differences with the purpose of arriving at agreement.” Barrick Realty Co. v. Bogan, 422 N.E.2d 1306, 1308 (Ind.App.1981). A one-sided meeting in which the hospital talks price and the Blues listen is not negotiation under this definition, the Hospitals say. They insist that the Blues had to propose acceptable prices or at least reveal what other hospitals, were bidding, so that each could see the competition.
The problem with this position is that few bargainers reveal their reservation prices. If the Blues had said: “A discount of 10% is enough to ensure your participation in the program,” few hospitals would have offered discounts of 12%. If the Blues had said: “Your competition has bid discounts of 8% and 7%,” few hospitals would have offered discounts steeper than 9%. See United States v. United States Gypsum Co., supra, in which the Supreme Court concluded that the exchange of price information among rivals would tend to stabilize or increase prices. It is unlikely that Indiana meant to require a form of disclosure that would have this effect. Often the most effective negotiation, from the buyer’s perspective, is to sit back and say: “Do more,” without saying how much more. The seller’s uncertainty induces it to bid down to marginal cost. This is hard bargaining, and it is negotiation notwithstanding its tendency to drive down prices. Competition has that effect.
A rule requiring a buyer to tell the seller how low is low enough would reduce the power of competition. See Great Atlantic & Pacific Tea Co. v. FTC, supra, 440 U.S. at 80-81, 99 S.Ct. at 933. In A&P the Supreme Court held that a buyer may obtain a discriminatorily low price from a seller without informing the seller that a higher price would have fetched the business anyway. It held that a requirement to tell the seller how low is low enough would be antithetical to the purposes of the antitrust laws. The price one will accept is usually a negotiator’s closely guarded secret. See Federal Open Market Committee v. Merrill, 443 U.S. 340, 361-64, 99 S.Ct. 2800, 2812-14, 61 L.Ed.2d 587 (1978). Perhaps an insurer could satisfy the Hospitals’ demand for price quotes by naming an unreasonable price, hoping to nudge each hospital closer to the one it really had in mind, but the statute does not require the observance of useless forms. Cf. Trans Alaska Pipeline Rate Cases, 436 U.S. 631, 653, 98 S.Ct. 2053, 2065, 56 L.Ed.2d 591 (1978).
“Individual negotiation” therefore means a process of dickering one hospital at a time, as opposed to announcing a price schedule for groups of hospitals. It contrasts with the geographic price schedules § 3(b) allows for “other providers” and specialists. The statute does not require the courts to supervise the handling of each individual negotiation. This conclusion disposes of the Hospitals’ arguments under § 3. The Blues did not violate the statute by conducting one-sided negotiations or by taking price into account.
Although we reject the Hospitals’ claims of discrimination based on price and location, our decision does not rest on any final construction of the statute; we have held only that no plaintiff hospital established that locational considerations were the cause of its exclusion from the Blues’ PPO plan. Because we have been able to decide this case without finally settling on a construction of § 3, we deny the Hospitals’ motion that we certify this question to the Supreme Court of Indiana.
VI
The Hospitals rely on a second statute, Ind.Code § 34-4-12.6-2(a). This part of the Peer Review Act provides that all proceedings of physicians reviewing the work of other physicians shall be confidential. The PPO contracts, the Hospitals tell us, violate this statute because they grant the Blues access to each hospital’s records and statistical information “without restriction.” The district court gave two replies: *1345first that the PPO Act preempts conflicting laws, see Ind.Code § 27-8-11-2; second that the PPO agreement “does not require preferred hospitals to breach any peer confidence under” § 12.6-2(a). Conclusions 26 and 28, 603 F.Supp. at 1087. The second of these grounds is all we need consider. The Blues distributed to each hospital material in question-and-answer form explaining the PPO contracts. Answer 89 states: “In no case would we [the Blues] be entitled by the contract language to access which you cannot legitimately provide because of statutory requirements placed upon you.” The contract, as the parties construe it, therefore does not violate § 12.6-2(a). There will be time for further consideration of preemption if the Blues should retract this assurance and seek access to information covered by the Peer Review Act.
VII
The district court not only construed the Indiana statutes but also held that the PPO contracts do not breach the provider agreements the Blues have with hospitals participating in their service benefit insurance plan. The Hospitals say that the declaration that there has been no breach must be set aside because the district court heard no evidence on the issue. Yet the Hospitals do not inform us what evidence the district court needed to hear, and they say that “as of the close of the evidence no breach had yet occurred.”
The Hospitals do not refer us to any language in their provider contracts that is inconsistent with any obligations in the PPO contracts. The Blues point out that hospitals who have joined the PPO plan have agreed to the terms of the contract, and hospitals who have not joined simply proceed under the terms of the old contracts without modification. If a patient covered by PPO insurance goes to a hospital that is not a member of the PPO, the Blues reimburse 75% of the hospital’s charges. The patient must pay the remainder. The Blues assert, without contradiction from the Hospitals, that their basic contracts provide for the care of patients with copayments of this character.
The very existence of the PPO therefore cannot be deemed an anticipatory breach of the contracts of the hospitals outside the PPO. Perhaps some cases will arise in which obligations the Blues have assumed under the PPO will produce violations of the existing non-PPO contracts. We do not interpret the district court’s judgment as ruling out this prospect. All the district court was entitled to hold — all it did hold— is that the PPO plan is not an inevitable violation of existing contracts. More specific claims of breach may be raised in cases presenting them.
VIII
One procedural issue requires our attention. During discovery the Hospitals requested access to the Blues’ data on prices bid by each hospital and the calculations the Blues performed to decide which hospitals to include in the PPO. The Blues replied that the data are trade secrets, that they had pledged confidentiality to the hospitals, and that they feared the Hospitals could use the comparative price information to raise their prices or collude in future years. See United States v. United States Gypsum Co., supra. The district court agreed that the data are sensitive and offered the Hospitals access under a protective order restricting access “only to trial counsel to this lawsuit who are engaged in the preparation for trial ... and who have neither represented nor [will] represent for 18 months any hospital, the Indiana Hospital Association, or any other entity in connection with the Blue Cross PPO on any matter other than the trial of the case.”
Counsel for the Hospitals refused to take the information with these strings attached. Witnesses for the Blues testified, based on portions of these data, about how the Blues decided which hospitals to include in the PPO. The Hospitals now say that their lack of access to the data prevented them from examining these witnesses or trying the case effectively, and they *1346ask for a new trial at which they will be able to use the data.
The price data are unquestionably relevant, and parties are entitled to information as important as this was. Deitchman v. E.R. Squibb & Sons, Inc., 740 F.2d 556 (7th Cir.1984). The price data are also unquestionably sensitive trade secrets of the Blues. Hospitals armed with the data could use it to advantage in the next round of negotiations. Access to the data could turn an antitrust suit into the basis of effective collusion, a concern we have expressed above. See also General Leaseways, Inc. v. National Truck Leasing Ass’n, 744 F.2d 588, 597 (7th Cir.1984). Confidential information is customarily made available, if at all, under a protective order, and the district court has substantial discretion to decide which information should be protected and to frame the order. Seattle Times Co. v. Rhinehart, 467 U.S. 20, 104 S.Ct. 2199, 2206-09, 81 L.Ed.2d 17 (1984); E.I. du Pont de Nemours Powder Co. v. Masland, 244 U.S. 100, 103, 37 S.Ct. 575, 576, 61 L.Ed. 1016 (1917).2
Counsel for the Hospitals balked at this protective order, they said, because it “restricts us [as] lawyers undertaking representations in future matters.”3 But the district court had reason to think the restriction necessary. Counsel for the Hospitals regularly represented them in price negotiations with the Blues. When counsel act as the negotiators, they become business agents of the Hospitals, and there is little difference between providing information to the president of a hospital and providing it to the hospital’s lawyer-agent.
The protective order did not prevent the Hospitals from retaining a separate law firm for the purpose of inspecting and using the data. It simply prevented the particular lawyers who had seen the data from working for a hospital as business agents during the next 18 months— the ensuing contract year and negotiation season. The Hospitals are represented by more than one firm, and they could have chosen from them (or still other firms) lawyers to examine and use the data while leaving the Hospitals’ regular firms free to serve in dual legal and business roles as before. The district court therefore did not abuse its discretion in framing the order.
Affirmed.
. People who cannot divert to preferred providers are unlikely customers of PPO plans. When a person's physician has privileges only at non-PPO hospitals, the incentives created by the plan have less effect, if they have any. Because the PPO plan adds a new option and does not withdraw any old one, patients who lack a choice among hospitals are best off sticking with their existing financing scheme. Throughout this opinion, when we discuss patients' incentives we refer only to those patients who have options or can persuade physicians to take advantage of these options.
. The data were not introduced into evidence, and therefore the presumption that evidentiary matters will be available to the public does not apply. See In re Continental Illinois Securities Litigation, 732 F.2d 1302 (7th Cir.1984) (announcing a presumption of availability). The D.C. Circuit has concluded that the panel in Continental Illinois got its history wrong, see In re Reporters Committee for Freedom of the Press, 773 F.2d 1325 (D.C.Cir.1985) (no presumption of access), but this is not the occasion to resolve the question.
. Counsel cited Disciplinary Rule 2-107, which says that counsel shall "not enter into an agreement that restricts his right to practice law” as part of a "settlement of a controversy or suit”. The protective order was not part of a "settlement," and we therefore need not consider whether it "restricts [the] right to practice law" within the meaning of the rule.