POSNER, Circuit Judge.
This appeal from a judgment of criminal contempt brings up to us issues of res judicata, pretrial discovery under Rules 26(c) and (d) of the Federal Rules of Civil Procedure, and the application of the Sherman Act’s “Rule of Reason” to denial of membership in a professional association.
The defendant-appellant in this case is the American Academy of Orthopaedic Surgeons, a private association to which most orthopaedic surgeons in the United States belong. The plaintiffs are two orthopaedic surgeons who practice in Evansville, Indiana, and Chicago, respectively, and were denied membership in the Academy without a hearing or a statement of reasons. Membership is alleged to confer certain professional advantages, but it is not a prerequisite either to being certified to practice as an orthopaedic surgeon or to obtaining hospital staff privileges. Each of the plaintiffs is certified to practice orthopaedic surgery and has staff privileges at several hospitals.
[1087]*1087After being refused membership the plaintiffs brought suit against the Academy in an Illinois state court. They claimed a right under Illinois law to a hearing on their application and to reasonable standards for membership. The Illinois Appellate Court held that the complaint failed to state a claim because membership in the Academy is not an “economic necessity.” Treister v. American Academy of Orthopaedic Surgeons, 78 Ill.App.3d 746, 755-56, 33 Ill.Dec. 501, 508, 396 N.E.2d 1225, 1232 (1979). Having lost in state court the plaintiffs then sued the Academy in federal district court for injunctive relief and damages under section 1 of the Sherman Act, 15 U.S.C. § 1, charging an illegal boycott. The complaint alleged that the Academy is “a monopoly in its field, possessed of substantial power to control the market for orthopaedic surgical services,” and that the plaintiffs, though fully qualified for membership under the announced criteria of the Academy, were excluded for “extraneous” reasons, which in the case of Dr. Treister (no particulars were given for Dr. Marrese) included “(a) his supposed willingness to offer expert testimony against other orthopaedic surgeons in medical malpractice cases; (b) his known willingness to consult surgical out-patients on a relatively high-volume basis; and (c) his nonconformity of personality and personal attitudes with those of most established orthopaedic surgeons and in particular those who were already members of the academy.” The complaint alleged that the Academy had refused to explain its actions and that the effect of those actions was “to limit competition and enforce conformity with current business practices” and to injure the plaintiffs in the practice of their profession.
The Academy moved to dismiss the complaint on two grounds: the judgment dismissing the plaintiffs’ state court action was res judicata in the federal proceeding; the complaint failed to state a claim under the Sherman Act. The motion was denied, 496 F.Supp. 236 (N.D.Ill.1980), and discovery began. The. plaintiffs asked the Academy to produce all of its correspondence and other documents relating both to the denial of the plaintiffs’ application for membership and to all other denials of membership applications between 1970 and 1980. The Academy refused. It persisted in its refusal even after the district court issued a protective order limiting access to the discovered materials to the plaintiffs and their counsel, and even after the court ordered the Academy to produce the requested documents. The court held the Academy in criminal contempt of its order and fined it $10,000. The Academy appeals from that judgment.
The Academy asks us to hold that the discovery order was an abuse of the district court’s discretion under Rule 26(c) of the Federal Rules of Civil Procedure., The plaintiffs point out that it is not a final order and argue that we cannot review it because the district court has not certified it for an immediate appeal under 28 U.S.C. § 1292(b). But the contempt judgment is a final order, re viewable by us; and a party who is willing to pay the price of being punished for contempt (or suffering an equivalent sanction such as dismissal of the complaint) if the validity of the order he has disobeyed is ultimately upheld may by that means get immediate review of the order. Many cases hold this. See, e.g., United States v. Ryan, 402 U.S. 530, 532-33, 91 S.Ct. 1580, 1581-82, 29 L.Ed.2d 85 (1971); Ryan v. Commissioner of Internal Revenue, 517 F.2d 13, 19-20 (7th Cir.1975); Hanley v. McHugh Constr. Co., 419 F.2d 955, 957 (7th Cir.1969); National Util. Serv., Inc. v. Northwestern Steel & Wire Serv., Inc., 426 F.2d 222 (7th Cir.1970); Hastings v. North East Independent School Dist, 615 F.2d 628, 631 (5th Cir.1980). But these cases coexist uneasily with another line of cases which holds that where as in this case the judgment is for criminal rather than civil contempt, the validity of the underlying order may not be questioned on appeal from the contempt judgment. See, e.g., United States v. United Mine Workers of America, 330 U.S. 258, 291-94, 67 S.Ct. 677, 694-96, 91 L.Ed. 884 (1947); ITT Community Development Corp. v. Barton, 569 F.2d 1351, 1356 (5th Cir.1978). Yet our decision in Hanley, for example, involved criminal contempt.
[1088]*1088The two lines of cases can be reconciled, see Hanley, supra, 419 F.2d at 957; United States v. Ryan, supra, 402 U.S. at 532 n. 4, 91 S.Ct. at 1582 n. 4; 13 Wright, Miller & Cooper, Federal Practice and Procedure § 3537, at pp. 340-41 (1975), by noting that in cases where the validity of the underlying order has been held not to be reviewable on appeal from the judgment of contempt, the order could have been appealed directly, while discovery orders cannot be, other than through discretionary or extraordinary procedures, such as section 1292(b) appeals and mandamus proceedings. The right to have a discovery order reviewed on appeal from a contempt judgment for disobeying it is thus one of the safety valves in the final-judgment rule, 28 U.S.C. § 1291. A discovery order may impose heavy and irrecoverable costs on a party; yet to make discovery orders appeal-able as of right would lead to intolerable delays in federal litigation. Confining the right to appellate review of discovery orders to cases where the party against whom the order was directed cared enough to incur a sanction for contempt is a crude but serviceable method of identifying really burdensome discovery orders, and waiving the finality rule only for them. The validity of the discovery order that the Academy disobeyed is therefore properly before us.
Rule 26(c) of the Federal Rules of Civil Procedure empowers the district court to “make any order which justice requires to protect a party or person from annoyance, embarrassment, oppression, or undue burden or expense, including .. . that discovery not be had .... ” The effective management of complex litigation requires that the district judge be allowed a broad discretion in guiding the discovery process, and therefore in exercising his powers under Rule 26(c). Chrysler Corp. v. Fedders Corp., 643 F.2d 1229, 1240 (6th Cir.1981). But his discretion is not unlimited and if he abuses it he commits reversible error. Silkwood v. Kerr-McGee Corp., 563 F.2d 433, 436 (10th Cir.1977).
In ruling on a motion under Rule 26(c) to limit discovery, the district court must compare the hardship to the party against whom discovery is sought, if discovery is allowed, with the hardship to the party seeking discovery if discovery is denied. Not only the magnitude of the hardship, but its nature, must be considered. Purely private interests should be given less weight than interests that are considered to have a special social value.
The Academy argues that the First Amendment immunizes from discovery its correspondence and other documents relating to denials of applications for membership. This is a way of saying that the interests it is asserting have a more than purely private value. But the Academy goes too far, as this court’s recent decision in Memorial Hospital for McHenry Cty. v. Shadur, 664 F.2d 1058, 1063 (7th Cir.1981) (per curiam), which rejected a claim of privilege for a hospital’s records of disciplinary proceedings against staff physicians, makes clear. Even if the Academy were engaged in advocating controversial views, and the publication of its internal files would expose members to retaliation for those views, it would not have an absolute privilege against discovery, though the plaintiffs would then have a heavy burden of showing that the information sought was vital to their case and unobtainable by means less likely to discourage such advocacy. See Hastings v. North East Independent School Dist., supra, 615 F.2d at 632; In re Petroleum Prods. Antitrust Litigation, 680 F.2d 5, 7 (2d Cir.1982).
Yet we cannot say that there is not lurking in this case if not a First Amendment right at least a First Amendment interest, which the discovery sought by the plaintiffs would impair and which differentiates this case from the usual antitrust case where discovery is sought of invoices or salesmen’s reports or the minutes of a corporation’s board of directors. The Academy may be engaged in a conspiracy to violate the antitrust laws, but it is also engaged in an exchange of information about surgical techniques and related matters of great public interest. This exchange may be inhibited if the Academy has to [1089]*1089disclose its membership files, even under a protective order; a protective order is the kind of lawyers’ arrangement that laymen instinctively distrust. If the Academy complies with the discovery order its members may be reluctant to offer candid evaluations of applicants in the future, and the atmosphere of mutual confidence that encourages a free exchange of ideas may be eroded.
It is also necessary to consider the probable impact of the discovery order, if it is upheld, on the purely private interests of the Academy and its members. One does not have to raise the ghosts of Aristotle and de Tocqueville to be reminded that voluntary associations are important to many people, Americans in particular, and that voluntary professional associations are important to American professionals (the last proposition is the very premise of the plaintiffs’ antitrust suit, as it was of their Illinois suit). Confidentiality in deliberations on membership is essential to the voluntary character of an association. Without it there are no candid evaluations of applicants, and without such evaluations the character of the association as a group of people associating by free choice is compromised. Apart, therefore, from any First Amendment considerations, which we have said are only indirectly implicated by the discovery order, the involuntary disclosure of the membership files of a voluntary association is inimical to the private (but worthy) objectives of the association, even when a protective order is issued. This is not to say that such disclosure may not be ordered; but it involves a significant hardship, which must in every case be balanced against its benefits.
On the benefit side of the ledger in a Rule 26(c) analysis the emphasis usually is on whether there are other, less burdensome means of getting the information that the party seeking discovery needs to prove his ease. But another consideration, less frequently emphasized but no less legitimate, is whether his case has any possible merit. If it does not, the only “value” of the requested discovery will be to give the party seeking it leverage to extort a settlement, by enabling him to impose on his opponent costs that may induce a settlement offer despite the lack of merit of the case. Predatory discovery is a serious problem in federal litigation today, and Rule 26(c) is one of the district court’s too' little used weapons against it. This is not to say that a Rule 26(c) movant may demand a trial of the merits on the spot. But suppose that after the pleadings, and the evidence obtained from all the other discovery in the case, are interpreted in the most favorable light to the party seeking the discovery order, and after every assumption favorable to that party with respect to what the requested order might turn up by way of additional evidence is indulged, it seems highly unlikely that the party’s claim or defense, as the case may be, has any possible merit. A motion for summary judgment might be premature because discovery was not yet complete. But if the discovery requested would be burdensome, the probable lack of merit of the underlying claim or defense is not irrelevant to the district court’s exercise of its discretion under Rule 26(c).
This point is not new. “The court has discretion whether or not to order discovery, however, and it would hardly be an abuse of discretion to refuse discovery when the claim or defense appears baseless and discovery would work a hardship on the other party.” 4 Moore’s Federal Practice ¶ 26.56[1], at pp. 26-124 to 26-125, n. 20 (Moore & Lucas eds. 1982). See Doe v. New York City Dept, of Social Services, 649 F.2d 134, 147 (2d Cir,1981); Spier v. Home Ins. Co., 404 F.2d 896, 899-900 (7th Cir.1968); Dionne v. Springfield School Comm., 340 F.Supp. 334, 335 (D.Mass.1972) (Wyzanski, J.) (“Federal courts do not stage academic tournaments merely for Don Quixotes to practice knighthood”). In a clear case, refusal to deny a discovery request on this ground (burdensome request in support of a probably baseless claim) could be an abuse of discretion; we must decide whether this is such a case.
[1090]*1090The Academy has advanced several grounds for believing that this antitrust suit has no conceivable merit, whatever the correspondence and membership files that the plaintiffs are trying to obtain may show about its motives in refusing to admit the plaintiffs and other applicants to membership. One is that the suit is barred by res judicata because of the earlier judgment against the plaintiffs in their state court action against the Academy. True, the plaintiffs did not include an antitrust count in that action. But res judicata bars not only the claims that a party actually made in the first suit but also the claims he could have made on the basis' of the facts that he alleged in that suit. A plaintiff may not split his cause of action by bringing a suit on one theory and then, after he has lost, trying again on a different theory. Federated Department Stores v. Moitie, 452 U.S. 394, 101 S.Ct. 2424, 69 L.Ed.2d 103 (1981); Harper Plastics, Inc. v. Amoco Chemicals Corp., 657 F.2d 939, 945 (7th Cir.1981). “Cause of action” in this context is simply the set of facts that establishes a right to relief under any theory that the plaintiff could have asserted in his first suit. Here that set of facts comprises the circumstances and consequences of the Academy’s refusal to admit Drs. Marrese and Treister to membership, and it is irrelevant that the plaintiffs chose not to allege the antitrust implications of those facts in their state court action.
But the plaintiffs point out that they could not have joined their Sherman Act claim to the other claims that they were prosecuting in state court. Blumenstock Bros. Adv. Agency v. Curtis Pub. Co., 252 U.S. 436, 440-41, 40 S.Ct. 385, 386-87, 64 L.Ed. 649 (1920), held that the federal courts have exclusive jurisdiction over suits alleging violations of the federal antitrust laws, and the holding is repeated in many modern cases, such as Kurek v. Pleasure Driveway & Park Dist. of Peoria, Ill., 583 F.2d 378, 379 (7th Cir.1978) (per curiam), in this circuit. This proposition has been questioned. See Note, Exclusive Jurisdiction of the Federal Courts in Private Civil Actions, 70 Harv.L.Rev. 509, 510 n. 13 (1957). It has no basis in the language or legislative history of the federal antitrust laws; and the fact that, as we are about to see, state courts adjudicate defenses based on federal antitrust law, with preclusive effect under the doctrine of collateral estoppel in any subsequent federal antitrust suit, shows that state courts are competent tribunals to adjudicate federal antitrust questions — and why should this not be so in cases brought, as well as defended, under federal antitrust law?
But however this question is answered, it cannot be doubted that the plaintiffs could have joined a claim under the Illinois Antitrust Act, Ill.Rev.Stat.1981, ch. 38, § 60, in their state court suit; and we must consider how close a substitute it would have been for their Sherman Act claim. The language of the Illinois act is not identical to that of the Sherman Act. Section 1 of the Sherman Act forbids conspiracies and other agreements to restrain trade. Section 3(2) of the Illinois Antitrust Act, which, with an exception noted later in this opinion, is the section under which a boycott would be challenged, see Bar Committee Comments— 1967, Smith-Hurd’s Ill.Ann.Stat. ch. 38, at p. 452 (1977), forbids conspiracies and other agreements to restrain trade “unreasonably.” Ill.Rev.Stat.1981, ch. 38, § 60-3(2). The difference in language has persuaded the Illinois Appellate Court that a boycott challenged under section 3(2) could never be pronounced illegal per se; it would have to be evaluated under the Rule of Reason. Blake v. H.F. Group Multiple Listing Serv., 36 Ill.App.3d 730, 743, 345 N.E.2d 18, 28 (1976). But that is of no moment in the present case. We shall see that the only tenable theory of an illegal boycott that these plaintiffs have is a Rule of Reason theory, not a per se theory; and so far as boycotts that are unlawful only if unreasonable are concerned, we can find no difference in substantive standards between the Illinois and federal acts. See Blake, supra, 36 Ill.App.3d at 739, 743, 345 N.E.2d at 25, 28; cf. People v. Crawford Distrib. Co., 53 Ill.2d 332, 339, 291 N.E.2d 648, 652 (1972), interpreting Ill.Rev.Stat.1981, eh. 38, § 60-11.
[1091]*1091But there is a difference in the remedies under the two acts. A prevailing plaintiff in a boycott ease under the Sherman Act as in any other federal antitrust case is entitled to an automatic trebling of his damages. See 15 U.S.C. § 15. But section 3(2) of the Illinois act is subject to the following provision: “if it is shown that [the] violation was willful, the court may, in its discretion, increase the amount recovered as damages up to a total of 3 times the amount of actual damages.” Ill.Rev.Stat.1981, ch. 38, § 60-7(2).
If it were not for this difference Nash County Bd. of Educ. v. Biltmore Co., 640 F.2d 484, 487-93 (4th Cir.1981), would be a direct precedent for holding that the plaintiffs’ federal antitrust claim is barred by res judicata. The Fourth Circuit thought it intolerable that a plaintiff should be able to bring an antitrust suit in state court under a state statute identical to the Sherman Act and then, after final judgment in that suit, bring the same suit in federal court under the Sherman Act. And, as we have said, it would have made no difference in Nash if the first suit had been brought under another theory and the plaintiff had reserved his antitrust theory for later use in federal court in the event that he failed to win in state court on the other theory.
Nash is a frankly innovative decision; it rejects contrary authority in three circuits. See Abramson v. Pennwood Inv. Corp., 392 F.2d 759, 762 (2d Cir.1968); Clark v. Watchie, 513 F.2d 994, 997 (9th Cir.1975); Hayes v. Solomon, 597 F.2d 958, 984 (5th Cir.1979). (Our circuit has never spoken to the question, although the brief per curiam opinion in Kurek v. Pleasure Driveway & Park Disk of Peoria, Ill., supra, could be read to imply, though it does not expressly adopt, an approach inconsistent with Nash.) But these cases predate Nash, and the courts that decided them may want to reexamine their position in' the light of the full and convincing discussion of the issue in Nash; the academic authorities that the court marshaled in support of its position, e.g., Currie, Res Judicata: The Neglected Defense, 45 U.Chi.L.Rev. 317, 347-48 (1978); the enigmatic but seemingly dispositive Supreme Court precedent on which the court relied, Becher v. Contorne Labs., Inc., 279 U.S. 388, 391-92, 49 S.Ct. 356, 357, 73 L.Ed. 752 (1929) (per Holmes, J.); and the preclusive effect of state court factfindings in subsequent federal antitrust suits.
The last point will bear some elaboration. The best reason for not giving preclusive effect to a judgment in an earlier state court action would be a belief that state courts were not competent to adjudicate federal antitrust questions, even if the state had enacted a statute, enforceable of course in its own courts, tracking the federal antitrust statutes. But this belief cannot be maintained given that state courts are allowed to decide federal antitrust questions, with preclusive effect under the doctrine of collateral estoppel, in suits in which a federal antitrust claim'is raised by way of defense. At least this is clear where the question is factual. See Lyons v. Westinghouse Elec. Corp., 222 F.2d 184, 188 (2d Cir.1955); Calvert Fire Ins. Co. v. American Mutual Reinsurance Co., 600 F.2d 1228, 1236 n. 18 (7th Cir.1979) (dictum). But the essential issues in most federal antitrust eases — relevant market, market power, effect on competition, monopolistic intent, and so forth — are factual. If state courts are competent to decide those issues with preclusive effect in any subsequent federal suit, and the state’s antitrust law is identical to the federal law, we do not see how an antitrust claimant can object to being required, in the interest of judicial economy, to litigate his antitrust claim in state court under state law when he is already in state court, by his choice, on another claim. That is the essential insight of Nash.
But Nash does not speak directly to a case such as this where the state antitrust statute is not identical to the federal. The question what to do in such a case was reserved in Nash. 640 F.2d at 490, 492. The plaintiffs like the answer given in Lyons, supra, which held that a state court judgment would not bar a federal antitrust suit, in part because treble damages could not have been obtained in the state suit. [1092]*1092See 222 F.2d at 189. That case is distinguishable from this one, though. For one thing, Lyons had no chance of getting treble damages in state court, whereas Marrese and Treister would have had some chance of getting treble damages under the Illinois Antitrust Act if they had sued under that act. But they would have had two hurdles to jump that would not face them in a federal antitrust suit. They would have had to prove a willful violation and would have had to appeal to the trial court’s discretion to award treble damages. The approach of the Illinois act is, much as in the typical criminal statute, to set a maximum penalty which the “sentencing” court presumably will impose only in extreme cases, rather than a definite penalty as under federal antitrust law.
But there is another and better distinction between this case and Lyons. Since Westinghouse had sued Lyons in state court for breach of contract and Lyons had interposed the defense that the contract was illegal under the Sherman Act, the question was whether a judgment in the state suit on the merits of this defense would bar Lyons from maintaining a Sherman Act suit against Westinghouse in federal court. The court held that it would not, in part as we have said because Lyons could not have obtained treble damages (or for that matter any damages) from successfully asserting his defense of antitrust illegality to Westinghouse’s state law contract claim. If res judicata had applied, Westinghouse would by suing in state court have succeeded in putting Lyons to the choice of either giving up his claim to treble damages or abandoning a possibly meritorious defense in the state suit. Cf. New York State Teamsters Conf. Pension & Retirement Fund v. Pension Benefit Guar. Corp., 591 F.2d 953, 957 (D.C.Cir.1979). The plaintiffs in the present case did not face this painful choice. If they wanted a state forum they could have joined a state antitrust count to their other state law counts and they would then have had a shot at treble damages, though not so sure a one as if they had brought a federal suit. Alternatively, they could have sued in federal court first, alleging a violation of the Sherman Act and joining to their federal claim the state law claims which they instead decided to sue on in state court. The federal court would have had jurisdiction to decide the state law claims under the doctrine of pendent jurisdiction.
Neither choice was as attractive to the plaintiffs as the one they made, which was to try the case piecemeal, litigating their state law claims in state court and their federal law claim in federal court with an adverse outcome in the first litigation having (they hoped) no preclusive effect on the second. No reason for this procedure has been suggested. And against its advantages to the plaintiffs must be balanced the inconvenience to the defendant, and to the state and federal judicial systems, of allowing a plaintiff to try his case in stages with the decision at one stage having no preclusive effect in subsequent stages (except that actual factfindings made at one stage would, as noted earlier, have collateral-estoppel effect at a subsequent stage). We think the balance favors preclusion where, as in this case, the relevant standard of liability is the same under state and federal law and the remedies are similar though not identical. The plaintiffs’ alternatives to piecemeal litigation were not so paltry that the courts should have to entertain such litigation at a time when both the state and federal judicial systems are staggering under caseloads of unprecedented magnitude.
We add that even if the district judge was correct in rejecting the Academy’s res judicata argument, he should have considered the piecemeal character of this litigation as part of his appraisal of the hardship to the Academy of complying with the plaintiffs’ discovery request. Because the plaintiffs put the Academy to unnecessary legal expense by splitting their claim as they did, their request for a form of discovery that would place a special burden on their opponent should have made a weaker appeal to the district judge’s equitable discretion under Rule 26(c) than if the plaintiffs had been litigating their claims [1093]*1093against the Academy in a more parsimonious manner.
There is an additional reason for believing that the plaintiffs’ antitrust suit may be groundless whatever the contested discovery might reveal. This reason has as we shall see different implications for the district court’s exercise of its Rule 26(c) powers: not that it should have denied the requested discovery altogether but that it should have postponed discovery of the membership files pending other, less burdensome discovery that might quickly have shown whether the suit was indeed groundless as a matter of antitrust law.
The great watershed of that law is the distinction between per se illegality and illegality under the Rule of Reason. If a practice is within the per se category, all you have to prove to establish a violation is that the defendant engaged in the practice; you do not have to show that in fact the practice has had or will have an adverse effect on competition. But if a practice is not within the per se category — if it is governed, in other words, by the Rule of Reason (the dichotomy is not perfect, as there are some practices, mergers for example, that are governed by another standard altogether, but this refinement is irrelevant to this case) — then the plaintiff must, to prevail, show not only that the defendant engaged in the practice but also that by doing so the defendant was injuring competition. “[A]ny rule of reason analysis requires a showing of anticompetitive market effect.” Lektro-Vend Corp. v. Vendo Co., 660 F.2d 255, 268 (7th Cir.1981); see Dos Santos v. Columbus-Cuneo-Cabrini Medical Center, 684 F.2d 1346, 1352 (7th Cir.1982); Havoco of America, Ltd. v. Shell Oil Co., 626 F.2d 549, 558 (7th Cir.1980).
The denial of the plaintiffs’ applications for membership in the Academy, a collective refusal by the Academy’s members to deal with the plaintiffs on the identical terms on which they deal with each other, was a form of boycott. It used to be said often that boycotts were illegal per se. But that was never entirely true. The best known statement of the Rule of Reason is found in a boycott case, Board of Trade of City of Chicago v. United States, 246 U.S. 231, 38 S.Ct. 242, 62 L.Ed. 683 (1918). The case involved the legality of an internal rule of a commodities exchange designed to limit competition from some of its members, and a violation of the rule would have been punishable by expulsion. A more accurate generalization today would be that boycotts are illegal per se only if they are used to enforce agreements that are themselves illegal per se — for example, price-fixing agreements. See United States Trotting Ass’n v. Chicago Downs Ass’n, 665 F.2d 781, 787-90 (7th Cir.1981) (en banc), for the general principle, and Spray-Bite Serv. Corp. v. Monsanto Co., 684 F.2d 1226, 1234-36 (7th Cir.1982), for the price-fixing exception. .At least that is the rule for organized associations having some lawful purposes; we need not consider whether a conspiracy or other ad hoc “association,” if anticompetitive in intent, would be treated under a harsher standard.
It is true that some early cases, such as United States v. Terminal Railroad Ass’n of St. Louis, 224 U.S. 383, 32 S.Ct. 507, 56 L.Ed. 810 (1912), hold or imply that a boycott which has the effect of totally excluding a competitor from a market is, for that reason alone, illegal; and this could be viewed as a species of per se illegality. As it is uncontested that the American Academy of Orthopaedic Surgeons has no power to prevent anyone from practicing orthopaedic surgery, these cases are of doubtful relevance here; so we need not consider to what extent, with their emphasis on the welfare of competitors rather than consumers (the excluded railroad in the Terminal Ass’n case appears to have been complaining about its exclusion from a cartel!), these cases can survive the consumer-oriented view of antitrust that prevails today. See, e.g., Reiter v. Sonotone Corp., 442 U.S. 330, 343, 99 S.Ct. 2326, 2333, 60 L.Ed.2d 931 (1979). Reviewing these cases recently this court held that boycotts cannot be deemed illegal per se on the basis of their impact on competitors (rather than on the consuming public) if there is no “direct [1094]*1094effort to influence the supply of, or demand for, a competitor’s product.” Phil Tolkan Datsun, Inc. v. Greater Milwaukee Datsun Dealers’ Advertising Ass’n, Inc., 672 F.2d 1280, 1286 (7th Cir.1982). None is alleged in this case. Treister and Marrese are competitors of Academy members and membership may be valuable in competing with both member and nonmember orthopaedists, but there is no contention that the Academy is interfering directly with the plaintiffs’ access to suppliers or customers, as by trying to get their hospital staff privileges revoked. The alleged effect of denial of membership on the plaintiffs’ ability to practice orthopaedic medicine is indirect, just as denying a trial lawyer membership in the American College of Trial Lawyers would have only an indirect effect on his ability to obtain clients. We are speaking of matters of degree of course, but an “indirect” effect, by which we mean one falling far short of complete exclusion from the market, will not support a per se attack on a boycott.
Nor do the plaintiffs allege a price-fixing conspiracy among orthopaedic surgeons, a conspiracy to which the exclusion of the plaintiffs from membership in the Academy might be ancillary. Even if, by taking the (concept of notice ’pleading to its furthest extreme, we could tease such an allegation out of the statement in the complaint that one of the reasons for denying Dr. Treister membership in the Academy was his “known willingness to consult surgical out-patients on a relatively high-volume basis” — “high volume” perhaps implying low price — it would not help the plaintiffs. If the boycott is a form of price-fixing, the plaintiffs probably could have gotten treble damages as a matter of right rather than grace in a suit under the Illinois Antitrust Act, which provides a right to treble damages in price-fixing cases. See Ill.Rev.Stat.1981, ch. 38, §§ 60-3(1), 60-7(2). Nash, a decision we consider correctly decided, would then be indistinguishable from this case and we would have no doubt at all that the suit was barred by res judicata.
What is far more likely, however, is that the plaintiffs are alleging a boycott that, if it is illegal, is so only by virtue of the Rule of Reason; and therefore they must show, as an essential element of illegality, some “anticompetitive market effect” from the boycott. Lektro-Vend Corp. v. Vendo Co., supra, 660 F.2d at 268. To see whether they have done or are likely to be able to do so, assume that all of the facts alleged by the plaintiffs are true. (This is too generous an assumption, for some of the plaintiffs’ allegations were denied in an affidavit that the plaintiffs did not counter with an affidavit of their own.) Then the plaintiffs were denied admission to the Academy because they are fierce competitors; others like them are also denied admission for this reason; and membership in the Academy carries with it such great professional advantages that most aspirants to membership — and all orthopaedic surgeons who are not already members are aspirants — will (if this suit fails) have an incentive to avoid engaging in the kind of competitive behavior that these plaintiffs engaged in and that cost them their chance of becoming members.
But even with all this granted the plaintiffs have not come close to showing the kind of competitive effects required in a Rule of Reason case. According to their own statements, the vast majority of orthopaedic surgeons in the U.S. who have been practicing for at least three years (at oral argument one of the plaintiffs’ counsel said 99 percent, though we suspect he was using poetic license) are members of the Academy, and it has 10,000 members in all. The Academy places no restrictions of any sort on its members. So far as it is concerned, they can compete as ferociously as they like. They are free to deal with nonmembers such as Marrese and Treister — to practice in association with them, to refer patients to them, and so forth. The Academy’s meetings are open to nonmembers such as Marrese and Treister, though they object, frivolously as it seems to us, to having to wear nonmember badges. The freedom of members to deal with nonmembers con-
[1095]*1095firms our earlier point that this is not a case where the association makes a “direct effort to influence the supply of, or demand for, a competitor’s product.” Phil Tolkan Datsun, supra, 672 F.2d at 1286.
In these circumstances it is clear that if all of the members of the Academy were in the same market — if all, that is to say, were competitors of one another — the consumer interest in effective competition could not be seriously harmed by the exclusion of these plaintiffs and others like them from the market (not that such exclusion is in the cards: the plaintiffs concede that the Academy has no power to exclude anyone from the practice of orthopaedic surgery). Ten thousand is a vast number of competitors. Unless they explicitly agreed not to compete, and backed up their agreement with elaborate enforcement machinery to prevent individual members from cheating on the agreement, the consumer would be assured the benefits of buying in a competitive market. While there is a sense in which the exclusion of any competitor reduces competition, it is not the sense of competition that is relevant to antitrust law as currently conceived. Products Liability Ins. Agency, Inc. v. Crum & Forster Ins. Cos., 682 F.2d 660, 663-64 (7th Cir.1982). The policy of competition is designed for the ultimate benefit of consumers rather than of individual competitors. The consumer is guaranteed competitive conditions but he is not guaranteed, and he has no interest in the preservation of, a fixed number of competitors far greater than the number required to assure his being able to buy in a competitive market. Maybe the older view would survive in a case of naked aggression resulting in the total exclusion of a competitor from the market; but that would be a per se case (if anything) and this is not.
The 10,000 members of the Academy do not, however, all compete in the same market. The market for orthopaedic surgery is local. Dr. Marrese competes with other orthopaedists in Evansville, Dr. Treister with other orthopaedists in Chicago. The plaintiffs will have the burden at trial of showing that in these local markets the number of orthopaedic surgeons who belong to the Academy is so few that competition among them — competition that the Academy is not alleged to limit or regulate in any way — is insufficient to assure the consuming public the benefits of competition. Unless they can show this they will be unable to ask the trier of fact to draw an inference that either the exclusion of an individual orthopaedist from each market, or the possible effect of that exclusion on the competitive behavior of other aspirants to membership, could result in a higher price or lower quality of orthopaedic surgery in these communities.
The plaintiffs have made no effort as yet in the discovery proceedings in this case to show a probable anticompetitive effect. We do not say they cannot make the required showing; but at this stage of the case, where the only facts of record bearing on competitive effect are the large number of orthopaedists who belong to the Academy and the fact that the Academy does not regulate their behavior in any way, it does not seem to us very likely that they will be able to do so. In these circumstances, we do not think the district court should have ordered discovery of the Academy’s membership files before there was any discovery on the issue of probable anticompetitive effect. A district court has the power under Rules 26(c) and (d) of the Federal Rules of Civil Procedure, and in a clear case the duty, to defer a burdensome discovery request pending completion of discovery on an issue that is likely to dispose of the entire case and thereby make the request moot. See 8 Wright, Miller & Cooper, Federal Practice and Procedure §§ 2040, 2047 (1976). We are not speaking here as we were earlier of denial, but only of postponement; if the former is sometimes appropriate, the latter is more often so. It should have been apparent to the district court, as it is to us, that the present case is likely to wash out over the issue of effect on competition. There would have been no conceivable hardship to the plaintiffs in requiring them to conduct their discovery on this issue before gaining access to the member[1096]*1096ship files. The sequence of discovery may well have been intended to coerce the Academy to settle, but in any event the balance of hardships is clear enough to make us conclude, independently of the res judicata issue, that the district court, in refusing to defer discovery of the membership files, abused its discretion.
Other issues are raised in the briefs, but on the view that we take of the case only one of them need be discussed. The Academy asks us to decide not only that the discovery order was an abuse of discretion but also that the district court’s refusal to dismiss the complaint on the basis of res judicata was erroneous. We have jurisdiction to decide whether the discovery order was an abuse of discretion, because it is the disobedience of that order, that the district judge punished by holding the Academy in criminal contempt. But we would have jurisdiction to review the district judge’s refusal to dismiss the complaint only if he had certified that interlocutory order for an immediate appeal under 28 U.S.C. 1292(b), and he did not do so. He did point out that the Academy could if it wished make its res judicata argument in its appeal from the contempt citation, and he was right, because the merit of that argument is, as we have seen, germane to whether the discovery order was an abuse of discretion. But his, statement did not satisfy the requirement of section 1292(b) that the district judge, if “of the opinion that such [nonappealable interlocutory] order involves a controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal from the order may materially advance the ultimate termination of the litigation, ... shall so state in writing in such order.” Since no such statement was made, we do not have jurisdiction to review the district court’s order refusing to dismiss the complaint, In re Watkins, 271 F.2d 771 (5th Cir.1959), and the appeal from that order is dismissed for lack of jurisdiction in this court. The judgment of contempt is reversed. Costs in this court to the appellant.
So Ordered.