Reset A A Font size: Print

United States Court of Appeals, District of Columbia Circuit.

Antony BROWN;  James Bishop;  John Buddenberg;  Gary Couch;  Craig Davis;  Ricky Andrews;  Thom Kaumeyer;  Wesley Pritchett;  and John Simpson, Individually and on Behalf of All Class Members, Appellees, v. PRO FOOTBALL, INC., d/b/a Washington Redskins;  The Five Smiths, Inc., d/b/a Atlanta Falcons;  Buffalo Bills, Inc., d/b/a Buffalo Bills;  Chicago Bears Football Club, Inc., d/b/a Chicago Bears;  Cincinnati Bengals, Inc., d/b/a Cincinnati Bengals;  Cleveland Browns, Inc., d/b/a Cleveland Browns;  Dallas Cowboys Football Club, Ltd., d/b/a Dallas Cowboys;  PDB Sports, Ltd., d/b/a Denver Broncos;  The Detroit Lions, Inc., d/b/a Detroit Lions;  Green Bay Packers, Inc., d/b/a Green Bay Packers;  Houston Oilers, Inc., d/b/a Houston Oilers;  Indianapolis Colts, Inc., d/b/a Indianapolis Colts;  Kansas City Chiefs Football Club, Inc., d/b/a Kansas City Chiefs;  The Los Angeles Raiders, Ltd., d/b/a Los Angeles Raiders;  Los Angeles Rams Football Company, Inc., d/b/a Los Angeles Rams;  Miami Dolphins, Ltd., d/b/a Miami Dolphins;  Minnesota Vikings Football Club, Inc., d/b/a Minnesota Vikings;  KMS Patriots, L.P., d/b/a New England Patriots;  The New Orleans Louisiana Saints Limited Partnership, d/b/a New Orleans Saints;  New York Football Giants, Inc., d/b/a New York Giants;  New York Jets Football Club, Inc., d/b/a New York Jets;  The Philadelphia Eagles Football Club, Inc., d/b/a Philadelphia Eagles;  B & B Holdings, Inc., d/b/a Phoenix Cardinals;  Pittsburgh Steelers Sports, Inc., d/b/a Pittsburgh Steelers;  The Chargers Football Company, d/b/a San Diego Chargers;  The San Francisco Forty-Niners, Ltd., d/b/a San Francisco Forty-Niners;  The Seattle Professional Football Club, d/b/a Seattle Seahawks;  Tampa Bay Area NFL Football, Inc., d/b/a Tampa Bay Buccaneers, Inc.;  and National Football League, Appellants.

Nos. 93-7165 et al., 94-7071.

Decided: March 21, 1995

Before:  EDWARDS, Chief Judge, WALD and RANDOLPH, Circuit Judges.Before:  EDWARDS, Chief Judge, and WALD, SILBERMAN, BUCKLEY, WILLIAMS, GINSBURG, SENTELLE, HENDERSON, RANDOLPH, ROGERS and TATEL, Circuit Judges. Gregg H. Levy, Washington, DC, argued the cause for appellants.   With him on the briefs were Herbert Dym and Sonya D. Winner, Washington, DC.   Richard W. Buchanan, Washington, DC, entered an appearance, for appellants. Joseph A. Yablonski, Washington, DC, argued the cause for appellees.   With him on the briefs were Daniel B. Edelman and John F. Colwell, Washington, DC. Howard L. Ganz, New York City and Warren L. Dennis, Washington, DC, filed amicus curiae brief for The Nat. Basketball Ass'n and The Nat. Hockey League. Peter D. Isakoff, Washington, DC and David G. Feher, New York City, filed amicus curiae brief for The Nat. Basketball Players Ass'n and The Nat. Hockey League Players Ass'n.

 This case poses a conflict between the policies underlying federal labor law and antitrust law in the context of a labor dispute involving professional football.   In the Sherman Act, 15 U.S.C. § 1 (Supp. II 1990), enacted in 1890, Congress proscribed certain practices and agreements inimical to free trade as a means “to promote the national interest in a competitive economy.”  Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 635, 105 S.Ct. 3346, 3358, 87 L.Ed.2d 444 (1985) (internal quotations omitted).   There was no unified federal labor policy at the time of the passage of the Sherman Act.   However, over fifty years later, when Congress passed the National Labor Relations Act (“NLRA”), 29 U.S.C. § 151 et seq. (1988), “it set down a federal labor policy ․ plainly meant to do more than simply alter the prevailing substantive law.   It sought as well to restructure fundamentally the processes for effectuating that policy, deliberately placing the responsibility for applying and developing this comprehensive legal system in the hands of an expert administrative body [the National Labor Relations Board (“NLRB”) ] rather than the federalized judicial system.”  Amalgamated Ass'n of Street, Elec. Ry. & Motor Coach Employees v. Lockridge, 403 U.S. 274, 288, 91 S.Ct. 1909, 1918, 29 L.Ed.2d 473 (1971).   A principal tenet of this federal labor policy is that settlement of collective bargaining disputes should be achieved by “subjecting labor-management controversies to the mediatory influence of negotiation,” not litigation.  Fibreboard Paper Prods. Corp. v. NLRB, 379 U.S. 203, 211, 85 S.Ct. 398, 402, 13 L.Ed.2d 233 (1964).   In this case, we must determine whether the nation's labor laws or antitrust policy control where, after bargaining in good faith to a point of impasse, an employer group takes unilateral action to impose a fixed salary for a category of employees as an otherwise lawful step in the collective bargaining process established by the NLRA.

In 1989, the 28 clubs of the National Football League (“NFL”) were engaged in collective bargaining with the NFL Players Association (“NFLPA”), the players' collective bargaining representative.   During the course of bargaining, the NFL proposed to pay a fixed salary of $1,000 per week to any player assigned to newly formed practice squads.   Had the parties been able to reach a settlement on this issue, they could have concluded an agreement establishing $1,000 per week as the salary for practice squad players, and this agreement would have posed no legal problems under the federal labor or antitrust laws.   Such was not to be the case, however, for the parties bargained to impasse over the issue, after which the clubs unilaterally imposed the fixed salary for the 1989 NFL season.   In response to the clubs' action, nine players who had been assigned to practice squads filed this class action antitrust lawsuit against the clubs and the NFL in the District Court, alleging that the fixed salary constituted an unreasonable restraint of trade in violation of the Sherman Act.   During four years of ensuing litigation, the District Court held that the defendants' agreement on a fixed salary violated the Sherman Act, and, after a trial to determine damages, entered a judgment against the clubs and the NFL in the amount of $30,349,642, and enjoined them from ever again setting a uniform salary for any class of players.

 While the clubs and the NFL raise a number of challenges to the District Court's actions, we need not address most of them, for we hold that the District Court erred in rejecting the appellants' claim that the nonstatutory labor exemption shields them from liability in this case.   This exemption, a judicially-created doctrine designed to reconcile federal labor and antitrust policies, has had an important place in federal jurisprudence for almost 30 years.   Although there has been much debate over the years regarding the scope of the exemption, there is at least one principle that seems clear:  restraints on competition lawfully imposed through the collective bargaining process are exempted from antitrust liability so long as such restraints primarily affect only the labor market organized around the collective bargaining relationship.   Thus, employees confronted with actions imposed lawfully through the collective bargaining process must respond not with a lawsuit brought under the Sherman Act, but rather with the weapons provided by the federal labor laws.

The NFL was free to take unilateral action after impasse (just as the NFLPA was free to strike), because the action was a legitimate economic weapon available to be used in an attempt to force a settlement.   This is exactly what federal labor policy condones, as the Supreme Court has often recognized:

[A] particular activity might be ‘protected’ by federal law not only [where it falls] within § 7 [of the NLRA], but also when it [is] an activity that Congress intended to be ‘unrestricted by any governmental power to regulate’ because it [is] among the permissible ‘economic weapons in reserve ․ actual exercise [of which] on occasion by the parties, is part and parcel of the system that the Wagner and Taft-Hartley Acts have recognized.’

Lodge 76, Int'l Ass'n of Machinists v. Wisconsin Employment Relations Comm'n, 427 U.S. 132, 141, 96 S.Ct. 2548, 2553, 49 L.Ed.2d 396 (1976) (quoting NLRB v. Insurance Agents' Int'l Union, 361 U.S. 477, 488, 489, 80 S.Ct. 419, 426, 427, 4 L.Ed.2d 454 (1960)).   And, as the Second Circuit has recently held:

[T]he antitrust laws do not prohibit employers from bargaining jointly with a union, from implementing their joint proposals in the absence of a [collective bargaining agreement], or from using economic force to obtain agreement to those proposals.   What limits on such conduct that exist are found in the labor laws.

National Basketball Ass'n v. Williams, 45 F.3d 684, 693 (2d Cir.1995).   We agree.

Because the NFL in this case acted lawfully within the framework of the collective bargaining process when it unilaterally imposed a fixed salary for practice-squad players, and because this action affected only the market for professional football player services, the nonstatutory labor exemption precludes any finding of liability under the Sherman Act.   Accordingly, we reverse the decision of the District Court.

I. Background

This case arises from a labor dispute over salaries for a limited number of professional football players whose jobs required them to practice with regular NFL players, and to replace regular players who became injured, but not otherwise to play in NFL football contests.   In 1987, a collective bargaining agreement governing the terms and conditions of employment for all professional football players expired, and the NFL and NFLPA began negotiations for a new agreement.   In early 1989, with the two sides making little progress toward such an agreement, the owners adopted an amendment to the NFL Constitution that altered the rules governing players on the clubs' injured-reserve lists and established new Developmental Squads of practice and replacement players.   The amendment, known as Resolution G-2, allowed each club to maintain a Developmental Squad of as many as six rookie or “first-year” 1 practice and replacement players in addition to its usual 47-player roster.   Resolution G-2 departed from the customary NFL practice of setting player salaries through individual negotiation.   It anticipated a fixed salary for the Developmental Squad players, though it did not establish the amount of that salary.

The NFL and NFLPA engaged in fruitless negotiations over Resolution G-2.   On April 7, NFL Management Committee (“NFLMC”) Executive Director Jack Donlan solicited a meeting with NFLPA Executive Director Gene Upshaw to negotiate the terms and conditions of employment for Developmental Squad players.   On May 17, 1989, a league management committee proposed that the salary for Developmental Squad players be set at $1,000 per week.   On May 18, Donlan again sought a meeting with Upshaw.   Subsequently, on May 30, 1989, Upshaw responded with a letter stating that the NFLPA agreed only “that players can be listed on a developmental squad of an NFL club if they have all of the benefits and protections which players on the Active List have,” including “the right to negotiate their own salaries.”   Letter from Gene Upshaw to Jack Donlan (May 30, 1989), reprinted in Joint Appendix (“J.A.”) 54.   Donlan and Upshaw met on June 16, 1989, to discuss the Developmental Squad portion of Resolution G-2.   In a letter to Donlan memorializing the discussions at that meeting, Upshaw rejected the fixed salary component of the Developmental Squad proposal, holding to the position that “all players, including developmental, should have the right to negotiate salary terms, and that no fixed wage for any group of players is acceptable to the NFLPA.”   Letter from Gene Upshaw to Jack Donlan (July 6, 1989), reprinted in J.A. 58.   As a result, Donlan concluded that the issue was “clearly at impasse” for “implementation purposes.”   Letter from Jack Donlan to Hugh Culverhouse et al. (June 16, 1989), reprinted in J.A. 55.

The NFL then unilaterally implemented the Developmental Squad program by distributing uniform contracts for Developmental Squad players to all teams.   Club officials were advised that paying any such player more or less than $1,000 per week would result in disciplinary action, including loss of future draft choices.   See Memorandum from Pete Rozelle, NFL Commissioner, to NFL General Managers, Player Personnel Directors (Aug. 24, 1989) at 1, reprinted in J.A. 2538.   Resolution G-2 allowed the clubs to form their Developmental Squads from players remaining available after each club reduced its regular roster to 47 players on September 4.   Under the resolution, the clubs could sign Developmental Squad players to contracts after 4 p.m. on September 5, 1989.   During the 1989 season, 236 players signed Developmental Squad contracts.

On May 9, 1990, appellee Antony Brown and eight other Developmental Squad players 2 (“the Players”) brought a class action lawsuit against all 28 NFL clubs and the NFL itself on behalf of 235 of the 1989 Developmental Squad players, alleging that the defendants engaged in price-fixing in violation of the Sherman Act by setting a $1,000 fixed salary for such players.   On June 4, 1991, the District Court granted the Players' motion for partial summary judgment, and denied the NFL's cross-motion, on the issue of whether the Players' suit was precluded by the nonstatutory labor exemption to the antitrust laws.  Brown v. Pro Football, Inc., 782 F.Supp. 125 (D.D.C.1991).   The District Court relied on three alternative rationales to support its judgment.   First, it held that the exemption ended when the parties' collective bargaining agreement expired in 1987.  Id. at 130-34.   Second, the court held that, even if the exemption survived the expiration of the collective bargaining agreement, it ended when the parties reached impasse regarding the issue of Developmental Squad player salaries.  Id. at 134-37.   Finally, the District Court held that, in any event, the exemption was inapplicable because it protects only restraints on competition contained in collective bargaining agreements, and the fixed salary had not previously been encompassed in any agreement between the NFL and NFLPA.  Id. at 137-39.

On March 10, 1992, the District Court granted the Players' motion for summary judgment on the issue of antitrust liability.  Brown v. Pro Football, Inc., 1992-1 Trade Cas. (CCH) ¶ 69,747, 1992 WL 88039 (D.D.C.1992), reprinted in J.A. 422.   With liability established, the District Court on September 21, 1992, began a ten-day jury trial on the issues of antitrust injury and damages.   The jury awarded damages to the players in the class that, when trebled in accordance with section 4 of the Clayton Act, 15 U.S.C. § 15 (1988), totaled $30,349,642.  Brown v. Pro Football, Inc., Civ.Action No. 90-1071 (D.D.C. Oct. 5, 1992) (judgment on the verdict), reprinted in J.A. 2714.   In the wake of this verdict, the District Court denied the clubs' motion for judgment as a matter of law, or a new trial, and granted the Players' request for a permanent injunction barring the clubs from ever again setting a uniform regular season salary for any category of players.  Brown v. Pro Football, Inc., 821 F.Supp. 20 (D.D.C.1993), reprinted in J.A. 2911.   Finally, on March 1, 1994, the District Court awarded the Players' counsel $1,744,578.41 in attorney's fees.  Brown v. Pro Football, Inc., 846 F.Supp. 108 (D.D.C.1994).3

II. Analysis

On appeal, the clubs and the NFL challenge each of the District Court's decisions.   However, we address only the District Court's rejection of appellants' nonstatutory labor exemption defense, because we deem it dispositive.   We review de novo the District Court's entry of summary judgment on this issue.  SEC v. Bilzerian, 29 F.3d 689, 695 (D.C.Cir.1994).

Appellants contend that the nonstatutory labor exemption applies to all restraints on competition imposed through the collective bargaining process, even those imposed unilaterally by an employer.   In their view, any other rule would disrupt the balance of power between unions and employers that exists under federal labor law.   The Players, meanwhile, argue that the exemption applies only where a union has consented to a restraint on competition.   They contend that the exemption must be narrowly construed to apply only where unions act in tandem with employers, as, for example, by entering into collective bargaining agreements.

After reviewing relevant Supreme Court precedent and the policies underlying both the NLRA and the Sherman Act, we conclude that the nonstatutory labor exemption shields from antitrust challenge alleged restraints on competition imposed through the collective bargaining process, so long as the challenged actions are lawful under the labor laws and primarily affect only a labor market organized around a collective bargaining relationship.   Because the fixed salary for Developmental Squad players is such an action, we hold that the exemption shields the clubs and the NFL from liability in this case.

A. Origin of the Exemption

 An exemption to the antitrust laws for activities related to collective bargaining traces its origin to sections 6 and 20 of the Clayton Act, 15 U.S.C. § 17 (1988), 29 U.S.C. § 52 (1988), and to the Norris-LaGuardia Act, 29 U.S.C. § 101 et seq. (1988).  Clayton Act section 6 excludes human labor from the definition of a commodity and provides that the antitrust laws do not prohibit labor organizations.   Section 20 of that Act, along with the Norris-LaGuardia Act, limits the authority of federal courts to enjoin specified union activities.   While none of these statutory provisions is phrased in terms of an exemption from the Sherman Act, the Supreme Court has interpreted them generally to waive antitrust liability for unilateral union conduct such as boycotts and picketing.   See H.A. Artists & Assoc., Inc. v. Actors' Equity Ass'n, 451 U.S. 704, 714-15, 101 S.Ct. 2102, 2108-09, 68 L.Ed.2d 558 (1981);  United States v. Hutcheson, 312 U.S. 219, 232, 61 S.Ct. 463, 466, 85 L.Ed. 788 (1941).   This exemption, known as the statutory labor exemption, does not exempt bilateral activity, such as “concerted action or agreements between unions and nonlabor parties.”  Connell Constr. Co., Inc. v. Plumbers & Steamfitters Local No. 100, 421 U.S. 616, 622, 95 S.Ct. 1830, 1835, 44 L.Ed.2d 418 (1975).   However, the Supreme Court has recognized

that a proper accommodation between the congressional policy favoring collective bargaining under the NLRA and the congressional policy favoring free competition in business markets requires that some union-employer agreements be accorded a limited nonstatutory exemption from antitrust sanctions.

Id.  Unlike the statutory exemption, this nonstatutory exemption is available to both unions and employers.   See Scooper Dooper, Inc. v. Kraftco Corp., 494 F.2d 840, 847 n. 14 (3d Cir.1974);  see also, e.g., Powell v. NFL, 930 F.2d 1293, 1303 (8th Cir.1989), cert. denied, 498 U.S. 1040, 111 S.Ct. 711, 112 L.Ed.2d 700 (1991).

Although the Supreme Court has recognized a nonstatutory labor exemption to the antitrust laws, the scope of the exemption never has been conclusively delimited.   Instead, the Court's cases in this area mark out only the general boundaries of the doctrine.   The Court's first major decision addressing a restraint on competition imposed through the collective bargaining process came in Allen Bradley Co. v. Local No. 3, Int'l Bhd. of Elec. Workers, 325 U.S. 797, 799-800, 65 S.Ct. 1533, 1535-36, 89 L.Ed. 1939 (1945), in which a local union of electrical workers in New York City reached an agreement with local manufacturers and contractors requiring the contractors to buy equipment only from manufacturers employing union members, and the manufacturers to sell only to contractors employing union members.   This arrangement severely curtailed competition from firms outside the city, inflating prices for electrical equipment in the local market.  Id. at 800, 65 S.Ct. at 1535.   When excluded manufacturers challenged the agreement under the Sherman Act, the union claimed an exemption from antitrust liability.   The Supreme Court rejected this claim, concluding that “Congress never intended that unions could, consistently with the Sherman Act, aid non-labor groups to create business monopolies and to control the marketing of goods and services.”   Id. at 808, 65 S.Ct. at 1539.

The Court reached a similar conclusion two decades later in United Mine Workers v. Pennington, 381 U.S. 657, 85 S.Ct. 1585, 14 L.Ed.2d 626 (1965), in which it addressed an antitrust challenge to a contract between a union and large coal mining companies by which the union agreed to impose a new, higher wage upon smaller coal companies as part of a concerted effort to drive the smaller firms from the industry.   A majority of the Court rejected the union's claim to an exemption from the Sherman Act.  Id. at 661, 85 S.Ct. at 1588.   While Justice White's opinion for the Court acknowledged that “a union may conclude a wage agreement with [a] multi-employer bargaining unit without violating the antitrust laws and ․ may as a matter of its own policy, and not by agreement with all or part of the employers of that unit, seek the same wages from other employers,” it held that “a union forfeits its exemption from the antitrust laws when it is clearly shown that it has agreed with one set of employers to impose a certain wage scale on other bargaining units.”  Id. at 664-65, 85 S.Ct. at 1590-91.

Pennington 's promise of a nonstatutory labor exemption to the Sherman Act was realized in a companion case, Local No. 189, Amalgamated Meat Cutters & Butcher Workmen v. Jewel Tea Co., Inc., 381 U.S. 676, 85 S.Ct. 1596, 14 L.Ed.2d 640 (1965).   There, the Court, in opinions by Justices White and Goldberg, applied the exemption to shield from antitrust liability a multi-union, multi-employer agreement to close food store meat departments at 6 p.m. so as to prevent butchers from being replaced by self-service markets or unskilled workers during night hours.   Justice White, writing for three justices, balanced the interest of union workers against the impact on the product market, finding the restriction on working hours to be of “immediate and direct” concern to union members and noting that the subject matter of the agreement-employees' working hours-was “well within the realm of ‘wages, hours, and other terms and conditions of employment’ about which employers and unions must bargain” under the NLRA.  Id. at 691, 85 S.Ct. at 1602.   Justice Goldberg, writing for three other justices, went further, concluding that all “collective bargaining activity concerning mandatory subjects of bargaining under the Labor Act is not subject to the antitrust laws.”  Id. at 710.

Finally, in its most recent opinion on the subject, the Court in Connell Construction, 421 U.S. at 616, 95 S.Ct. at 1832, held the nonstatutory exemption inapplicable where a plumbers union forced a general contractor to sign an agreement requiring the contractor to subcontract construction work only to firms maintaining collective bargaining agreements with the union.   The Court found that this arrangement “indiscriminately excluded nonunion subcontractors from a portion of the market, even if their competitive advantages were not derived from substandard wages and working conditions but rather from more efficient operating methods.”  Id. at 623, 95 S.Ct. at 1835.   The Court stated:

This kind of direct restraint on the business market has substantial anticompetitive effects, both actual and potential, that would not follow naturally from the elimination of competition over wages and working conditions.   It contravenes antitrust policies to a degree not justified by congressional labor policy, and therefore cannot claim a nonstatutory exemption from the antitrust laws.

Id. at 625, 95 S.Ct. at 1836.

In assessing the relevance of these cases to the issue before us, we note initially that the Court never has considered whether the exemption applies where, as here, an antitrust claim arises from a dispute over a negotiable condition of employment between parties to the collective bargaining process.   Nevertheless, the Players contend that the Supreme Court's precedents dictate a rule to govern this case.   They seek to extract from the Court's opinions a rule that the exemption applies only where a union has manifested its consent to a restraint on trade by signing a collective bargaining agreement.   They emphasize that both the Connell Construction and Jewel Tea opinions refer to the nonstatutory exemption as a shield from antitrust liability for union-employer “agreements.” 4  Jewel Tea Co., 381 U.S. at 689, 85 S.Ct. at 1601 (White, J.);  Connell Constr. Co., 421 U.S. at 622, 95 S.Ct. at 1835.   Absent union consent, the Players contend, the exemption is inapplicable.   We disagree.   Indeed, we believe that the Players' view is based on a distorted reading of the case law.

The language in the opinions upon which the Players rely must be read in context.   As the previous discussion indicates, each of the pertinent Supreme Court cases involved an antitrust challenge to a union-employer agreement.   In such circumstances, the Court's reference to the nonstatutory labor exemption in terms of a potential shield from antitrust liability for such agreements is hardly surprising.5  Certainly, no Supreme Court case expressly limits the exemption in the manner suggested by the appellees.   Thus, nothing in any of the Court's opinions, from Allen Bradley to Connell Construction, disposes of the issue before us.

While not dispositive, however, the Court's opinions do provide significant guidance for our reconciliation of the competing antitrust and labor policies in this case.   First, in its most extensive discussion of the nonstatutory labor exemption, the Court characterized the doctrine as representing a “proper accommodation between the congressional policy favoring collective bargaining under the NLRA and the congressional policy favoring free competition in business markets.”  Connell Constr. Co., 421 U.S. at 622, 95 S.Ct. at 1835 (emphasis added).   Thus, the Court recognized that the juxtaposition of policies giving rise to the exemption focuses on collective bargaining as a process, not merely on the product of that process-the collective bargaining agreement.

We also observe that, in applying the nonstatutory labor exemption, the Court consistently has struggled to balance the interests of those involved in collective bargaining against the impact of their activities on the product market, paying little attention to any impact on the labor market.   See, e.g., Pennington, 381 U.S. at 663, 85 S.Ct. at 1589 (stating that union and mining companies who agreed on price at which companies would sell their coal would be unshielded by exemption because “the restraint on the product market is direct and immediate”);  Jewel Tea, 381 U.S. at 690 n. 5, 85 S.Ct. at 1602 n. 5 (stating that “crucial determinant” for application of exemption to an agreement is the agreement's “relative impact on the product market and the interests of union members”);  see also Wood v. National Basketball Ass'n, 809 F.2d 954, 963 (2d Cir.1987) (“Each of the [Supreme Court's] decisions [applying the nonstatutory labor exemption] involved injuries to employers who asserted that they were being excluded from competition in the product market.”  (emphasis omitted)).   Nothing in the Court's opinions supports appellees' attempt to import antitrust principles into a labor market organized around a collective bargaining relationship.  “Quite the contrary, the issue has been whether even to go so far as to impose antitrust sanctions for the kind of product market activities involved in Pennington.”   Michael S. Jacobs & Ralph K. Winter, Jr., Antitrust Principles and Collective Bargaining by Athletes:  Of Superstars in Peonage, 81 Yale L.J. 1, 27 (1971).   In short, we agree with those commentators who have observed that

[f]rom Allen Bradley to Pennington, the majority of the Court has insisted that one factor be present before the Sherman Act applies to arrangements arrived at through collective bargaining:  one group of employers must conspire to use the union to hurt their competitors.   The line the Court has consistently sought to draw, therefore, is the line between the product market and the labor market.

Id. at 26.

 Thus, from the Court's opinions we derive two principles to guide our application of the nonstatutory labor exemption.   First, the exemption must be broad enough in scope to shield the entire collective bargaining process established by federal law.   Second, the case for applying the exemption is strongest where a restraint on competition operates primarily in the labor market and has no anti-competitive effect on the product market.   We believe these principles find support not only in the Supreme Court's precedents, but also in the policies underlying both the NLRA and the Sherman Act.


 The NLRA makes clear that federal labor policy focuses on collective bargaining as a process, rather than collective bargaining agreements alone.   In the NLRA, Congress established “the mutual obligation of the employer and the representative of the employees to meet at reasonable times and confer in good faith with respect to wages, hours, and other terms and conditions of employment.”  29 U.S.C. § 158(d).   This obligation “is premised on the belief that collective discussions backed by the parties' economic weapons will result in decisions that are better for both management and labor and for society as a whole.”  First Nat'l Maintenance Corp. v. NLRB, 452 U.S. 666, 678, 101 S.Ct. 2573, 2580, 69 L.Ed.2d 318 (1981).   Thus, two factors define the collective bargaining process under federal law:  the “necessity for good-faith bargaining between parties, and the availability of economic pressure devices to each to make the other party incline to agree on one's terms.”  Insurance Agents' Int'l Union, 361 U.S. at 489, 80 S.Ct. at 427.

 As the foregoing suggests, collective bargaining under the NLRA is a carefully defined bilateral process.   Unions and employers alike enjoy protections under the collective bargaining process.   Unions are not the sole beneficiaries of the NLRA.   For example, unions may strike, see NLRB v. Washington Aluminum Co., 370 U.S. 9, 14, 82 S.Ct. 1099, 1102, 8 L.Ed.2d 298 (1962);  see also 29 U.S.C. § 163, but employers may lock out their workers, see American Ship Bldg. Co. v. NLRB, 380 U.S. 300, 310-11, 85 S.Ct. 955, 963-64, 13 L.Ed.2d 855 (1965).   Further, unions are protected against unilateral action by employers with respect to subjects of mandatory bargaining while negotiations over such subjects are ongoing, see NLRB v. Katz, 369 U.S. 736, 743, 82 S.Ct. 1107, 1111, 8 L.Ed.2d 230 (1962), and continue to enjoy this protection until negotiations end in impasse, see Southwestern Steel & Supply, Inc. v. NLRB, 806 F.2d 1111, 1113 (D.C.Cir.1986).   At that point, however, “an employer does not violate the Act by making unilateral changes that are reasonably comprehended within his pre-impasse proposals.”  NLRB v. McClatchy Newspapers, Inc., 964 F.2d 1153, 1165 (D.C.Cir.1992) (Edwards, J., concurring in denial of petition for enforcement) (emphasis omitted) (quoting American Fed'n of Television & Radio Artists v. NLRB, 395 F.2d 622, 624 (D.C.Cir.1968)).   Employers may not fire workers engaged in economic strikes, and must rehire such workers at the conclusion of such a strike absent “legitimate and substantial business justifications” for doing otherwise.   NLRB v. Fleetwood Trailer Co., Inc., 389 U.S. 375, 378, 88 S.Ct. 543, 545, 19 L.Ed.2d 614 (1967) (quoting NLRB v. Great Dane Trailers, 388 U.S. 26, 34, 87 S.Ct. 1792, 1797, 18 L.Ed.2d 1027 (1967)).   Under an exception to this rule, however, employers may hire permanent replacements for strikers.  NLRB v. Mackay Radio & Telegraph Co., 304 U.S. 333, 345-46, 58 S.Ct. 904, 910-11, 82 L.Ed. 1381 (1938).   Finally, while the NLRA requires both employers and unions to bargain collectively, it does not require either side to agree.   See 29 U.S.C. § 158(d) (providing that duty to bargain “does not compel either party to agree to a proposal or require the making of a concession”).   Rather, the statute generally leaves the outcome of negotiations to the parties, with government intervention largely proscribed.   See Insurance Agents' Int'l Union, 361 U.S. at 488, 80 S.Ct. at 426 (“Congress intended that the parties should have wide latitude in their negotiations, unrestricted by any governmental power to regulate the substantive solution of their differences.”);  H.K. Porter Co. v. NLRB, 397 U.S. 99, 108, 90 S.Ct. 821, 826, 25 L.Ed.2d 146 (1970) (stating that NLRB's role under the Act is “to oversee and referee the process of collective bargaining, leaving the results of the contest to the bargaining strengths of the parties”).

 Thus, a careful reading of the federal labor laws militates strongly against the Players' argument that the nonstatutory labor exemption shields restraints on competition only when unions have consented to them by signing collective bargaining agreements, for that argument necessarily incorporates the premise that federal labor policy somehow favors unions in the collective bargaining process.   As the terms of the NLRA amply demonstrate, federal labor policy favors neither party to the collective bargaining process, but instead stocks the arsenals of both unions and employers with economic weapons of roughly equal power and leaves each side to its own devices.

 Some commentators have suggested that union agreement ought to be a precondition to any invocation of the nonstatutory labor exemption, because, in their view, the federal labor laws are primarily designed to foster employee rights to engage in collective bargaining.   At one level, this argument is a nonsequitur, for the right to engage in collective bargaining does not encompass the right of agreement.   At another level, the argument expresses more an aspiration than a reality.   It is true that the NLRA strongly protects the right to join and form unions, but at least since Congress passed the Labor Management Relations Act of 1947, 29 U.S.C. § 141 et seq. (1988) (“LMRA”), federal labor law has expressed a “policy of voluntary unionism.”  Pattern Makers' League v. NLRB, 473 U.S. 95, 105, 105 S.Ct. 3064, 3070, 87 L.Ed.2d 68 (1985).   The LMRA expressly gave employees the right to refrain from collective bargaining activity, see 29 U.S.C. § 157, and, while it maintained the obligation of employees to pay union dues under union security agreements, it established that an employer would commit an unfair labor practice by discharging an employee “for failing to abide by union rules or policies with which he disagrees.”  Pattern Makers' League, 473 U.S. at 106, 105 S.Ct. at 3071;  see 29 U.S.C. § 158(a)(3).   Thus, under the NLRA, employees have an equal right to join or refrain from joining unions and engaging in collective bargaining.

 Accordingly, to accommodate federal labor policy, we must preserve the delicate balance of countervailing power that characterizes the process.   Injecting the Sherman Act into the collective bargaining process would disrupt this balance by giving unions a powerful new weapon, one not contemplated by the federal labor laws.   With such a weapon at their disposal, union workers could do what the plaintiff class has done here:  invoke the antitrust laws and their threat of treble damages to gain an advantage in bargaining over a salary provision about which union members do not care deeply enough to strike.   In sum, a proper respect for the national labor policy expressed in the NLRA and the LMRA requires us to recognize the nonstatutory labor exemption as a potential shield for all lawful actions taken by either unions or employers pursuant to the collective bargaining process.   Indeed, only the most crude accommodation of the federal labor and antitrust policies would shield union-employer agreements from Sherman Act liability, but leave exposed the lawful means employed in the process to reach those agreements.

In this regard, we find it telling that those judges and commentators who favor leaving certain aspects of the collective bargaining process unshielded by the exemption cannot agree on any point at which the exemption must expire in order to properly accommodate federal labor policy.   See, e.g., Powell v. NFL, 678 F.Supp. 777, 788 (D.Minn.1988) (“[P]roper accommodation of labor and antitrust interests requires that a labor exemption relating to a mandatory bargaining subject survive expiration of the collective bargaining agreement until the parties reach impasse as to that issue. ”), rev'd, 930 F.2d 1293 (8th Cir.1989);  Bridgeman v. National Basketball Ass'n, 675 F.Supp. 960, 967 (D.N.J.1987) (holding that nonstatutory labor exemption for restraint on competition contained in expired collective bargaining agreement survives impasse in negotiations over new agreement “only as long as the employer continues to impose that restriction unchanged, and reasonably believes that the practice or a close variant of it will be incorporated in the next collective bargaining agreement”);  Kieran M. Corcoran, When Does The Buzzer Sound?:  The Nonstatutory Labor Exemption In Professional Sports, 94 Colum.L.Rev. 1045, 1072 (1994) (suggesting that “the exemption should extend until it becomes clearly unreasonable for both of the parties to believe the particular provision will continue to exist in that form in a succeeding agreement”);  Ethan Lock, The Scope Of The Labor Exemption In Professional Sports, 1989 Duke L.J. 339, 400 (1989) (“[T]he proper accommodation of federal antitrust and labor law requires that the labor exemption expire simultaneously with the collective bargaining agreement.”).   The one common thread in this dizzying array of options is that not one of the proposals makes the slightest sense under established labor law principles.6

We also find it telling that the more recent decisions construing the nonstatutory labor exemption establish a clear trend in favor of shielding the collective bargaining process in its entirety.   Most notably, in Powell v. NFL, 930 F.2d at 1293, the Eighth Circuit applied the nonstatutory labor exemption to bar an antitrust attack on free agency restrictions contained in an expired collective bargaining agreement even after negotiations on a new agreement ended in impasse.   The Powell court concluded, as we do here, that

[t]he labor arena is one with well established rules which are intended to foster negotiated settlements rather than intervention by the courts.   The League and the Players have accepted this “level playing field” as the basis for their often tempestuous relationship, and we believe that there is substantial justification for requiring the parties to continue to fight on it, so that bargaining and the exertion of economic force may be used to bring about legitimate compromise.

Id. at 1303;  see also, e.g., Williams, 45 F.3d at 688 (holding that antitrust laws do not prohibit employers from bargaining jointly with a union and implementing their joint proposals in the absence of a collective bargaining agreement);  Wood, 809 F.2d at 959 (“[N]o one seriously contends that the antitrust laws may be used to subvert fundamental principles of our federal labor policy as set out in the [NLRA].”).

 Finally, we observe that the Players have offered no coherent theoretical framework in support of their view that the exemption expires with the collective bargaining agreement.   Under their view, a multi-employer bargaining unit could be held liable under the antitrust laws for adhering to the terms of an alleged restraint on competition embodied in an expired collective bargaining agreement prior to impasse, despite the fact that employers face an unfair labor practice charge if they alter the terms of agreement during that period.7  See Southwestern Steel & Supply, Inc., 806 F.2d at 1113.   At the same time, however, the Players submit that such employers could not similarly be held liable for locking out employees to gain an advantage in negotiations.   See Tr. of Oral Argument at 42.   We find this reasoning incomprehensible.   Both an employer's adherence to expired terms prior to impasse and its decision to lock out employees are unilateral but lawful aspects of the collective bargaining process established by the NLRA-as, for that matter, is an employer's unilateral implementation after impasse of a term encompassed within its pre-impasse proposals.

In an effort to find a theoretical framework to support the Players' position, the dissent relies heavily on a purported distinction between the unilateral implementation of employment terms after impasse and other bargaining “tactics” utilized by both employers and employees in the collective bargaining process.   Not only is such a distinction completely without support in the case law, but, as even the dissent itself acknowledges, such a view is patently contrary to Supreme Court precedent.   On this point, the Court in American Ship Building Co., 380 U.S. at 316, 85 S.Ct. at 966, made it clear that an employer's right to “institute unilaterally the working conditions which he desires” is among the lawful “tools of economic self-help” available to be utilized as a tactic in collective bargaining.   It is beyond the authority of this court to find otherwise.

C. The Sherman Act

 As the policy underlying the NLRA supports the principle that the nonstatutory labor exemption shields the entire collective bargaining process, so too the policy underlying the Sherman Act supports the principle that the case for applying the exemption is strongest where a restraint on competition operates primarily in the labor market.   The Sherman Act states simply that “[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal.”  15 U.S.C. § 1.   While the statutory language speaks broadly of every restraint of trade, the Supreme Court has long since limited its application to only those restraints deemed unreasonable.   See Standard Oil Co. v. United States, 221 U.S. 1, 62, 31 S.Ct. 502, 516, 55 L.Ed. 619 (1911) (stating that “the criteria to be resorted to in any given [Sherman Act] case for the purpose of ascertaining whether violations of the section have been committed, is the rule of reason”).   In an effort to give content to this reasonableness standard, courts have conceived of the Sherman Act as a “consumer welfare prescription.”  Reiter v. Sonotone Corp., 442 U.S. 330, 343, 99 S.Ct. 2326, 2333, 60 L.Ed.2d 931 (1979) (internal quotations omitted).   Thus, the Sherman Act's primary focus is the product market, not the labor market.   As one preeminent commentator has stated, “the antitrust laws are not concerned with competition among laborers or with bargains over the price or supply of labor-its compensation or hours of service or the selection and tenure of employees.”   Archibald Cox, Labor and the Antitrust Laws-A Preliminary Analysis, 104 U.Pa.L.Rev. 252, 255 (1955).   Indeed, in considering the standing of labor unions to bring antitrust claims, the Supreme Court has stated that “a union, in its capacity as bargaining representative, will frequently not be part of the class the Sherman Act was designed to protect, especially in disputes with employers with whom it bargains.”  Associated Gen. Contractors v. California State Council of Carpenters, 459 U.S. 519, 540, 103 S.Ct. 897, 909, 74 L.Ed.2d 723 (1983) (emphasis added).

 We recognize, of course, that, as a general matter, the antitrust laws may apply to restraints on competition in non-unionized labor markets.   See, e.g., Radovich v. NFL, 352 U.S. 445, 449-52, 77 S.Ct. 390, 392-94, 1 L.Ed.2d 456 (1957).   However, we think the inception of a collective bargaining relationship between employees and employers irrevocably alters the governing legal regime.   Once employees organize a union, federal labor law necessarily limits the rights of individual employees to enter into negotiations with their employer.   See NLRB v. Allis-Chalmers Mfg. Co., 388 U.S. 175, 180, 87 S.Ct. 2001, 2006, 18 L.Ed.2d 1123 (1967);  see also 29 U.S.C. § 159(a) (providing that representatives selected to bargain for employees in a collective bargaining unit “shall be the exclusive representatives of all the employees in such unit for the purposes of collective bargaining”).   Indeed, employers are positively prohibited from seeking to bargain with individual employees, absent consent from the union.   See J.I. Case Co. v. NLRB, 321 U.S. 332, 338-39, 64 S.Ct. 576, 580-81, 88 L.Ed. 762 (1944).   Moreover, employers may lawfully reduce competition in the labor market by forming multi-employer bargaining units, allowing for standardization of wage rates and working conditions within an industry.   See generally NLRB v. Truck Drivers Local Union No. 449, 353 U.S. 87, 94-96, 77 S.Ct. 643, 646-48, 1 L.Ed.2d 676 (1957).   Thus, once collective bargaining begins, the Sherman Act paradigm of a perfectly competitive market necessarily is replaced by the NLRA paradigm of organized negotiation-a paradigm that itself contemplates collusive activity on the parts of both employees and employers.   Stubborn adherence to antitrust principles in such a market can only result in “a wholesale subversion” of federal labor policy.  Wood, 809 F.2d at 959.

When this case is put in proper perspective, it is very difficult to take seriously the Players' claim that the action of the NFL violated the Sherman Act.   If the Players had been negotiating with a single football team (instead of the multi-employer group), there is no doubt whatsoever that that single employer lawfully could have taken unilateral action to impose a fixed salary for practice team players after bargaining in good faith to a point of impasse.   Indeed, the players do not assert otherwise, for they recognize that this is standard fare in labor law.   And nothing in the antitrust law is intended to proscribe such unilateral action by a single employer.   So the heart of the Players' position appears to be that the presence of a multi-employer bargaining unit should make a difference under the antitrust law.   This is a wholly untenable position, we think, for the reasons given in a compelling decision recently issued by the Second Circuit:

The existence of employer organizations that bargained with unions of employees pre-dated the passage of the Sherman Act in 1890․  Indeed, at that time these organizations existed at the national as well as local level․  Nevertheless, we find during the 104 succeeding years not a single instance, prior to this very case, of a union or an individual employee asserting the stark claim that the routine practices of multiemployer bargaining violate the antitrust laws.   It is true that recent antitrust challenges in the professional sports industry have at times involved facts very similar to those in the instant matter, see Powell, 930 F.2d at 1293;  Brown v. Pro Football, Inc., 782 F.Supp. 125 (D.D.C.1991);  Bridgeman v. National Basketball Ass'n, 675 F.Supp. 960 (D.N.J.1987), but the multiemployer bargaining issue appears to have been raised obliquely, if at all.

To the extent Congress has focused on the legality of multiemployer collective bargaining, it has always indicated its approval, or at least an assumption of legality.   In 1920, Congress sought in the Clayton Act to prevent federal courts from interfering in labor disputes.   Although Congress was largely concerned with the effect of such interference on unions, the statute was phrased in an evenhanded fashion to protect employer conduct in labor disputes as well as that of unions.   Section 20 thus exempted from federal prohibition “persons ․ terminating any relation of employment ․ or withholding ․ moneys or things of value,” language that would permit multiemployer lockouts.  29 U.S.C. § 52.   In 1932, Congress passed the Norris-LaGuardia Act, which barred federal courts from issuing injunctions against firms seeking to join, or remain members of, employer organizations, 29 U.S.C. § 104(b), and used language similar to that of the Clayton Act quoted supra, 29 U.S.C. § 104(c).   The definition of persons involved in a labor dispute also recognizes the existence of employer organizations.   See 29 U.S.C. § 113(b).  See also California State Council of Carpenters v. Associated Gen. Contractors, 107 L.R.R.M. 2724, 2725 (9th Cir.1981) (“[E]ven if the antitrust laws had been interpreted so as to bring multiemployer bargaining units within the scope of the Sherman Act, the statutory exemption found in section 4 of the Norris-LaGuardia Act, 29 U.S.C. § 104, when read together with section 20 of the Clayton Act, 29 U.S.C. § 52, clearly exempts ‘[b]ecoming or remaining a member ․ of any employer organization’ from the antitrust laws.  29 U.S.C. § 104(b).”  (footnote omitted)).   In 1947, Congress refused to limit multiemployer bargaining because it concluded that such bargaining “was a vital factor in the effectuation of the national policy of promoting labor peace through strengthened collective bargaining.”  [NLRB v. Truck Drivers Local Union No. 449 (“Buffalo Linen”), 353 U.S. 87, 95, 77 S.Ct. 643, 647, 1 L.Ed.2d 676 (1957) ] (discussing congressional debate over the Taft-Hartley amendments of 1947).

The lack of any antitrust challenge to, or congressional action restricting, multiemployer bargaining, for a century during which it prominently existed, grew, and flourished, strongly suggests some kind of general understanding about the legality of multiemployer bargaining that is fundamentally inconsistent with [the Players'] claim.

 *   *   *   *   *   *

We believe that the history recounted here strongly suggests that Congress never intended that the antitrust laws prohibit multiemployer bargaining with a common union.   Congress seemed to assume, and then to act on that assumption in 1947, see Buffalo Linen, 353 U.S. at 95-96, 77 S.Ct. at 647-48, that multiemployer bargaining was both efficient and a necessary counterweight to union power.   Such a belief arguably fits within Rule of Reason analysis that permits “ancillary restraints” necessary to a legitimate transaction.   See United States v. Addyston Pipe & Steel Co., 85 F. 271, 282 (6th Cir.1898), modified, 175 U.S. 211, 20 S.Ct. 96, 44 L.Ed. 136 (1899).   Whether as a court we would reach that result if writing on a clean slate is not the issue, however, because whatever doubt might have existed as to Congress' intent was entirely eliminated by the passage of federal labor laws.

 *   *   *   *   *   *

[M]ultiemployer bargaining has been a conspicuous feature of collective bargaining since the very formation of unions.   To hold at this late date that it-or the essence of practice under it-is illegal under the antitrust laws would cause a massive reshaping of the institution of collective bargaining․  The only basis for the claim asserted by [the Players] is that multiemployer bargaining is illegal.

 *   *   *   *   *   *

Buffalo Linen simply cannot be reconciled with [the Players'] antitrust claim.   The facts in that case plainly involved conduct that, if antitrust principles apply, were price-fixing and a joint boycott, per se violations of the Sherman Act.   However, the decision not only permitted the conduct but also stated that Congress had expressly considered multiemployer bargaining and had resolved to allow it subject only to the limitations of labor law as interpreted by the Labor Board.

Williams, 45 F.3d at 692.

 We therefore conclude that when federal labor policy collides with federal antitrust policy in a labor market organized around a collective bargaining relationship, antitrust policy must give way.   No one disputes that this conclusion is correct where a restraint on trade is embodied within a valid collective bargaining agreement.   However, we see no need for a different rule in cases in which a multi-employer bargaining unit lawfully implements a term rejected by the union after bargaining breaks down.   In such a case, “the gist of the antitrust theory is little more than that hard bargaining by employers with unions violates the Sherman Act.”   Jacobs & Winter, supra, at 27.

D. Application of the Exemption

 Based on our review of relevant Supreme Court precedent and the policies underlying both the NLRA and the Sherman Act, we conclude that injecting antitrust liability into the system for resolving disputes between unions and employers would both subvert national labor policy and exaggerate federal antitrust concerns.   Accordingly, we hold that the nonstatutory labor exemption waives antitrust liability for restraints on competition imposed through the collective bargaining process, so long as such restraints operate primarily in a labor market characterized by collective bargaining.   Our conclusion dictates that the exemption bars the plaintiff class in this case from challenging the fixed salary for Developmental Squad players under the antitrust laws.   The clubs and the NFL imposed the fixed salary only after negotiations over the issue reached impasse, and the fixed salary was encompassed within their pre-impasse proposals.   Thus, their action was a lawful part of the collective bargaining process established by the NLRA.   Further, the fixed salary operates as a restraint on competition in the labor market, specifically, the market for Developmental Squad player services, which is organized around the collective bargaining relationship between the NFL and the NFLPA.   Indeed, counsel for the plaintiff class at oral argument acknowledged that the fixed salary for Developmental Squad players had no effect on the product market.8

 In our view, the nonstatutory labor exemption requires employees involved in a labor dispute to choose whether to invoke the protections of the NLRA or the Sherman Act.   If employees wish to seek the protections of the Sherman Act, they may forego unionization or even decertify their unions.   We note that the NFL players took exactly this latter step after the Eighth Circuit's Powell decision.   See Releasing Superstars, supra, at 883 (describing NFLPA decertification after Powell ).   Decertification also occurs when newly organized unions are unable to negotiate their first contract.   We do not mean to encourage this practice, but we believe that employees, like all other economic actors, must make choices.   If they choose to avail themselves of the advantages of the collective bargaining process, their protections are as defined by the federal labor laws.   The system established by those statutes offers employees many benefits:  recognition of organized workers as a bargaining unit, thereby giving them bargaining strength;  establishment of mandatory subjects of bargaining;  protection of the right to strike;  allowance for the possibility of negotiated grievance procedures and pooled benefit plans;  and judicial enforcement of collective bargaining agreements.   Further, it establishes a significant list of employer actions that, if taken, constitute unfair labor practices for which employees and unions may seek redress before the NLRB.   However, under the system established by the federal labor laws, employees win concessions not by filing antitrust lawsuits, but with shrewd bargaining, favorable grievance settlements, victories in arbitration, and, when necessary, by striking.

 We recognize that the history of bargaining between the NFL and the NFLPA, which includes a failed strike by the players during the 1987 season, has prompted some commentators to conclude that “[t]he union cannot effectively strike.”   Ed Garvey, Foreword To The Scope Of The Labor Exemption In Professional Sports:  A Perspective On Collective Bargaining In The NFL, 1989 Duke L.J. 328, 337 (1989);  see also Lock, supra, at 354-59 (discussing factors creating “an inherently unequal bargaining relationship between players and owners” in professional football).   While we express no opinion on the NFLPA's ability to win the concessions it desires through organized activity, we reject the notion that an inequality of bargaining power in any industry justifies judicial intervention under the auspices of the Sherman Act.   Federal labor law guarantees a process, not any particular result.   Thus, it does not “contain a charter for the National Labor Relations Board”-or, we believe, for the courts-“to act at large in equalizing disparities of bargaining power between employer and union.”  Insurance Agents' Int'l Union, 361 U.S. at 490, 80 S.Ct. at 428.

In this regard, we think the recently resolved labor dispute in the National Hockey League illustrates that collective bargaining, instead of litigation, can operate as an effective means of union-employer dispute resolution in professional sports.   On January 11, 1995, after a 103-day lockout, professional hockey's players and owners agreed upon a new collective bargaining agreement.   Neither side received everything it wanted.   Club owners yielded on the issue of a player salary cap.   Players were required to accept limits on free agency.   See generally Len Hochberg, Last-Minute Deal Saves NHL Season, Wash.Post, Jan. 12, 1995, at A1.   Such is the nature of the collective bargaining process.   Most importantly, however, the league now is operating pursuant to terms endorsed by both sides-not terms dictated by a federal court.   While this model for dispute resolution may not be perfect, Congress has selected it to govern labor relations in this country, and it certainly is more balanced than the system that would prevail in professional football if its players were free to wield the threat of Sherman Act liability as a negotiating tool.

III. Conclusion

In sum, the clubs and the NFL are exempt from Sherman Act liability in this case.   Our holding does not mean that, freed of the threat of treble damages, employers will henceforth seek to force every set of negotiations with employees to impasse so that they may unilaterally implement the employment terms they desire.   Employers always face the threat of a strike for such actions.   Perhaps more importantly, they face the pressures toward settlement that inhere in an ongoing employer-employee relationship in which each side needs the other to accomplish its goals.

Initially it may be only fear of the economic consequences of disagreement that turns the parties to facts, reason, a sense of responsibility, a responsiveness to government and public opinion, and moral principle;  but in time these forces generate their own compulsions, and negotiating a contract approaches the ideal of informed persuasion.

Archibald Cox, The Duty To Bargain In Good Faith, 71 Harv.L.Rev. 1401, 1409 (1958).

Nor does our holding mean that the antitrust laws are completely without force where labor agreements are at issue.   Employers and unions certainly face liability when they conspire to use a collective bargaining agreement as an economic weapon against their competitors in the product market.   See, e.g., Pennington, 381 U.S. at 665-66, 85 S.Ct. at 1590-91;  Allen Bradley, 325 U.S. at 808, 65 S.Ct. at 1539.   However, where, as here, an alleged restraint on competition imposed through the collective bargaining process affects only the bargaining parties, and has no impact on the product market, the nonstatutory labor exemption shields those parties from antitrust liability.   Accordingly, the judgment of the District Court is reversed and the case is remanded for further proceedings consistent with this opinion.

So ordered.

Today the majority holds that the so-called “nonstatutory labor exemption” totally immunizes employers from antitrust liability when they unilaterally impose industry-wide terms of employment, in restraint of competition in the labor market, so long as the employers have previously engaged in unsuccessful collective bargaining with their employees over those issues.   In so holding, the majority decides that employees who exercise their statutory rights to bargain collectively automatically forfeit antitrust rights to which they otherwise would be entitled.   Thus, employees must now choose between foregoing collective bargaining altogether, thereby retaining antitrust protection against employer restraints on the labor market;  or engaging in collective bargaining at the risk of forfeiting all antitrust remedies if bargaining fails and the employers unilaterally foist unagreed-to industry-wide terms upon them.

The majority insists its ruling does no more than maintain a level playing field in employer-employee relations and carry out the congressional mandate favoring collective bargaining as the primary means of resolving labor disputes.   I do not think so.   The reality is that today's decision sharply tilts the playing field in employers' favor, and because of that, will erode the vitality of collective bargaining itself.   The rule announced today gives multiemployer groups both renewed incentives and unprecedented power to impose industry-wide terms of employment regardless of union consent and without fear of antitrust liability.   Multiemployer groups will enjoy new freedom to harden bargaining positions, secure that if the employees do not submit at the bargaining table, the employers may impose their negotiating proposals after bargaining has irretrievably stalled.   Employees in a weak bargaining position will likely opt in favor of clinging to antitrust protection as the less risky course of action;  these employees will henceforth have powerful incentives not to engage in collective bargaining at all, since any collective bargaining opens the door to employer imposition of labor market restraints if bargaining fails.   It could in extreme cases drive employees to decertify their unions, as the only guarantee against threats by multiemployer groups to unilaterally impose industry-wide wage caps and unacceptable working conditions.

All this suggests less, not more, collective bargaining and fewer, not more, successfully negotiated collective bargaining agreements.   The majority decision thus poses a threat to the central purpose of our labor laws of promoting collective bargaining as the primary vehicle for ensuring labor peace.   This result neither harmonizes labor law with antitrust law, nor comports with the congressional design of our statutory labor law regime.

I. Applicability of Antitrust Laws to Restraints on the Labor Market

Absent special statutory or judge-made exemption, a multiemployer agreement unilaterally imposing uniform industry-wide terms of employment, and thereby restraining competition in the labor market, runs afoul of the antitrust laws.1  Concededly, employer-imposed restraints on the labor market are most prominent in professional sports.   Employers in other fields generally favor open competition in the labor market, believing that, in the absence of chronic labor shortages, competition for jobs will generally hold down wages and produce higher quality and better motivated workers.   Historically, restraints on competition in the labor market have originated in employees' attempts to counter this downward pressure on wages through collective bargaining agreements fixing uniform wages and terms of employment.   In professional sports, however, the situation is reversed.   Because top-level athletic talent is scarce, competition tends to drive salaries up, at least for the top performers.   Thus professional athletes have sought to maximize competition among employers for their unique athletic skills.2  Predictably, the owners have reacted with multiemployer efforts to restrain competition for professional athletes' services.3  See, e.g., Flood v. Kuhn, 407 U.S. 258, 92 S.Ct. 2099, 32 L.Ed.2d 728 (1972) (baseball “reserve clause”);  Mackey v. NFL, 543 F.2d 606 (8th Cir.1976) (football “Rozelle rule” limiting player mobility);  Smith v. Pro-Football, 420 F.Supp. 738 (D.D.C.1976) (football player draft), aff'd in part & rev'd in part,593 F.2d 1173 (D.C.Cir.1978);  McCourt v. California Sports, Inc., 600 F.2d 1193 (6th Cir.1979) (hockey reserve system);  Bridgeman v. National Basketball Ass'n, 675 F.Supp. 960 (D.N.J.1987) (basketball draft, salary cap, and right of first refusal).

Nonetheless, it is important to remember that multiemployer labor market restraints are not entirely confined to professional sports, so that today's majority rule has implications beyond the football gridiron.   In Anderson v. Shipowners' Ass'n, 272 U.S. 359, 47 S.Ct. 125, 71 L.Ed. 298 (1926), the Supreme Court held that an industry-wide agreement setting uniform terms and conditions of employment for seamen was subject to antitrust scrutiny.   Lower courts have similarly recognized that the antitrust laws presumptively apply to employer-imposed restraints on labor markets.   See, e.g., Ostrofe v. H.S. Crocker Co., Inc., 740 F.2d 739, 742-43 (9th Cir.1984) (multiemployer agreement not to hire “whistleblowers”);  Hennessey v. NCAA, 564 F.2d 1136, 1147-51 (5th Cir.1977) (multicollege agreement to limit employment of athletic coaches);  Newburger, Loeb & Co., Inc. v. Gross, 563 F.2d 1057, 1082 (2d Cir.1977) (employee noncompetition clauses), cert. denied, 434 U.S. 1035, 98 S.Ct. 769, 54 L.Ed.2d 782 (1978);  Quinonez v. National Ass'n of Sec. Dealers, Inc., 540 F.2d 824, 827-28 (5th Cir.1976) (multiemployer blacklisting agreement);  Nichols v. Spencer International Press, 371 F.2d 332, 334 (7th Cir.1967) (multiemployer “non-switching” agreement restricting mobility of employees);  Cordova v. Bache & Co., 321 F.Supp. 600 (S.D.N.Y.1970) (multiemployer action to reduce compensation paid to securities brokers), later opinion sub nom.  Jacobi v. Bache & Co., 377 F.Supp. 86 (S.D.N.Y.1974), aff'd, 520 F.2d 1231 (2d Cir.1975), cert. denied, 423 U.S. 1053, 96 S.Ct. 784, 46 L.Ed.2d 642 (1976).   See also Gardella v. Chandler, 172 F.2d 402, 408 (2d Cir.1949) (L. Hand, J.) (“[W]hatever other conduct the [antitrust] Acts may forbid, they certainly forbid all restraints of trade which were unlawful at commonlaw, and one of the oldest and best established of these is a contract which unreasonably forbids anyone to practice his calling.”).

Admittedly, a few commentators have suggested that antitrust laws should not apply to restraints on labor markets.   See, e.g., Archibald Cox, Labor and the Antitrust Laws-A Preliminary Analysis, 104 U.Pa.L.Rev. 252, 254-55 (1954);  Milton Handler, Labor and Antitrust:  A Bit of History, 40 Antitrust L.J. 233, 235 (1971);  Theodore J. St. Antoine, Connell:  Antitrust Law at the Expense of Labor Law, 62 Va.L.Rev. 603, 606 (1976).   They argue that, despite the broad language of the antitrust statutes prohibiting every restraint on trade, the central purpose of the antitrust laws is not to promote competition per se, but to protect consumers from anticompetitive practices.   See generally Robert Bork, The Antitrust Paradox (1978).   Thus if an anticompetitive restraint affects only input markets, e.g., labor, and not output markets, e.g., products, it should not be subject to antitrust scrutiny.   See, e.g., Elinor R. Hoffmann, Labor and Antitrust Policy:  Drawing a Line of Demarcation, 50 Brooklyn L.Rev. 1 (1983) (tracing distinction between input and output markets to legislative history of Sherman Act).

It is, however, not possible to square this minority view with the development of our antitrust jurisprudence.   See II Phillip Areeda & Donald F. Turner, Antitrust Law ¶ 338b & 338c (1978) (employer agreements restricting the labor market are subject to antitrust liability);  James M. Altman, Antitrust:  A New Tool for Organized Labor?, 131 U.Pa.L.Rev. 127 (1982) (reviewing cases and common law antecedents, concluding that employer restraints on labor markets are subject to antitrust laws).   The Supreme Court has clearly held that antitrust laws apply not only to restraints on output markets, but to input markets as well, including both labor, Anderson, supra;  Radovich v. National Football League, 352 U.S. 445, 77 S.Ct. 390, 1 L.Ed.2d 456 (1957), and input commodities, e.g., Mandeville Island Farms, Inc. v. American Crystal Sugar Co., 334 U.S. 219, 236, 68 S.Ct. 996, 1006, 92 L.Ed. 1328 (1948) (sugar refiners' agreement to fix input prices paid for sugar beets violates Sherman Act, which “does not confine its protection to consumers, or to purchasers, or to competitors, or to sellers” but instead “is comprehensive in its terms and coverage, protecting all who are made victims of the forbidden practices by whomever they may be perpetrated”).   See also United States v. Socony-Vacuum Oil Co., 310 U.S. 150, 223, 60 S.Ct. 811, 844, 84 L.Ed. 1129 (1940) (Sherman Act prohibits any “combination formed for the purpose and with the effect of raising, depressing, fixing, pegging, or stabilizing the price of a commodity” and agreements to make “purchases [of inputs] at or under the market are one species of [unlawful] price-fixing”);   American Tobacco Co. v. United States, 328 U.S. 781, 66 S.Ct. 1125, 90 L.Ed. 1575 (1946) (applying antitrust law to restraints on input market for tobacco as well as output market for tobacco products);  Swift & Co. v. United States, 196 U.S. 375, 25 S.Ct. 276, 49 L.Ed. 518 (1905) (same, for livestock inputs and meat outputs).   Cf. National Macaroni Mfr's Ass'n v. Federal Trade Comm'n, 345 F.2d 421, 426 (7th Cir.1965) (purpose and effect of agreement to fix composition of macaroni product was to depress input price of semolina, thereby violating Sherman Act).

Economists too have long recognized that market inefficiencies created by anticompetitive restraints on input markets can be as destructive of a free market economy (and therefore ultimately damaging to consumers) as restraints on output markets.   See, e.g., Roger D. Blair & Jeffrey L. Harrison, Monopsony 36-43 (1993);  see also Mandeville Farms, 334 U.S. at 241-44, 68 S.Ct. at 1008-10 (describing how restraints on input market adversely affect output market).   While antitrust prosecutions for restraints on input markets are relatively rare, this is explained by the fact that restraints on input markets arise only in the unusual circumstance of an effective monopsony-a single purchaser, or a group of purchasers acting in concert.   And monopsony in turn arises only when the resource is uniquely valuable in its current use, so that even if the price is depressed by monopsony, sellers are unable to find alternative buyers.   Richard A. Posner & Frank H. Easterbrook, Antitrust 150 (1981) (monopsony can arise when “resources hav[e] substantially greater value in some uses than in others” so the “sole purchaser ․ will ․ be able to force the [seller] ․ to accept a monopsony price”);  Blair & Harrison, supra, at 43-44 (inelasticity of supply facilitates collusive monopsony).   Thus, for example, in the Mandeville Island Farms case, the sugar refiners' monopsonistic price-fixing scheme was effective because growers could not easily find other buyers or profitably switch to other crops when refiners conspired to fix the price of sugar beets.   See 334 U.S. at 240-42, 68 S.Ct. at 1008-09.   Nonetheless, according to the economists, there is a dead-weight loss associated with imposition of monopsony pricing restraints.   Some producers will either produce less or cease production altogether, resulting in less-than-optimal output of the product or service, and over the long run higher consumer prices, reduced product quality, or substitution of less efficient alternative products.   See, e.g., Herbert Hovenkamp, Economics and Federal Antitrust Law § 1.2, at 17-18 (1985);  Blair & Harrison, supra, at 42-43, 72.   So, even proceeding from the premise that antitrust laws aim only at protecting consumers, monopsonies fall under antitrust purview because monopsonistic practices will eventually adversely affect consumers.

Athletic prowess is, of course, a unique and highly specialized resource, of precisely the genre vulnerable to monopsony manipulation.   With a few notable exceptions, athletes typically excel in a single sport, and their labor has greater market value in that sport than in any other profession.   If team owners join together to suppress the price of athletic services through monopsony practices, most athletes will not be able to switch profitably to other lines of work.   Thus, the labor market for professional athletes' services is one of a very few areas where there is real potential for anticompetitive monopsonistic practices.   Blair & Harrison, supra, at 72.   Economic theory tells us, however, that monopsony will diminish output over time.   Because some talented athletes will switch to other sports or other professions, or decide not to enter the field, the overall quality of athletic performance in the sport will then decline, to the detriment of consumers.   Thus, both legal precedent and economic theory instruct that-absent some special exemption-antitrust principles do and should apply to such monopsonistic practices.   As the majority acknowledges, Maj. op. at 103, professional athletes presumptively have antitrust protection against monopsonistic restraints on the market for their services,4 as long as they forego collective bargaining with their employers.

In sum, antitrust law has a significant role to play as an antidote to anticompetitive restraints on labor markets.   That role should be ousted only as necessary to implement clear congressional directives.

II. The Statutory and Nonstatutory Exemptions

Our next inquiry is whether any special exemption-statutory or judge-made-applies to immunize labor market restraints from expected antitrust scrutiny.   In resolving this question, we are informed by the familiar nostrum that “exemptions from the antitrust laws are to be narrowly construed,” Group Life & Health Insurance Co. v. Royal Drug Co., 440 U.S. 205, 231, 99 S.Ct. 1067, 1083, 59 L.Ed.2d 261 (1979).   This principle applies equally to express and implied exemptions.  Id.  Implied exemptions to the “fundamental national policies embodied in the antitrust laws,” Federal Maritime Comm'n v. Seatrain Lines, Inc., 411 U.S. 726, 733, 93 S.Ct. 1773, 1779, 36 L.Ed.2d 620 (1973), in particular, should be found “only if necessary to make the [conflicting statutory scheme] work, and even then only to the minimum extent necessary,” Silver v. New York Stock Exchange, 373 U.S. 341, 357, 83 S.Ct. 1246, 1257, 10 L.Ed.2d 389 (1963).

The so-called statutory labor exemption is clearly inapplicable here.   The “statutory exemption” is embraced in two enactments:  sections 6 and 20 of the Clayton Act, 15 U.S.C. § 17 and 29 U.S.C. § 52, enacted in 1914, exempting unions from coverage of the Sherman Act and prohibiting the issuance of injunctions against specified labor activities;  and sections 1 through 15 of the Norris-LaGuardia Act, 29 U.S.C. §§ 101-115, enacted in 1932, barring federal courts from enjoining certain other specified union activities.  “These statutes declare that labor unions are not combinations or conspiracies in restraint of trade, and exempt specific union activities, including secondary picketing and boycotts, from the operation of the antitrust laws.”  Connell Constr. Co., Inc. v. Plumbers & Steamfitters Local No. 100, 421 U.S. 616, 621-22, 95 S.Ct. 1830, 1835, 44 L.Ed.2d 418 (1975).   The statutory labor exemption thus shields unions and specified union activities.   H.A. Artists & Assocs. v. Actors' Equity Ass'n, 451 U.S. 704, 713-15, 101 S.Ct. 2102, 2108-09, 68 L.Ed.2d 558 (1981).   Section 4(b) of the Norris-LaGuardia Act, however, does bar injunctions against “any person or persons participating in or interested in [a labor] dispute ․ [b]ecoming or remaining a member of ․ any employer organization․”  29 U.S.C. § 104(b) (emphasis added).   Thus, the statutory exemption appears to immunize membership in a multiemployer bargaining unit.   See California State Council of Carpenters v. Associated Gen'l Contractors, 648 F.2d 527, 544-45 (9th Cir.1980), rev'd on other grounds, 459 U.S. 519, 103 S.Ct. 897, 74 L.Ed.2d 723 (1983).   But beyond bare membership, the Clayton and Norris-LaGuardia Acts do not identify any protected class of multiemployer conduct.   This is in marked contrast to their detailed specifications of protected union and employee conduct.  Id. at 534-36.   The multiemployer conduct at issue in this case therefore falls beyond the narrow scope of any statutory labor exemption.

The only other possible source of exemption from the antitrust statutes is the nonstatutory labor exemption, a judicially-created doctrine that attempts to harmonize the potentially inconsistent requirements of labor law and antitrust law.  “The nonstatutory exemption has its source in the strong labor policy favoring the association of employees to eliminate competition over wages and working conditions.”  Connell, 421 U.S. at 622, 95 S.Ct. at 1835.   Because the exercise of labor law rights and duties-generally encouraging collective action-may be inconsistent with prohibitions on concerted action of any kind under the antitrust laws, courts have carved out an implied exemption to antitrust coverage.   Courts presume that Congress implicitly intended to exempt from antitrust scrutiny those concerted activities explicitly authorized by labor laws that are essential to collective bargaining.   The prime example is the ordinary collective bargaining agreement.   Although no statutory exemption shields such agreements from antitrust scrutiny, the collective bargaining scheme established by the national labor laws would be mortally undercut if collective bargaining agreements were subject to collateral attack under the antitrust laws as “combinations in restraint of trade.”   See Local No. 189, Amalgamated Meat Cutters & Butcher Workmen v. Jewel Tea Co., Inc., 381 U.S. 676, 689, 85 S.Ct. 1596, 1601-02, 14 L.Ed.2d 640 (1965) (Opinion of White, J.) (“Employers and unions are required to bargain about wages, hours, and working conditions, and this fact weighs heavily in favor of antitrust exemption for agreements on these subjects.”);  id. 381 U.S. at 711-12 (Opinion of Goldberg, J.) (“To tell the parties that they must bargain about a point but may be subject to antitrust penalties if they reach an agreement is to stultify the congressional scheme.”).

The few cases in which the Supreme Court has addressed the nonstatutory labor exemption involved employer-employee agreements which also had discernible anticompetitive effects on product output markets.   See United Mine Workers v. Pennington, 381 U.S. 657, 85 S.Ct. 1585, 14 L.Ed.2d 626 (1965) (exemption does not apply to agreement to impose uniform industry-wide terms on employers not party to the agreement);  Jewel Tea, 381 U.S. at 676 (exemption applies to agreement limiting marketing hours despite ancillary anticompetitive effect on product market, where restraint is tailored to protect direct interests of bargaining unit members);  Connell, 421 U.S. at 635, 95 S.Ct. at 1841 (exemption does not apply to agreement barring subcontracting to nonunion subcontractors).   As the majority accurately notes, several of these cases held the nonstatutory exemption inapplicable when its primary purpose or effect was to restrain product output markets.   Maj. op. at 1050-51.   But nothing in these cases supports the majority's reverse contention that antitrust law is totally unconcerned with labor markets once collective bargaining has been undertaken but failed to resolve the dispute.   In truth, Supreme Court precedent is wholly inconclusive as to whether the nonstatutory labor exemption embraces anything like the unilateral employer-imposed labor market restraints at issue here.   At best, the relevant Supreme Court cases establish only that there is a nonstatutory labor exemption, and that its umbrella protects at least some employer-employee agreements, while others, i.e., those whose primary purpose or effect is to restrain product output markets, are not immune.

Two sister circuits have recently adopted expanded interpretations of the nonstatutory labor exemption to cover post-impasse situations similar to the one here.   See National Basketball Association v. Williams, 45 F.3d 684 (2d Cir.1995);  Wood v. National Basketball Ass'n, 809 F.2d 954 (2d Cir.1987);  Powell v. National Football League, 930 F.2d 1293 (8th Cir.1989), cert. denied, 498 U.S. 1040, 111 S.Ct. 711, 112 L.Ed.2d 700 (1991).   But other courts have refused to extend it beyond employer-employee agreements.   See e.g., McCourt, 600 F.2d at 1197-98;  Bridgeman, 675 F.Supp. at 964-65;  Smith, 420 F.Supp. at 741-42.

Precedent is of limited assistance, then, in deciding whether to extend the nonstatutory exemption doctrine to the situation at hand.   Striking the appropriate balance requires us to go further and examine the underlying policies and principles of both labor law and antitrust law, effecting “a proper accommodation between the congressional policy favoring collective bargaining under the NLRA and the congressional policy favoring free competition in business markets,” Connell, 421 U.S. at 622, 95 S.Ct. at 1835.

The majority presses a deceptively simple solution:  whenever antitrust coverage potentially overlaps with activities authorized by labor law, and only the labor market is affected by the challenged restraint, antitrust law must give way.   The majority's is a bright-line rule, no mistake.   Additionally, it keeps federal courts out of messy labor-management disputes.   There is much to be said for the twin virtues of economy and elegance, but they may be purchased at too high a price-in this case the substantial interests of players and fans in guarding against uncontrolled labor market restraints.   Over the long haul, the majority's approach will not serve the interests of labor law by promoting collective bargaining.   And it certainly will not comport with our judicial duty to carve out only the narrowest antitrust exemption compatible with the effective implementation of the labor laws.

III. Maintaining the Employer-Employee Balance

The majority insists its rule granting employers an antitrust exemption for terms unilaterally imposed after a bargaining impasse is required so that employers can engage in hard bargaining as permitted by the labor laws, and thereby “preserve the delicate balance of countervailing power,” Maj. op. at 100, that is necessary to keep unions from gaining the upper hand.   The majority rightly points out that the labor laws contemplate a system of countervailing economic pressure between employers and employees, under which both sides have incentives to bargain.   See Julius G. Getman, The Protection of Economic Pressure by Section 7 of the NLRA, 115 U.Pa.L.Rev. 1195, 1196 (1967).   But we must also be mindful that “a primary purpose of the National Labor Relations Act was to redress the perceived imbalance of economic power between labor and management ․ by conferring certain affirmative rights on employees and by placing certain enumerated restrictions on the activities of employers.”  American Ship Bldg. Co. v. NLRB, 380 U.S. 300, 316, 85 S.Ct. 955, 966, 13 L.Ed.2d 855 (1965) (emphasis added).   Once Congress established this scheme, it contemplated that neither the courts nor the National Labor Relations Board would intervene on a case-by-case basis to adjust the parties' relative bargaining power, but rather leave them to fight it out at the bargaining table.  American Ship Bldg., 380 U.S. at 317-18, 85 S.Ct. at 966-67;  NLRB v. Insurance Agents Int'l Union, 361 U.S. 477, 497-500, 80 S.Ct. 419, 431-33, 4 L.Ed.2d 454 (1960).   Of course, Congress itself from time to time adjusts the balance, as it did in enacting the Taft-Hartley Act of 1947 and the Landrum-Griffin Act of 1957, further limiting union conduct.   Insurance Agents, 361 U.S. at 500, 80 S.Ct. at 433.

In my view, the majority's rule does much more than “preserve the delicate balance” in labor-management relations already established by labor law.   It decisively tips the balance by giving employers important new rights, together with new incentives to harden their bargaining positions.   Because employers can now unilaterally impose labor market restraints that heretofore might have exposed them to antitrust liability,5 the majority's rule provides new incentives to raise the stakes in bargaining in hope of winning industry-wide labor market restraints that might never before have been attempted.   It also encourages employer groups to hang tough on their demands, since at impasse they may be able to impose any bargaining proposals the employees have rejected without fear of antitrust liability.   The majority's rule does not just protect hard bargaining;  it positively encourages it.   It makes impasse more likely and successful collective bargaining less likely.6

On the employees' side, the majority's rule also creates new incentives, none of them helpful to the bargaining process.   First, some employees-especially those in a weak bargaining position-are under pressure not to enter into collective bargaining at all, lest the existence of a bargaining relationship license unilateral employer imposition of anticompetitive terms.   See Lee Goldman, The Labor Exemption to the Antitrust Laws as Applied to Employers' Labor Market Restraints in Sports and Non-sports Markets, 1989 Utah L.Rev. 617, 658-59 (describing disincentives to collective bargaining if unilaterally-imposed terms are exempt from antitrust coverage).   Second, as the majority recognizes, Maj. op. at 105, some employees will be encouraged to decertify their unions rather than risk unilateral multiemployer imposition of terms at impasse.   See also Goldman, supra, at 658-59.   Such a consequence is not mere speculation, but represents the actual upshot of the Eighth Circuit's decision in Powell upholding a broad nonstatutory labor exemption like that of the majority here.   Faced with anticompetitive industry-wide terms forced upon them without their consent, the players in Powell promptly responded by terminating union representation.   See Powell v. National Football League, 764 F.Supp. 1351 (D.Minn.1991) (recognizing termination of nonstatutory labor exemption when NFL players voted to end union representation in response to Powell decision);  The Five Smiths, Inc., et al. v. National Football League Players Ass'n, No. Civ. 3-90-177 (D.Minn., filed Mar. 30, 1990) (suit by NFL seeking judicial invalidation of players' union decertification).   New incentives for employees not to engage in collective bargaining-and the bizarre prospect of employers attempting to force employees to remain in a union so as to preserve the employers' valuable antitrust exemption-run directly contrary to the overarching purpose of the labor laws to encourage bona fide collective bargaining.

The majority seems to suggest that because the labor laws mandate a process and not an agreement, Maj. op. at 99, 100, we should not be concerned if the rule ends up encouraging more nonagreements than agreements.   I do not see how they can be right.  “The basic theme of ․ [the National Labor Relations] Act was that through collective bargaining the passions, arguments, and struggles of prior years would be channeled into constructive, open discussions leading, it was hoped, to mutual agreement.”  H.K. Porter Co. v. NLRB, 397 U.S. 99, 103, 90 S.Ct. 821, 823, 25 L.Ed.2d 146 (1970).   A rule that ousts antitrust law only to encourage fewer collective bargaining agreements seems contrary to the purposes of both the labor laws and the antitrust laws.

The majority's rule also creates new assymetries in bargaining relationships still in effect.   Employers, now shielded from antitrust liability, can unilaterally impose terms restraining the labor market, while employees have no such capability because unions do not hold sufficient economic power to unilaterally force terms on unwilling employers.   See Archibald Cox, The Duty to Bargain in Good Faith, 71 Harv.L.Rev. 1401, 1424 (1958).   Prior to today's ruling, employers and employees could jointly agree to restraints on the labor market through collective bargaining agreements, and those restraints would be free from antitrust scrutiny.   In that limited sense, employees had the right to bargain away (or waive) their antitrust protection, but always with the prospect of getting something in return.   Employers could also acquire antitrust immunity for labor market restraints through the quid pro quo of collective bargaining.   But today's ruling “effectively gives the party benefitting from a potentially anticompetitive restraint [i.e., employers] ․ the benefit of the antitrust exemption without having to pay for it” through concessions at the bargaining table, Brief of the United States as Amicus Curiae Opposing Certiorari in NFL v. Powell (“Amicus br.”), at 15.   The symmetry, mutuality, and give-and-take that lie at the heart of the bargaining process are irretrievably lost.

Under the majority's rule it is clear that a multiemployer agreement setting industry-wide uniform terms of employment in a single industry-precisely what was prohibited in Anderson v. Shipowners' Ass'n -would be exempt from antitrust scrutiny, provided the employers first proposed the scheme in collective bargaining and implemented it after impasse.   Until today, the shipowners could immunize those industry-wide terms from antitrust scrutiny only through a collective bargaining agreement;  after today, the shipowners need give nothing up in order to get around the antitrust roadblock.   Rather than being in a position to demand a quid pro quo for their assent to the employers' proposed terms, the employees must now either strike or make concessions to prevent the employers from doing what the antitrust laws would have previously deterred them from doing.   The employers thus enter collective bargaining in a stronger position, and the employees enter in a commensurately weaker position.   Although strong unions may still be able to successfully fend off employers' unilateral imposition of terms, even the strongest unions are left in a weaker bargaining position after today's ruling.   See Goldman, supra, at 661.

In the end, the wide open field given employers by so broad an exemption from antitrust laws cannot be justified by any rationale based on the requirements of collective bargaining.

IV. “Terms” Versus “Tactics”

The majority's analysis goes astray, I believe, by relying on a too-facile equation of terms of employment that restrain the labor market and are unilaterally imposed by employers at impasse, with “bargaining tactics,” “hard bargaining,” and the “economic weapons” used by each side in jockeying for position at the bargaining table.   Unilaterally-imposed terms are not mere bargaining tactics.   See Goldman, supra, at 682-83 (distinguishing bargaining “tactics” from unilaterally-imposed “terms” and urging nonstatutory exemption for “tactics” but not for “terms”);  Leonard L. Scheinholtz & Kenneth C. Kettering, Exemption Under the Antitrust Laws for Joint Employer Activity, 21 Duquesne L.Rev. 347, 354 (1983) (distinguishing multiemployer bargaining “tactics” from “terms”).   It is true, of course, that “the use of economic pressure by the parties to a labor dispute is ․ part and parcel of the process of collective bargaining,” Insurance Agents', 361 U.S. at 495, 80 S.Ct. at 430, and the Supreme Court itself has suggested that post-impasse unilateral imposition of terms may qualify as an economic pressure tactic, see American Ship Bldg., 380 U.S. at 316, 85 S.Ct. at 966 (dicta).   Still, the statutory structure of the National Labor Relations Act and the logic of collective bargaining suggest that an employer's right to unilaterally impose terms after impasse is best understood not as a “bargaining tactic” but as part of the employer's residual right to continue operating as dictated by business necessity once her statutory duty to bargain has been exhausted, see NLRB v. Tex-Tan, Inc., 318 F.2d 472, 480-81 & n. 20 (5th Cir.1963).   Prior to enactment of the NLRA (and even today outside the collective bargaining context), employers remained free to unilaterally impose new terms of employment at any time;  but those unilaterally-imposed terms were subject to antitrust scrutiny.   The NLRA, however, restricts an employer's freedom to unilaterally impose terms.   The employer's central duty under § 8(a)(5) of the NLRA is to bargain in good faith over mandatory subjects of bargaining.   See Milton Handler & William C. Zifchak, Collective Bargaining and the Antitrust Laws:  The Emasculation of the Labor Exemption, 81 Colum.L.Rev. 459, 500 (1981).   To give substance to that obligation, § 8(a)(5) has been interpreted to prohibit an employer from unilaterally imposing at impasse new terms not previously submitted to collective bargaining and rejected by the union, for to do so “is necessarily inconsistent with a sincere desire to conclude a [ ] [collective bargaining] agreement,” NLRB v. Katz, 369 U.S. 736, 745, 82 S.Ct. 1107, 1112, 8 L.Ed.2d 230 (1962).   Moreover, because unilateral imposition of new terms during bargaining “tends to subvert the union's position as the representative of the employees,” Insurance Agents', 361 U.S. at 485, 80 S.Ct. at 425, the employer must observe the status quo while bargaining continues.   But after impasse, the employer is free to unilaterally impose terms reasonably encompassed in bargaining proposals already rejected by the union, Katz, 369 U.S. at 745, 82 S.Ct. at 1112, because at that point the employer has exhausted its statutory duty to bargain, Taft Broadcasting, 163 N.L.R.B. 475, 478 (1967), (“[A]fter bargaining to impasse, that is, after good-faith negotiations have exhausted the prospects of concluding an agreement, an employer does not violate the Act by making unilateral changes that are reasonably comprehended within his pre-impasse proposals”) (emphasis added), petition for review denied sub. nom.  American Fed'n of Television & Radio Artists v. NLRB, 395 F.2d 622 (D.C.Cir.1968) (“AFTRA ”);  AFTRA, 395 F.2d at 628;  NLRB v. McClatchy Newspapers, Inc., 964 F.2d 1153, 1157 (D.C.Cir.1992) (at impasse, “the duty to bargain is at least temporarily suspended, and the parties, typically the employer, may enact any change in a mandatory subject [of bargaining] reasonably contained within its final proposal”).

Why then is it necessary, or even helpful in advancing the collective bargaining process, to grant additional antitrust immunity to terms unilaterally imposed by the employer as part of his residual right to conduct his business, after he has been relieved of his statutory duty to bargain?   I should think that the antitrust interest in unilaterally-imposed industry-wide terms remains as great after impasse, when the employer's duty to bargain over those terms has ceased under the labor laws, as at any other time.   Congress could, of course, assign some new, significant labor law role to an employer's right to unilaterally impose terms at impasse, such that the terms must be immune from antitrust scrutiny.   But the majority can point to nothing in the statute or in its legislative history to support any such intention.   See Amicus br. at 13.  (“The text of the NLRA does not mention the antitrust laws, and the legislative history demonstrates no intent to restrict the preexisting rights of workers as the price of participation in a collective bargaining relationship․”).  In short, there is simply no fundamental conflict between the well-established interest of the antitrust laws in preventing employers from unilaterally imposing restraints on the labor market through anticompetitive terms, and the interest of the labor laws in promoting collective bargaining.

Despite the majority's rhetoric, terms of employment unilaterally imposed at impasse bear little resemblance to “bargaining tactics,” even extreme ones like strikes and lockouts.7  Unlike terms of employment, strikes and lockouts have instrumental value only within the bargaining context, as tools to put pressure on the other bargaining party to reach an agreement.   Strikes and lockouts make sense only as long as hope exists that an agreement may be reached.   Unilaterally-imposed terms, however, go well beyond that purpose.   Although occasionally displaying ancillary “tactical” significance, unilaterally-imposed terms are inescapably nontactical in nature-they are simply the way the employer decides to run his business in the absence of any bargaining agreement.   They are not bargaining tactics but the end-product of an unsuccessful bargaining process, addressed not to the other bargaining party (the union) but imposed upon the employees because the employer believes they represent good business judgment.   Unilaterally-imposed terms, then, are a substitute for an agreement, not a means of reaching one.

Contrary to the majority's contention, Maj. op. at 101-102, the Supreme Court has repeatedly recognized this fundamental distinction between terms of employment-“the end result of bargaining”-and the “bargaining process” itself, Metropolitan Life Ins. Co. v. Massachusetts, 471 U.S. 724, 752, 105 S.Ct. 2380, 2395, 85 L.Ed.2d 728 (1985).   In the analogous context of implied preemption by the NLRA of state laws setting minimum labor standards, the Court has held that statutes setting substantive limits on terms of employment-including terms unilaterally imposed at impasse-do not impermissibly intrude on an employers' “lawful ‘tools of economic self-help’ available to be utilized as a tactic in collective bargaining,” Maj. op. at 102.   Earlier cases had established that state laws interfering with the exercise of self-help “economic weapons” in a labor dispute were preempted by federal labor law because they would “frustrate effective implementation of the [National Labor Relations] Act's processes,” Machinists v. Wisconsin Employment Relations Comm'n, 427 U.S. 132, 148, 96 S.Ct. 2548, 2557, 49 L.Ed.2d 396 (1976) (quotation and citation omitted);  see also Golden State Transit Corp. v. City of Los Angeles, 475 U.S. 608, 106 S.Ct. 1395, 89 L.Ed.2d 616 (1986) (citing cases).   In Metropolitan Life, employers advanced the argument-similar to the majority's conclusion here-that this implied preemption also extended to state laws establishing minimum terms of employment, on the theory that the NLRA allows employers and unions to settle terms through free exercise of the “self-help weapon[s]” that form the “balance of power between labor and management expressed in our national labor policy.”  Id. 471 U.S. at 750-51, 105 S.Ct. at 2394-95 (citation and quotation omitted).   The Court roundly rejected the employers' contention that “Congress' ultimate concern in the NLRA was in leaving the parties free to reach agreement about contract terms.”  Id. at 752, 105 S.Ct. at 2395.   Despite the “surface plausibility to appellants' argument” that statutory minimum terms interfere with the bargaining process, id., the Court said that “[t]he evil Congress was addressing [in enacting the NLRA] ․ was entirely unrelated to local or federal regulation establishing minimum terms of employment,” and there is “[n]o incompatibility” between the purpose of NLRA to “restore the equality of bargaining power,” and “state or federal legislation that imposes minimal substantive requirements on contract terms ․”  Id. at 754, 105 S.Ct. at 2396 (emphasis added).

In a subsequent case, Fort Halifax Packing Co. v. Coyne, 482 U.S. 1, 107 S.Ct. 2211, 96 L.Ed.2d 1 (1987), the Court held that a state statute setting minimum terms in the absence of a collective bargaining agreement was not preempted by the NLRA.   In Fort Halifax, a Maine statute provided that employers and employees could alter severance pay terms from the statutory minimum through collective bargaining agreements;  but once bargaining reached impasse, the statutory minimum terms applied.   Employers again complained that the statute interfered with the collective bargaining process (including presumably their right to impose terms unilaterally at impasse), but the Court held that the statute “does not impermissibly intrude upon the collective bargaining process.” 8  Id. at 23, 107 S.Ct. at 2223.   As the Court explained:

Both employers and employees come to the bargaining table with rights ․ that form a “backdrop” for their negotiations (citation omitted).   Absent a collective-bargaining agreement, for instance, state common law generally permits the employer to run the workplace as it wishes․  The parties may enter negotiations designed to alter this state of affairs, but, if impasse is reached, the employer may rely on pre-existing state law to justify its authority to make employment decisions;  that same state law defines the rights and duties of employees․  If impasse is reached, ․ preexisting state law determines the right[s] of employees․

Id. at 21, 107 S.Ct. at 2222.

The principles of Metropolitan Life and Fort Halifax apply here.   Federal antitrust law forms a “backdrop” against which collective bargaining takes place.   The nonstatutory labor exemption allows employers and employees to alter these standards through collective bargaining agreements;  but at impasse, preexisting antitrust rights apply.   Here, just as in Metropolitan Life and Fort Halifax, the substantive limitations placed on terms of employment do not impair the collective bargaining process, even though they may limit employers' ability to impose the terms of their choice unilaterally at impasse.

In some cases, we are told, impasse “ ‘is eventually broken, either through a change of mind or the application of economic force,’ ” and impasse “may be ‘brought about intentionally by one or both parties as a device to further, rather than destroy, the bargaining process.’ ”  Bonanno Linen, 454 U.S. at 412, 102 S.Ct. at 725 (quoting Charles D. Bonanno Linen Serv., Inc., 243 N.L.R.B. 1093, 1093-94).   But even if unilateral imposition of terms at impasse may sometimes have tactical significance, the majority's rule sweeps much too broadly, immunizing unilaterally-imposed terms whether or not they are used as a bargaining tactic;  whether or not they break an impasse or otherwise promote further bargaining;  and regardless of how identical terms, imposed for identical reasons, would fare under the antitrust laws in the absence of failed collective bargaining.

The present case offers a striking illustration of how employers may now execute an end run around the antitrust laws.   Here, there is no indication that the employers' unilateral imposition of a fixed salary of $1,000 per week on rookie and first-year developmental squad players was a “tactic” or “economic weapon” aimed at bringing an intransigent union back to the bargaining table.   The employers frankly admitted they wanted a $1,000 fixed salary for developmental squad players to save money, and because they could not enforce existing prohibitions against “stashing” players on the injured reserve list.   Thus their primary purpose for unilaterally imposing the fixed salary had nothing to do with tactical advantages in the bargaining process or pressuring the union to resume bargaining.   They basically thought the new terms were a better way to run their business.   The purpose, effect, and content of their unilaterally-imposed terms would have been the same had there been no union in the picture at all.

V. Striking the Balance Between Antitrust and Labor Law

The majority suggests that without the rule adopted today, employees will rely on antitrust litigation to secure gains they could not win at the bargaining table but were unwilling to strike over.   Maj. op. at 100-101.   That argument is a nonstarter.

First, the majority's rule applies whether or not the employees strike.   Second, denying employers antitrust exemption for unilaterally-imposed terms does not place any new offensive plays in the employees' playbook.   Rather, it affords them a defense against unilateral employer actions that offend antitrust principles.   Preserving antitrust protection does not guarantee unions a win at impasse because multiemployer bargaining units still have many other options.9  They can maintain the status quo, and employ economic pressure tactics such as joint lockouts, Buffalo Linen, 353 U.S. at 97, 77 S.Ct. at 648, or hiring temporary replacements for strikers, NLRB v. Brown, 380 U.S. 278, 284, 85 S.Ct. 980, 984, 13 L.Ed.2d 839 (1965).   A multiemployer group can be disbanded (by mutual consent with the union) upon the failure of bargaining, and each employer can resume bargaining separately.   Although generally not free to unilaterally withdraw from the multiemployer bargaining unit absent “unusual circumstances,” individual employers are still free to negotiate interim agreements with the union, Bonanno Linen, 454 U.S. at 413, 102 S.Ct. at 725.   Finally, even if the multiemployer group does proceed to unilaterally impose anticompetitive terms, thereby courting an antitrust lawsuit, the outcome of that litigation is far from a foregone conclusion under a rule-of-reason balancing.   Under any of these scenarios, employees will be hard-pressed to gain any easy victory.   Indeed, the range of choices left to employers without post-impasse antitrust immunity more closely tracks the “delicate balance” of the collective bargaining process as we now know it, than any wholesale exemption such as that announced by the majority today.

For all these reasons, I would hold ultimately, in accord with the district court, that terms of employment unilaterally imposed by employers after impasse are not exempt from antitrust scrutiny under the nonstatutory labor exemption.10  Under our rule, the symmetry and mutuality of the bargaining process are preserved.   Either side may propose and bargain for terms restraining the labor market, but can win those restraints only at the bargaining table by making concessions and securing agreement from the opposing side.   Incentives to bargain remain.   But at the point of impasse-when an agreement is no longer in sight or even being sought-immunity from antitrust liability for terms employers unilaterally impose should terminate.   Extending antitrust immunity undercuts the substantial interests of antitrust law in protecting the freedom of the labor market, and serves no ascertainable purpose within the collective bargaining framework.

This outcome, I believe, accomplishes a superior balancing of the competing interests of antitrust and labor policy.   Rather than casting out antitrust principles whenever any aspect of labor law enters the scene, it preserves antitrust protection alongside the collective bargaining process in situations where they do not fundamentally conflict.   Cf. Silver v. New York Stock Exchange, 373 U.S. at 357, 83 S.Ct. at 1257 (“the proper approach [to resolving conflict between antitrust laws and a competing statutory scheme] ․ is an analysis which reconciles the operation of both statutory schemes with one another rather than holding one completely ousted”).   And it better comports with our duty to construe implied antitrust exemptions narrowly;  the majority's global interpretation of the nonstatutory exemption goes far beyond what is necessary to make the collective bargaining process work.   Cf. id.   Finally, because it preserves the symmetry, mutuality, and balance of the collective bargaining process as crafted by Congress over the past 60 years, it more faithfully follows the design of our labor laws as well.

With time running out in a bitterly fought scoreless tie, this court would allow the owners an unearned critical fifth down.

I respectfully dissent.


June 12, 1995

Appellees' Suggestion For Rehearing In Banc and the brief of the United States in support thereof have been circulated to the full court.   The taking of a vote was requested.   Thereafter, a majority of the judges of the court in regular, active service did not vote in favor of the suggestion.   Upon consideration of the foregoing, it is

ORDERED, by the Court in banc, that the suggestion is denied.

This case presents antitrust and labor issues of national significance.   The issues have been fully engaged and developed by the majority and dissenting opinions.   Supreme Court review is essential to the resolution of these issues.   See Sup. Ct. R. 10.1(c).

I would grant the petition for rehearing in banc for the reasons set out in my panel dissent.   I believe further that the case involves a significant issue of statutory accommodation between two premier pieces of legislation incorporating our national policies on labor relations and competition and as such merits final resolution by the Supreme Court.


1.   A first-year player is a player who, in a previous year, attended an NFL training camp but played in fewer than three regular season games.

2.   The other players representing the plaintiff class were James Bishop, John Buddenberg, Gary Couch, Craig Davis, Ricky Andrews, Thom Kaumeyer, Wesley Pritchett, and John Simpson.

3.   Meanwhile, the NFLMC and the NFLPA finally agreed to a new seven-year collective bargaining agreement on January 6, 1993.

4.   Appellees also rely heavily on the Eighth Circuit's discussion of the nonstatutory labor exemption in Mackey v. NFL, 543 F.2d 606 (8th Cir.1976), cert. dismissed, 434 U.S. 801, 98 S.Ct. 28, 54 L.Ed.2d 59 (1977), which, they contend, “repeatedly expressed the understanding that the Supreme Court decisions precondition the exemption on an agreement.”   Brief of Appellees at 29.   However, they overlook the Eighth Circuit's subsequent statement in Powell, 930 F.2d at 1301, that “[a] collective bargaining agreement is not always essential to a finding that challenged employment terms fall within the labor exemption.”   We also note, as did the Powell court, id., that other decisions of the courts of appeals have applied the nonstatutory exemption to actions not embodied in collective bargaining agreements.   See, e.g., Amalgamated Meat Cutters v. Wetterau Foods, Inc., 597 F.2d 133, 136 (8th Cir.1979) (holding that employer's agreement with supplier to replace striking workers was protected by nonstatutory labor exemption);  Prepmore Apparel, Inc. v. Amalgamated Clothing Workers, 431 F.2d 1004, 1006-07 (5th Cir.1970) (holding that employer's refusal to deal with union was exempt from Sherman Act), cert. dismissed, 404 U.S. 801, 92 S.Ct. 21, 30 L.Ed.2d 34 (1971).

5.   In this regard, we note that even one commentator who favors imposing a union consent requirement for application of the nonstatutory labor exemption acknowledges that a “precedent-oriented” approach to this issue is “problematic ․ because in no Supreme Court case has union consent been contested.”   Note, Releasing Superstars From Peonage:  Union Consent and the Nonstatutory Labor Exemption, 104 Harv.L.Rev. 874, 881-82 (1991) [hereinafter Releasing Superstars ].

6.   We also note that each of these alternatives has its own set of problems.   In particular, those that require a court to determine when it is reasonable to believe a restraint on competition will be incorporated into a future collective bargaining agreement pose severe problems of administrability.   Further, they are likely to prompt the negotiating parties to distort their positions through posturing designed solely to influence this determination.

7.   The employers' duty to maintain the status quo even after expiration of a collective bargaining agreement fosters the bargaining process by protecting the union's position as the representative of the employees on subjects of mandatory bargaining, and by maintaining a stable environment for negotiations.   See Katz, 369 U.S. at 747, 82 S.Ct. at 1113;  Insurance Agents' Int'l Union, 361 U.S. at 485, 80 S.Ct. at 424.   We cannot fathom how subjecting actions taken pursuant to this duty to the threat of antitrust liability could possibly further the goals of the NLRA.   Employers faced with such a threat might well hesitate to agree with unions on important terms when they know the union may immediately challenge the terms under the antitrust laws once their agreement expires.

8.   Counsel acknowledged that the restraint on competition at issue in this case had no impact on the quality of professional football competition, and, thus, professional football's ability to compete with other entertainment products.   Specifically, counsel stated:[W]hat we are talking about here are the salaries that were paid to players who didn't even suit up, much less play.   For them to say that somehow or other the competitive contest[s] were improved and they could compete better with ballet is like a bootstrap leap into wonderland.   There is just no relationship.Tr. of Oral Argument at 33.As the quoted statement suggests, the clubs and the NFL contend that the fixed salary does impact the market for professional football contests, but they argue that it has the entirely procompetitive effect of improving the overall quality of NFL competition.   Specifically, they assert that the fixed salary provides a disincentive for the most talented rookie and first-year players to agree to being “stashed” on the Developmental Squads of top NFL teams, and thereby makes such players available to less successful teams that need their services immediately.   See Brief of Appellants at 8-9, 39-40.   Significantly, neither party ever has argued that the fixed salary for Developmental Squad players has any anticompetitive impact on the market for the NFL's product-the only market involved in this case with which the Sherman Act is concerned.

1.   The majority characterizes the players' position as an ill-begotten effort to “import” and “inject” antitrust principles into collective bargaining.   Majority opinion (“Maj. op.”) at 1050, 1052.   This of course turns the question inside-out.   Historically and logically, antitrust comes first;  the antitrust statutes antedate our labor laws and are of general applicability, applying to labor markets unless some statutory or judge-made exemption creates an exception.   The question before us then, is whether we should now extend a judicially-created exemption from antitrust coverage for collective bargaining agreements into the previously uncharted territory of post-collective bargaining imposition of the employers' terms.   Thus we might better ask to what extent we must “import” and “inject” labor law practices beyond the collective bargaining context into antitrust laws that would otherwise apply.

2.   Ironically, athletes have sought to achieve employer competition through the same collective bargaining processes that other employees use to restrain labor market competition.   Players' unions typically bargain for such mobility-enhancing and competition-maximizing measures as “free agency” and the elimination of reserve clauses, salary caps and player drafts.   Simultaneously, however, players' unions have embraced more traditional union demands such as league-wide minimum salaries and standardized pension plans.   See Ethan Lock, The Scope of the Labor Exemption in Professional Sports, 1989 Duke L.J. 339, 341 n. 12.

3.   I express no opinion as to whether these labor market restraints actually violate the antitrust laws.   In most cases, a rule of reason analysis will decide, see Smith v. Pro Football Inc., 593 F.2d 1173, 1182-83 (D.C.Cir.1978), requiring a fact-specific judgment as to whether the procompetitive benefits, if any, of the labor market restraint outweigh its anticompetitive costs.   In the present case, appellants colorably argued below that the wage restraint imposed on rookie and first-year nonroster players was necessary to enhance on-field “competitive balance” among teams, thereby making NFL football a more attractive and economically competitive entertainment product.   Cf. NCAA v. Board of Regents, 468 U.S. 85, 104 S.Ct. 2948, 82 L.Ed.2d 70 (1984).   The district court may have erred in ruling such alleged procompetitive benefits “irrelevant” and granting summary judgment to the players on the question of antitrust liability.   Although the majority does not reach the question, I would be inclined to remand for further proceedings on the question of antitrust liability under rule of reason analysis.

4.   Baseball is an exception for historical reasons.   See Flood v. Kuhn, 407 U.S. 258, 92 S.Ct. 2099, 32 L.Ed.2d 728 (1972).

5.   The applicability of antitrust law to unilaterally-imposed terms of employment was not a settled question prior to today's decision.   But since antitrust coverage presumptively applies to labor markets absent a special exemption, and no exemption clearly applied before today, it is fair to say that until now employers at least had to contemplate the risk of antitrust liability if they unilaterally imposed industry-wide terms of employment.

6.   A related concern is that some employers may be emboldened to go beyond hard bargaining and engage in “surface bargaining,” merely “going through the motions of negotiating,” K-Mart Corp. v. NLRB, 626 F.2d 704, 706 (9th Cir.1980), in order to position themselves to unilaterally impose industry-wide labor market restraints when the bargaining fails.   Surface bargaining is a form of bad-faith bargaining which “is often difficult to detect,” Seattle-First Nat. Bank v. NLRB, 638 F.2d 1221, 1227 n. 9 (9th Cir.1981), because, as we noted in NLRB v. Cauthorne, 691 F.2d 1023, 1026 n. 5 (D.C.Cir.1982), the duty to bargain in good faith does not prohibit an “employer's adamant insistence on pro-management terms,” and “neither side is required to agree to a proposal or make concessions.”   Thus it will often be difficult to distinguish bad faith from mere good-faith “adamant insistence” on favorable terms.   But whether or not some employers bargain in bad faith, the majority's rule clearly shifts the terms of employers' calculations:  entering bargaining confident of their ability to circumvent the constraints of the antitrust laws either through employees' capitulation or through post-impasse unilateral imposition of terms may not be bad faith, but merely a reasonable calculation as to the likely outcome of the negotiation process.

7.   Although the question is not directly at issue in this case, I share the majority's view that if the collective bargaining process is to work without undue interference from antitrust law, multiemployer “bargaining tactics” must fall within the scope of the nonstatutory labor exemption.   See Maj. op. at 1056-57.   Although multiemployer bargaining is not expressly authorized by the National Labor Relations Act, neither is it prohibited, and the fact that multiemployer bargaining predates the Act strongly suggests that it was contemplated as part of the collective bargaining process.  NLRB v. Truck Drivers Union Local 449, 353 U.S. 87, 94-95, 77 S.Ct. 643, 646-47, 1 L.Ed.2d 676 (1957) (“Buffalo Linen ”);  see also Williams, 45 F.3d at 688.   Multiemployer bargaining strengthens the bargaining process not only by facilitating bargaining on the employers' side, Buffalo Linen, 353 U.S. at 94-95, 77 S.Ct. at 646-647, and allowing employers to counter “whipsaw” tactics by unions, Charles D. Bonanno Linen Service, Inc. v. NLRB, 454 U.S. 404, 409-10 & n. 3, 102 S.Ct. 720, 723-24 & n. 3, 70 L.Ed.2d 656 (1982), but it also makes available joint benefit programs and frees unions from the burden of negotiating separately with many small employers.  Id.  Thus the statutory scheme of the NLRA may well require that employers be exempt from antitrust liability for basic bargaining tactics such as joining multiemployer bargaining units;  developing joint bargaining positions;  jointly proposing and negotiating for those positions;  and for joint actions, such as lockouts, aimed at pressuring the union in hopes of securing an agreement.   Cf. Jewel Tea, 381 U.S. at 712, 85 S.Ct. at 1616 (Opinion of Goldberg, J.) (suggesting multiemployer bargaining must be exempt from antitrust laws);  Volkswagenwerk v. Federal Maritime Comm'n, 390 U.S. 261, 287 n. 5, 88 S.Ct. 929, 943 n. 5, 19 L.Ed.2d 1090 (1968) (Harlan, J., concurring) (it is “obvious that the employers are not violating the antitrust laws either when they confer about wage policy preparatory to bargaining or when they sign an agreement”).   See also Goldman, supra, at 674-76 (urging nonstatutory labor exemption for multiemployer bargaining tactics, but not for unilaterally-imposed terms).   Indeed, there is broad precedent holding multiemployer tactical actions directed at the union within a collective bargaining framework exempt from antitrust scrutiny.   See, e.g., Amalgamated Meat Cutters & Butchers Workmen, Local No. 576 v. Wetterau Foods, Inc., 597 F.2d 133 (8th Cir.1979) (agreement to supply temporary replacement workers to struck employer);  Prepmore Apparel, Inc. v. Amalgamated Clothing Workers of Amer., 431 F.2d 1004, 1007 (5th Cir.1970) (alleged conspiracy to refuse to deal with union), cert. dismissed, 404 U.S. 801, 92 S.Ct. 21, 30 L.Ed.2d 34 (1971);  Newspaper Drivers & Handlers' Local No. 372 v. NLRB, 404 F.2d 1159 (6th Cir.1968) (mutual lockout agreement), cert. denied, 395 U.S. 923, 89 S.Ct. 1775, 23 L.Ed.2d 240 (1969);  Kennedy v. Long Island R. Co., 319 F.2d 366, 372-73 (2d Cir.) (multiemployer strike insurance pact), cert. denied, 375 U.S. 830, 84 S.Ct. 75, 11 L.Ed.2d 61 (1963).

8.   Fort Halifax thus contradicts the majority's contention, Maj. op. at 1054, that this case is controlled by dicta in American Ship Building, 380 U.S. at 316, 85 S.Ct. at 966, naming unilateral imposition of terms among the employers' “tools of economic self-help.”   Even if we are bound by the American Ship Building dicta, Fort Halifax -decided later-rejects the implication the majority would draw, that any independent statutory limitation on the terms an employer may impose at impasse impermissibly interferes with the employer's exercise of self-help tools.

9.   In that sense, this case is distinguishable from Williams, where employees mounted an antitrust challenge to employers' maintenance of the status quo after expiration of a collective bargaining agreement.   There, the employers' labor law duty to maintain the status quo squarely conflicted with their alleged antitrust duty not to maintain the status quo.   Rather than deciding the case on that narrow issue, however, the Williams court announced a broad ouster of antitrust law, similar to the majority's here.

10.   In reaching this conclusion, I do not find it necessary to reach some closely-related questions that have arisen in other cases.   The most important of these is when, if at all, the nonstatutory labor exemption attaching to a collective bargaining agreement ends-at expiration of the collective bargaining agreement, at impasse, or at some other post-impasse point.   Since the employees never agreed to the labor market restraints at issue here in the first place, no such questions arise in this case.

Opinion for the Court filed by Chief Judge EDWARDS. Dissenting opinion filed by Circuit Judge WALD.

Copied to clipboard