THOMAS, Chief Judge

: We consider an appeal and cross-appeal from an order enjoining the National Collegiate Athletic Association (the "NCAA") from enforcing rules that restrict the education-related benefits that its member institutions may offer students who play Football Bowl Subdivision ("FBS") football and Division I ("D1") basketball (collectively, "Student-Athletes"). See In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. (Alston ), 375 F. Supp. 3d 1058 (N.D. Cal. 2019). We have jurisdiction under 28 U.S.C. § 1291, and we affirm.

We conclude that the district court properly applied the Rule of Reason in determining that the enjoined rules are unlawful restraints of trade under section 1 of the Sherman Act, 15 U.S.C. § 1. We further conclude that the record supports the factual findings underlying the injunction and that the district court’s antitrust analysis is faithful to our decision in O’Bannon v. NCAA (O’Bannon II ), 802 F.3d 1049 (9th Cir. 2015).

I

A. The NCAA and its Compensation Rules

Founded in 1905, the NCAA regulates intercollegiate sports. Id. at 1053. Its mission statement is to "maintain intercollegiate athletics as an integral part of the educational program and the athlete as an integral part of the student body and, by so doing, retain a clear line of demarcation between intercollegiate athletics and professional sports." NCAA regulations govern, among other things, the payments that student-athletes may receive in exchange for and incidental to their athletic participation as well as in connection with their academic pursuits.

The NCAA divides its member schools into three competitive divisions. D1 schools—some 350 of the NCAA’s approximately 1,100 member schools—sponsor the largest athletic programs and offer the most financial aid. D1 football has two subdivisions, one of which is the FBS.

In August 2014, the NCAA amended its D1 bylaws (the "Bylaws") to grant the so-called "Power Five" conferences—the FBS conferences that generate the most revenue—autonomy to adopt collectively legislation in certain areas, including limits on athletic scholarships known as "grants-in-aid." In January 2015, the Power Five voted to increase the grant-in-aid limit to the cost of attendance ("COA") at each school. Since August 2015, the Bylaws have provided that a "full grant-in-aid" encompasses "tuition and fees, room and board, books and other expenses related to attendance at the institution up to the [COA]," as calculated by each institution’s financial aid office under federal law. See 20 U.S.C. §§ 1087kk, ll . The Bylaws also contain an "Amateurism Rule," which strips student-athletes of eligibility for intercollegiate competition if they "[u]se[ ] [their] athletics skill (directly or indirectly) for pay in any form in [their] sport." "[P]ay" is defined as the "receipt of funds, awards or benefits not permitted by governing legislation."

However, governing legislation permits a wide range of above-COA payments—both related and unrelated to education. Without losing their eligibility, student-athletes may receive, for instance: (i) awards valued at several hundred dollars for athletic performance ("athletic participation awards"), which may take the form of Visa gift cards; (ii) disbursements—sometimes thousands of dollars—from the NCAA’s Student Assistance Fund ("SAF") and Academic Enhancement Fund ("AEF") for a variety of purposes, such as academic achievement or graduation awards, school supplies, tutoring, study-abroad expenses, post-eligibility financial aid, health and safety expenses, clothing, travel, "personal or family expenses," loss-of-value insurance policies, car repair, personal legal services, parking tickets, and magazine subscriptions; (iii) cash stipends of several thousands of dollars calculated to cover costs of attendance beyond the fixed costs of tuition, room and board, and books, but used wholly at the student-athlete’s discretion; (iv) mandatory medical care (available for at least two years after the athlete graduates) for an athletics-related injury; (v) unlimited meals and snacks; (vi) reimbursements for expenses incurred by student-athletes’ significant others and children to attend certain athletic competitions; and (vii) a $30 per diem for "unitemized incidental expenses during travel and practice" for championship events.

The NCAA has carved out many of these exceptions in the past five years. For example, before 2015, athletic participation awards did not take the form of cash-like Visa gift cards. And once the NCAA permitted grants-in-aid for the full COA, effective August 2015, many more student-athletes began to receive above-COA payments, such as cash stipends, Pell Grants, and AEF as well as SAF distributions.

This expansion of above-COA compensation has coincided with rising revenue from D1 basketball and FBS football for the NCAA and its members. In the 2015–16 academic year, these programs generated $4.3 billion in revenue (a $300 million increase from the previous year) for the Power Five. And in 2016, the NCAA negotiated an eight-year extension (until 2032) of its multimedia contract for the broadcasting rights to March Madness, the annual D1 men’s basketball tournament. Under that agreement, the NCAA will receive $1.1 billion per year (an annual increase of over $325 million).

B. The O’Bannon Litigation

The NCAA is no stranger to antitrust litigation arising from its compensation rules. In 2009, Ed O’Bannon, a former UCLA basketball player, sued the NCAA after learning that a college basketball video game featured an avatar that resembled him and sported his jersey number. O’Bannon II , 802 F.3d at 1055. "The gravamen of [his] complaint" was that the NCAA illegally restrained trade, in violation of section 1, by preventing FBS football and D1 men’s basketball players from receiving compensation for the use of their names, images, and likenesses ("NILs"). Id . After a bench trial, the district court agreed under the Rule of Reason and entered relief for the plaintiffs. See O’Bannon v. NCAA (O’Bannon I ), 7 F. Supp. 3d 955, 962–63 (N.D. Cal. 2014), aff’d in part, rev’d in part , O’Bannon II , 802 F.3d at 1079. The district court acknowledged the NCAA’s evidence that college athletics’ "amateur tradition" helps maintain their popularity as a product distinct from professional sports. Id. at 999. It nevertheless concluded that this procompetitive benefit did not justify the NCAA’s "sweeping prohibition" on NIL compensation. Id. Based on evidence that "school loyalty and geography" primarily drive consumer demand and a lack of proof that small payments to student-athletes would diminish college sports’ popularity, the district court determined that the NCAA could justify, at most, restrictions on large payments. Id. at 1000–01.

After identifying two less restrictive alternatives ("LRAs") to the challenged rules, id. at 1004–07, the district court implemented those LRAs through an injunction that required the NCAA to permit its schools to (i) "use the licensing revenue generated from the use of their student-athletes’ [NILs] to fund stipends covering the [COA]"; and (ii) to make deferred, post-eligibility cash payments in NIL revenue, not to exceed $5,000, to student-athletes. Id. at 1007–08 ; see also id. at 1008 (finding no evidence that "such a modest payment" would "undermine[ ]" NCAA’s "legitimate procompetitive goals"). The NCAA appealed.

A majority of a Ninth Circuit panel concluded that the district court’s decision, the first of its kind, was "largely correct." O’Bannon II , 802 F.3d at 1053 ; id. at 1079 (Thomas, C.J., concurring in part and dissenting in part). The panel unanimously affirmed the injunction insofar as it required the NCAA to permit athletic scholarships for the full COA, but a panel majority reversed and vacated the injunction’s requirement that the NCAA allow deferred NIL payments. Id . at 1053.

In pertinent part, the panel rejected the NCAA’s threshold argument that its amateurism rules, including those governing compensation, are "valid as a matter of law" under NCAA v. Board of Regents of the University of Oklahoma , 468 U.S. 85, 104 S.Ct. 2948, 82 L.Ed.2d 70 (1984). O’Bannon II , 802 F.3d at 1061. The panel acknowledged the Supreme Court’s observation, in "dicta," that the NCAA has historically preserved its product by, inter alia , prohibiting payments to student-athletes. Id. at 1063 (citing Bd. of Regents , 468 U.S. at 102, 104 S.Ct. 2948 ). But it declined to read that statement as perpetual blanket approval for the NCAA’s compensation rules, which were not at issue in Board of Regents . Id. Though conceding that the NCAA’s "amateurism rules are likely to be procompetitive," id. at 1053, the panel refused to exempt them from antitrust scrutiny, see id. at 1064 (explaining that a procompetitive rule "can still be invalid under the Rule of Reason").

The panel then affirmed much of the district court’s analysis. See id. at 1069–76. As is relevant here, it found, based on the record, "a concrete procompetitive effect in the NCAA’s commitment to amateurism: namely that the amateur nature of collegiate sports increases their appeal to consumers." Id. at 1073. As to LRAs, it agreed that the ban on funding COA scholarships with NIL revenue was "patently and inexplicably stricter" than necessary to differentiate college from professional sports. Id. at 1075 ("[B]y the NCAA’s own standards, student-athletes remain amateurs as long as any money paid to them goes to cover legitimate educational expenses."). It clarified that courts must invalidate such restraints but may not "micromanage organizational rules" or "strike down largely beneficial market restraints[.]" Id.

A panel majority, however, found error in the district court’s adoption of deferred NIL compensation "untethered to [student-athletes’] education expenses" as a viable LRA. Id. at 1076. It explained that "not paying student-athletes is precisely what makes them amateurs " and disagreed that "being a poorly-paid professional" is " ‘virtually as effective’ for that market as being a[n] amateur." Id. To avert a "transition[ ]" to "minor league status" and to heed the "Supreme Court’s admonition that [courts] must afford the NCAA ‘ample latitude’ to superintend college athletics," the majority vacated this portion of the injunction. Id. at 1079 (quoting Bd. of Regents , 468 U.S. at 120, 104 S.Ct. 2948 ). In closing, it "emphasize[d] the limited scope of [its] decision," explaining that "in th[at] case," the Rule of Reason did "not require" anything "more" of the NCAA than to permit student-athletes to receive scholarships for the COA. Id.

C. The Alston Litigation

In March 2014, while the NCAA was litigating O’Bannon I , FBS football and D1 men’s and women’s basketball players filed several antitrust actions against the NCAA and eleven D1 conferences that were transferred to and, with one exception, consolidated before the same district court presiding over O’Bannon I . Rather than confining their challenge to rules prohibiting NIL compensation, Student-Athletes sought to dismantle the NCAA’s entire compensation framework.

In December 2015, the district court certified three injunctive relief classes comprised of (i) FBS football players, (ii) D1 men’s basketball players, and (iii) D1 women’s basketball players. Each subclass consists of student-athletes who have received or will receive a full grant-in-aid during the pendency of this litigation.

Nearly a year after our decision in O’Bannon II , the NCAA sought judgment on the pleadings, invoking res judicata. It argued that O’Bannon II "requires nothing more of the NCAA than that it permit its member schools to provide student-athletes with their full education-related [COA]." Because the NCAA had already amended its rules to satisfy that requirement, it reasoned that any post-O’Bannon antitrust challenges to its compensation rules must fail. The district court denied the motion. It explained that Student-Athletes, unlike the O’Bannon plaintiffs, had challenged, among other things, limits on non-cash, education-related benefits. It acknowledged the possibility that O’Bannon forecloses a type of relief—lifting restrictions on cash payments untethered to educational expenses—but declined to read it more broadly than that.

Cross-motions for summary judgment followed. The district court again rejected the NCAA’s preclusion arguments. As to the merits, it adopted, at the parties’ request, the market definition from O’Bannon I : the market for a college education or, alternatively, student-athletes’ labor. It then granted Student-Athletes summary judgment at the Rule of Reason’s first step, as the NCAA did not meaningfully dispute that the challenged rules have anticompetitive effects in the relevant markets. At the Rule of Reason’s second step, it determined that the NCAA had raised triable issues as to whether its rules have the procompetitive effect(s) of maintaining the popularity of its elite college basketball and football products or integrating student-athletes into the wider campus community. Last, the district court found that Student-Athletes had proffered sufficient evidence to support their two proposed LRAs: (i) allowing individual conferences, but not the NCAA, to regulate student-athlete compensation; or (ii) enjoining NCAA rules that restrict both non-cash education-related benefits and benefits that are incidental to athletic participation.

D. The District Court’s Decision

After a ten-day bench trial, the district court entered judgment for Student-Athletes, in part. The court concluded that NCAA limits on education-related benefits are unreasonable restraints of trade, and accordingly enjoined those limits; however, the court declined to hold that NCAA limits on compensation unrelated to education likewise violate section 1. Alston , 375 F. Supp. 3d at 1109.

1. Determination that O’Bannon Is Not Preclusive

At the outset of its conclusions of law, the district court again declined to dismiss the case on res judicata grounds. Id. at 1092–96. It identified "material factual differences" between O’Bannon and the Alston litigation, id. at 1095, including in the identity of class members and the rules and rights at issue, see id. at 1093–94 (explaining that "[t]he crux of the O’Bannon case was the right to student-athletes’ NIL[s]," whereas "[t]he conduct at issue here is not connected to NIL rights" but to limits on above-COA compensation and benefits); id. at 1094 (noting that challenged rules either did not exist or have "materially changed" since O’Bannon ). The district court then proceeded to its Rule of Reason analysis.

2. The Relevant Market

To begin, the district court accepted Student-Athletes’ trial theory narrowing the relevant market to one in which Student-Athletes sell their "labor in the form of athletic services" to schools in exchange for athletic scholarships and other payments permitted by the NCAA. Id. at 1067, 1097.

3. Anticompetitive Effects

Next, the court reiterated its summary judgment finding of "significant anticompetitive effects in the relevant market." Id. at 1067, 1097. It relied on Student-Athletes’ economic analyses reflecting that schools, as buyers of athletic services, exercise monopsony power to artificially cap compensation at a level that is not commensurate with student-athletes’ value. Id. at 1068. Based on these analyses, it also found that, but for the challenged restraints, schools would offer recruits compensation that more closely correlates with their talent. Id. at 1068–69, 1098.

The district court also highlighted additional trial evidence demonstrating the challenged rules’ anticompetitive effects. This included testimony that, in 2013, the Power Five began to urge the NCAA to loosen its compensation restrictions based on a concern that existing rules incongruously allowed schools to spend on virtually anything, including palatial athletic facilities and seven-figure coaches’ salaries, except direct financial support for student-athletes. Id. at 1068–69. In the district court’s view, the Power Five’s concerns constituted further proof that, absent the NCAA’s rules, student-athletes would receive higher compensation. Id. at 1069. Although the NCAA granted the Power Five autonomy to create new forms of compensation and to expand previously available compensation and benefits in 2015, the district court observed that these conferences remain constrained by "overarching NCAA limits" that cap compensation at an artificially low level. Id.

4. Procompetitive Effect

The district court then turned to the NCAA’s asserted procompetitive justifications. In pertinent part, the NCAA argued that the challenged rules implement "amateurism," which drives consumer interest in college sports because "consumers ‘value amateurism.’ " Id. at 1070 (internal citation omitted). The district court accepted this justification with respect to the NCAA’s limits on cash compensation untethered to education, but not as to its limits on non-cash education-related benefits. Id. at 1082–83, 1101–02.

As a preliminary matter, the district court found no proof that the challenged rules directly foster consumer demand. Id. at 1070. It acknowledged the NCAA’s theory that its rules safeguard "amateurism" for consumers’ benefit, but the meaning of that term eluded the court. See id. at 1070–71 (noting former SEC commissioner’s testimony that he "do[es not] even know what [amateurism] means" (internal citation omitted)). Though the NCAA defined amateurism during the litigation as " ‘not paying’ the participants," id. at 1071 (internal citation omitted), the district court observed that this purported pay-for-play prohibition is riddled with exceptions. See id. at 1071–74.

After cataloguing the long list of above-COA payments that the NCAA permits, the court then reached two conclusions: (i) the challenged rules "do not follow any coherent definition of amateurism ... or even ‘pay,’ " and (ii) these payments (many of which post-date O’Bannon ) have not diminished demand for college sports, which "remain[ ] exceedingly popular and revenue-producing." Id. at 1074.

On the question of consumer demand, the district court found Student-Athletes’ evidence regarding the effect (or lack thereof) of above-COA compensation on demand more compelling than the NCAA’s. For instance, in the battle of economic experts, the district court found the NCAA’s only demand expert, Dr. Kenneth Elzinga, unreliable because he failed to study "standard measures of consumer demand, such as revenues, ticket sales, or ratings," but instead relied on interviews with NCAA affiliates introduced to him by defense counsel. Id. at 1075. The district court further found his analysis irrelevant as he refused to study consumer response to historical changes in compensation levels based on the false premise that the NCAA’s amateurism rules have not materially changed over time. Id.

By contrast, the district court credited Student-Athletes’ expert Dr. Daniel Rascher’s demand analysis, which was based on two natural experiments and, in some respects, corroborated by defense witnesses. Id. at 1076–78, 1100. The first experiment—comparing consumer demand before and after the August 2015 increase to the grant-in-aid limit, which resulted in "thousands of class members receiving significant" above-COA payments, including SAF and AEF distributions—demonstrated "no negative impact on consumer demand." Id. at 1076. In fact, Dr. Rascher found that revenues from D1 basketball and FBS football, "one of the best economic measures of consumer demand," have increased since 2015. Id. at 1076–77 ; see also id. at 1078 (noting corroborating testimony by an NCAA Rule 30(b)(6) witness and a Big 12 Rule 30(b)(6) witness). The second experiment—comparing demand before and after the University of Nebraska (of the Big Ten) began providing athletes up to $7,500 in post-eligibility education-related aid—likewise did not demonstrably reduce interest in Nebraska sports or FBS football and D1 basketball more broadly. Id. at 1077–78.

The district court also found Student-Athletes’ survey expert, Dr. Hal Poret, considerably more persuasive than the NCAA’s, Dr. Bruce Isaacson. Id. at 1078–80, 1100–01. Dr. Isaacson asked respondents why they watch college sports and listed "amateurs and/or not paid" as one possible reason, but failed to indicate that "amateurs" means "not paid" or to otherwise define "amateurs," thus "render[ing] the responses hopelessly ambiguous." Id. at 1078. Moreover, he measured only consumer preference and conceded that he did not attempt to study behavior . Id. at 1079. By contrast, Dr. Poret tested behavior and found that consumers would continue to view or attend college athletics (at the same rate) even if eight types of compensation that the NCAA currently prohibits or limits were individually implemented. Id. at 1079–80. The district court credited this conclusion. Id. at 1079–80 & n.24.

Testimony by NCAA lay witnesses that "student" status drives demand also failed to persuade the district court of a connection between the challenged compensation regime and demand. Id. at 1082, 1101. It reasoned that "student-athletes would continue to be students in the absence of the challenged rules," id. at 1082, relying on O’Bannon II ’s observation that higher education "would still be available to student-athletes if they were paid some compensation in addition to their athletic scholarships," id. at 1101 (quoting O’Bannon II , 802 F.3d at 1073 ). It also underscored the absence of evidence that the NCAA had promulgated its rules based on demand analyses. Id. at 1080, 1100–01.

Despite finding the NCAA’s procompetitive theory largely unpersuasive, the district court "credit[ed] the importance to consumer demand of maintaining a distinction between college sports and professional sports." Id. at 1082. The court then found that some NCAA rules—the COA limit on the grant-in-aid, limits on compensation unrelated to education, and limits on cash awards for graduating or other academic achievements—serve that purpose by precluding "unlimited payments unrelated to education, akin to salaries seen in professional sports leagues." Id. at 1082–83 ; see also id. at 1101–02. But the court concluded that limits on "non-cash education-related benefits," such as post-eligibility graduate scholarships or tutoring, do not have that effect; it reasoned that such benefits "could not be confused with a professional athlete’s salary" and would only "emphasize that the recipients are students." Id. at 1083.

5. Less Restrictive Alternative

At the Rule of Reason’s third step, the district court considered whether three potential alternatives to the challenged restraints were less restrictive but virtually as effective in preventing "demand-reducing unlimited compensation indistinguishable from that observed in professional sports." Id. at 1086. The district court rejected two proposed LRAs, both of which would have permitted individual conferences to limit above-COA compensation, but would have otherwise invalidated either (i) all NCAA compensation limits or (ii) NCAA limits on education-related compensation and existing caps on benefits incidental to athletics participation, such as healthcare, pre-season expenses, and athletic participation awards. Id. at 1086–87. The district court found that both these alternatives would enable professional-style cash payments, thus threatening the distinction between college and professional sports. Id. at 1087. The court acknowledged the possibility that conferences could "discover" demand-preserving compensation levels. Id. But it rejected these LRAs to avoid demand-reducing "miscalculations" during "the inevitable trial-and-error phase." Id.

The district court then identified a viable LRA:

(1) allow the NCAA to continue to limit grants-in-aid at not less than the [COA]; (2) allow the [NCAA] to continue to limit compensation and benefits unrelated to education; (3) enjoin NCAA limits on most compensation and benefits that are related to education, but allow it to limit education-related academic or graduation awards and incentives, as long as the limits are not lower than its limits on athletic performance awards now or in the future.

Id. The court enumerated specific education-related benefits that the NCAA would be unable to prohibit or limit under the LRA: "computers, science equipment, musical instruments and other items not currently included in the [COA] but nonetheless related to the pursuit of various academic studies"; post-eligibility scholarships for undergraduate, graduate, and vocational programs at any school; tutoring; study-abroad expenses; and paid post-eligibility internships. Id. at 1088.

The district court explained that this LRA would permit some NCAA regulation of cash graduation or academic awards because these payments could otherwise morph into professional-like salaries. Id. It instructed that the cap on such awards should not fall below the existing limit on aggregate athletic participation awards (currently, $5,600), as receipt of the latter "has been shown not to decrease consumer demand and not to be inconsistent with the NCAA’s understanding of amateurism." Id. Under this LRA, individual conferences may continue to limit all payment types because "no individual conference dominates nearly the entire market, like the NCAA does." Id. The district court further reasoned that this LRA would not "greatly impact[ ]" the NCAA’s "latitude to superintend college sports," as it "would affect only a small fraction of [its] rulemaking jurisdiction." Id.

The district court concluded that this LRA would be virtually as effective as the challenged rules at preserving student-athletes’ status as students (and thus demand), analogizing it to the LRA affirmed in O’Bannon II : Both require the NCAA to permit members "to cover legitimate education-related costs." Id. at 1105 (citing O’Bannon II , 802 F.3d at 1075 ). Finally, it determined that, far from resulting in significantly increased costs, the LRA’s elimination of a category of rules would decrease the NCAA’s enforcement costs. Id. at 1090–91, 1105.

6. Remedy

The district court implemented this LRA via a permanent injunction. See In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. , 2019 WL 1593939 (N.D. Cal. Mar. 8, 2019). The injunction provides that the parties may move to modify its list of education-related benefits and that the NCAA may move to incorporate a definition of compensation and benefits that are "related to education" if it chooses to adopt one. Id. at *1. It also allows the NCAA to regulate how its members provide education-related benefits. Id. ; see also Alston , 375 F. Supp. 3d at 1107 ("[T]he NCAA could require schools to pay for these items directly or to reimburse student-athletes for [equipment] expenses if adequate proof of purchase is shown."). The court reiterated that NCAA members remain free to independently restrict pay. Alston , 375 F. Supp. 3d at 1109. And it stayed the injunction pending resolution of a timely appeal. Id. at 1110.

E. Post-Appeal Developments

After the NCAA timely appealed, California enacted the Fair Pay to Play Act (the "FPP Act"). See Cal. S.B. 206 (Sept. 30, 2019), Cal. Educ. Code § 67456. The FPP Act requires the NCAA and its member institutions to permit student-athletes enrolled in California colleges and universities to earn compensation from the use of their NILs. Id. § 67456(a), (g). It takes effect on January 1, 2023. Id. § 67456(h).

In response to the FPP Act, the NCAA created a working group that has recommended permitting NIL benefits so long as they are tethered to education and otherwise preserve the distinction between college and professional sports recognized in O’Bannon II . See Fed. and State Leg. Working Grp. Report 4 (Oct. 23, 2019), available at https://tinyurl.com/working-grp-report. In recent testimony before the Senate Commerce Subcommittee on Manufacturing, Trade and Consumer Protection, NCAA President Dr. Mark Emmert denied that the NCAA would be "taking any action that is contrary to the position advocated by the NCAA or accepted by the Ninth Circuit with respect to the type of NIL payments that were at issue in the O’Bannon case[.]" See Test. of Dr. Mark Emmert 6 (Feb. 11, 2020), available at https://tinyurl.com/Emmert-Test-y.

II

The application of stare decisis and res judicata are questions of law that we review de novo. See In re Watts , 298 F.3d 1077, 1079 (9th Cir. 2002) ; Media Rights Techs., Inc. v. Microsoft Corp. , 922 F.3d 1014, 1020 (9th Cir. 2019).

We review factual findings for clear error and legal conclusions de novo. See O’Bannon II , 802 F.3d at 1061. Under clear error review, we must "accept the district court’s findings of fact unless we are left with the definite and firm conviction that a mistake has been committed." Id. (quoting FTC v. BurnLounge, Inc. , 753 F.3d 878, 883 (9th Cir. 2014) ); see also United States v. Alexander , 106 F.3d 874, 877 (9th Cir. 1997) ("We must not reverse as long as the findings are plausible in light of the record viewed in its entirety[.]"). In other words, a decision is not clearly erroneous unless it "strike[s] us as wrong with the force of a five-week-old, unrefrigerated dead fish." Prete v. Bradbury , 438 F.3d 949, 968 n.23 (9th Cir. 2006) (internal citation omitted).

Last, "[w]e review a district court’s decision to grant a permanent injunction for an abuse of discretion"; the "factual findings underpinning the award" for clear error; and the "rulings of law relied upon by the district court in awarding injunctive relief" de novo. Ollier v. Sweetwater Union High Sch. Dist. , 768 F.3d 843, 867 (9th Cir. 2014) (internal citations and quotation marks omitted).

III

The district court correctly concluded O’Bannon II did not foreclose this litigation as a matter of stare decisis and res judicata.

A

Stare decisis binds "today’s Court" to "yesterday’s decisions." Danielson v. Inslee , 945 F.3d 1096, 1097 (9th Cir. 2019) (quoting Kimble v. Marvel Entm’t, LLC , 576 U.S. 446, 135 S. Ct. 2401, 2409, 192 L.Ed.2d 463 (2015) ). "In determining whether [we are] bound by an earlier decision," we consider "not only the rule announced, but also the facts giving rise to the dispute, other rules considered and rejected and the views expressed in response to any dissent or concurrence." Hart v. Massanari , 266 F.3d 1155, 1170 (9th Cir. 2001). "Insofar as there may be factual differences between the current case" and O’Bannon II , we "must determine whether those differences are material to the application of the rule or allow the precedent to be distinguished on a principled basis." Id. at 1172 ; see also In re Osborne , 76 F.3d 306, 309 (9th Cir. 1996) (explaining that decisions "furnish[ ] the rule for the determination of a subsequent case involving identical or similar material facts" (internal citation omitted)).

Antitrust decisions are particularly fact-bound. The Supreme Court has long emphasized that the Rule of Reason "contemplate[s]" "case-by-case adjudication." See Leegin Creative Leather Prods., Inc. v. PSKS, Inc. , 551 U.S. 877, 899, 127 S.Ct. 2705, 168 L.Ed.2d 623 (2007) ; see also Maple Flooring Mfrs.’ Ass’n v. United States , 268 U.S. 563, 579, 45 S.Ct. 578, 69 L.Ed. 1093 (1925) ("[E]ach case arising under the Sherman Act must be determined upon the particular facts disclosed by the record, and ... opinions in those cases must be read in the light of their facts"); Phillip Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application , ¶ 1205c3 (4th ed. 2018) ("Continuing contracts in restraint of trade," are "typically subject to continuing reexamination," and "even a judicial holding that a particular agreement is lawful does not immunize it from later suit or preclude its reexamination as circumstances change.").

O’Bannon II was a decision of "limited scope," which the panel majority summarized as follows:

[W]e reaffirm that NCAA regulations are subject to antitrust scrutiny and must be tested in the crucible of the Rule of Reason. ... [T]he NCAA is not above the antitrust laws, and courts cannot and must not shy away from requiring

the NCAA to play by the Sherman Act’s rules. In this case , the NCAA’s rules have been more restrictive than necessary to maintain its tradition of amateurism in support of the college sports market. The Rule of Reason requires that the NCAA permit its schools to provide up to the [COA] to their student athletes. It does not require more.

802 F.3d at 1079 (emphasis added).

In arguing that the last two sentences of this passage foreclose the current litigation, the NCAA ignores the inherently fact-dependent nature of a Rule of Reason analysis, which evaluates dynamic market conditions and consumer preferences; the panel majority’s manifest effort to limit its decision to the record before it; and the majority’s mandate that courts must continue to subject NCAA rules, including those governing compensation, to antitrust scrutiny. See id. at 1064 ("The amateurism rules’ validity must be proved, not presumed.").

Far from straying outside O’Bannon II ’s bounds, the district court here sought to toe the line that the panel majority drew. The court uncapped education-related benefits, but left in place NCAA limits on compensation unrelated to education, consistent with the majority’s observation that "student-athletes remain amateurs as long as any money paid to them goes to cover legitimate educational expenses ." Id. at 1075 (emphasis added); see also id. at 1076 (vacating injunction only insofar as it forced NCAA to permit "cash payments untethered to ... education expenses").

The district court meaningfully and properly distinguished O’Bannon II from the current litigation as a narrow challenge to restrictions on NIL compensation. See id. at 1052 (introducing challenged rules as those that "prohibit student-athletes from being paid for the use of their [NILs]"); id. at 1055 (stating that the "gravamen of O’Bannon’s complaint was that the NCAA’s amateurism rules, insofar as they prevented student-athletes from being compensated for the use of their NILs, were an illegal restraint of trade"); id. at 1073 n.17 ("The correct inquiry under the Rule of Reason is: What procompetitive benefits are served by the NCAA’s existing rule banning NIL payments?"). Additionally, the proposed LRAs in O’Bannon were expressly limited to "licensing revenue generated from the use" of student-athletes’ NILs. See O’Bannon I , 7 F. Supp. 3d at 1007. By contrast, this action more broadly targets the "interconnected set of NCAA rules that limit the compensation [student-athletes] may receive in exchange for their athletic services." Alston , 375 F. Supp. 3d at 1062. And Student-Athletes sought LRAs that would uncap above-COA compensation, regardless whether their NILs have, will, or could generate any revenue that would fund such compensation. See id. at 1086.

The NCAA’s argument that it should not incur antitrust liability for relaxing its compensation limits since O’Bannon is not persuasive. The district court rightly concluded that this argument misses the mark: "It is the fact that the prices of student-athlete compensation are fixed, as opposed to the amount at which these prices are fixed, that renders the agreements at issue anticompetitive." Id. at 1095 (citing O’Bannon II , 802 F.3d at 1071 ("It is no excuse that the prices fixed are themselves reasonable.") (quoting Catalano, Inc. v. Target Sales, Inc. , 446 U.S. 643, 647, 100 S.Ct. 1925, 64 L.Ed.2d 580 (1980) )).

Additionally, the NCAA’s concession that it has relaxed its compensation limits since O’Bannon only underscores that the instant litigation is materially factually different from O’Bannon . Indeed, as Student-Athletes argue, the changes to compensation limits since O’Bannon "alter the factual assumption that drove the result in O’Bannon : they show that non-education-related cash payments in excess of the [COA] are no longer a ‘quantum leap’ from current NCAA practice[.]" See O’Bannon II , 802 F.3d at 1078 ("The difference between offering student-athletes education-related compensation and offering them cash sums untethered to educational expenses is not minor; it is a quantum leap.").

In O’Bannon II , the majority addressed only two types of above-COA allowances: Pell Grants and prize money for tennis recruits. See id. at 1058–59. It distinguished Pell Grants, which are "intended for education-related expenses," from "pure cash compensation" for athletic performance. Id. at 1078 n.24. And it declared that "award money from outside athletic events implicates amateurism differently than allowing schools to pay student-[athletes] directly." Id. at 1077 n.21. Neither of these above-COA allowances is analogous to the post- O’Bannon II forms of compensation—provided by schools and unrelated to education—that the district court cited to support its conclusion that the NCAA, contrary to its theory of amateurism, does provide at least some "pay for play." See Alston , 375 F. Supp. 3d at 1071–74. For example, the court found that, after the O’Bannon record closed, student-athletes have received, inter alia , athletic participation awards in the form of Visa gift cards, SAF disbursements in the thousands of dollars to pay for loss-of-value insurance, and personal expenses unrelated to education. Id. at 1095. Based on these innovations, the court fairly concluded that the compensation landscape has meaningfully changed since O’Bannon . See id. at 1094.

In sum, because O’Bannon II "was decided on a narrow set of facts that are distinguishable from the present case," we "decline to adopt" the NCAA’s "broad interpretation" of that decision. United States v. Silver , 245 F.3d 1075, 1079 (9th Cir. 2001).

B

Res judicata, also known as "claim preclusion," "bars a party in successive litigation from pursuing claims that ‘were raised or could have been raised in [a] prior action.’ " Media Rights Techs. , 922 F.3d at 1020 (internal citation omitted). It applies when there is: (i) an identity of claims between the prior and subsequent actions; (ii) a final judgment on the merits; and (iii) identity or privity between the parties. Id. at 1020–21. The NCAA bears the burden of proving all three elements. Id . at 1021. The NCAA fails to carry its burden with respect to the first element.

"Claim preclusion does not apply to claims that were not in existence and could not have been sued upon ... when the allegedly preclusive action was initiated." Id. (internal citation omitted). That bright-line rule is dispositive here. Because Student-Athletes’ antitrust claim "arose from events that occurred after" the O’Bannon record closed in August 2014—that is, the above-described proliferation of permissible above-COA payments alongside a growth in revenues from FBS football and D1 basketball—it is "not barred." Howard v. City of Coos Bay , 871 F.3d 1032, 1040 (9th Cir. 2017) ; see also Harkins Amusement Enters., Inc. v. Harry Nace Co. , 890 F.2d 181, 183 (9th Cir. 1989) ("Failure to gain relief for one period of time does not mean that the plaintiffs will necessarily fail for a different period of time"); California v. Chevron Corp. , 872 F.2d 1410, 1415 (9th Cir. 1989) (providing that "conduct of the parties since the first judgment[ ] must be considered" in connection with successive antitrust suits).

IV

The district court properly granted judgment on the Student-Athletes’ Sherman Act § 1 claim. The Sherman Act prohibits, inter alia , agreements "in restraint of" interstate trade or commerce. 15 U.S.C. § 1. The Supreme Court has interpreted section 1 "as ‘outlaw[ing] only unreasonable restraints’ of trade." See In re Nat’l Football League’s Sunday Ticket Antitrust Litig. , 933 F.3d 1136, 1149 (9th Cir. 2019) (alteration in original) (quoting State Oil Co. v. Khan , 522 U.S. 3, 10, 118 S.Ct. 275, 139 L.Ed.2d 199 (1997) ). "[W]hen considering agreements among entities involved in league sports, such as here, [we] must determine whether the restriction is unreasonable under the [R]ule of [R]eason." Id. at 1150 n.5 ; see also O’Bannon II , 802 F.3d at 1069 ("[T]he appropriate rule is the Rule of Reason.").

As applied here, under the Rule of Reason’s "three-step framework:"(1) Student-Athletes "bear[ ] the initial burden of showing that the restraint produces significant anticompetitive effects within a relevant market"; (2) if they carry that burden, the NCAA "must come forward with evidence of the restraint’s procompetitive effects"; and (3) Student-Athletes "must then show that any legitimate objectives can be achieved in a substantially less restrictive manner." O’Bannon II , 802 F.3d at 1070 (quoting Tanaka v. Univ. of S. Cal. , 252 F.3d 1059, 1063 (9th Cir. 2001) ). Throughout this analysis, we remain mindful that, although "the NCAA is not above the antitrust laws," id. at 1079, courts are not "free to micromanage organizational rules or to strike down largely beneficial market restraints," id. at 1075. Accordingly, a court must invalidate a restraint and replace it with an LRA only if the restraint is "patently and inexplicably stricter than is necessary to accomplish all of its procompetitive objectives." Id. at 1075.

A

The district court properly concluded that the Student-Athletes carried their burden at the first step of the Rule of Reason. The district court found that the NCAA’s rules have "significant anticompetitive effects in the relevant market" for Student-Athletes’ labor on the gridiron and the court. See Alston , 375 F. Supp. 3d at 1070 ("[B]ecause elite student-athletes lack any viable alternatives to [D1], they are forced to accept, to the extent they want to attend college and play sports at an elite level after high school, whatever compensation is offered to them by [D1] schools, regardless of whether any such compensation is an accurate reflection of the competitive value of their athletic services."). These findings "have substantial support in the record," O’Bannon II , 802 F.3d at 1070 ; see Alston , 375 F. Supp. 3d at 1067–70, and the NCAA does not dispute them, see O’Bannon II , 802 F.3d at 1072.

B

The NCAA does, however, quarrel with the district court’s analysis at the Rule of Reason’s second step, where the NCAA bears a "heavy burden" of "competitively justify[ing]" its undisputed "deviation from the operations of a free market." Bd. of Regents , 468 U.S. at 113, 104 S.Ct. 2948 ; see also O’Bannon II , 802 F.3d at 1064 (explaining that the NCAA is not entitled to a presumption that its restraints are procompetitive). On appeal, the NCAA advances a single procompetitive justification: The challenged rules preserve "amateurism," which, in turn, "widen[s] consumer choice" by maintaining a distinction between college and professional sports.

"Improving customer choice is procompetitive." Paladin Assocs., Inc. v. Mont. Power Co. , 328 F.3d 1145, 1157 (9th Cir. 2003) ; see also O’Bannon II , 802 F.3d at 1072 ("[A] restraint that broadens choices can be procompetitive."). Thus, the district court properly "credit[ed] the importance to consumer demand of maintaining a distinction between college and professional sports." Alston , 375 F. Supp. 3d at 1082.

The district court concluded, however, that only some of the challenged rules serve that procompetitive purpose: limits on above-COA payments unrelated to education, the COA cap on athletic scholarships, and certain restrictions on cash academic or graduation awards and incentives. Id. at 1101–02 (recognizing that removal of these restrictions could result in unlimited cash payments akin to professional salaries). It explained that the remaining rules—those restricting "non-cash education-related benefits"—do nothing to foster or preserve demand because "[t]he value of such benefits, like a scholarship for post-eligibility graduate school tuition, is inherently limited to its actual value, and could not be confused with a professional athlete’s salary." Id. at 1083.

The record amply supports these findings. The district court reasonably relied on demand analyses, survey evidence, and NCAA testimony indicating that caps on non-cash, education-related benefits have no demand-preserving effect and, therefore, lack a procompetitive justification. See id. at 1076–80.

First, Dr. Rascher’s and Dr. Noll’s demand analyses demonstrate that the NCAA has loosened its restrictions on above-COA, education-related benefits since O’Bannon without adversely affecting consumer demand. These benefits include SAF and AEF distributions to cover fifth-and sixth-year aid, postgraduate scholarships, tutoring, international student fees, educational supplies, academic achievement or graduation awards, graduate school exam fees, and fees for internship programs. Id. at 1072 n.15.

Second, Student-Athletes’ survey evidence reflects that individually implementing seven types of education-related benefits—limited or forbidden under the challenged rules—would not diminish the survey respondents’ viewership or attendance.

Third, NCAA witnesses confirmed that the NCAA set limits on education-related benefits without consulting any demand studies. See id. at 1080 ("Indeed, [Kevin] Lennon, who has worked for the NCAA for more than thirty years, testified that he does not recall any instance in which any study on consumer demand was considered by the NCAA membership when making rules about compensation"); see also id. at 1074 ("Defendants have not provided any cogent explanation for why the NCAA generally prohibits financial aid for graduate school at another institution, or for why the Senior Scholar Awards are limited in quantity and amount.").

Notwithstanding this evidence, the NCAA accuses the district court of straying from a purported "judicial consensus" that the NCAA expands consumer choice by enforcing an amateurism principle under which student-athletes "must not be paid" a penny over the COA. This sweeping procompetitive justification—the "Not One Penny" standard, in Dr. Noll’s parlance—lacks support in both precedent and the record.

Although both Board of Regents and O’Bannon II define amateurism to exclude payment for athletic performance, neither purports to immortalize that definition as a matter of law. In fact, O’Bannon II recognizes that Board of Regents ’ discussion of amateurism is "dicta." 802 F.3d at 1063. And to the extent the O’Bannon II majority accepted the NCAA’s conception of amateurism, it did so based on the record, which demonstrated a "concrete procompetitive effect," id. at 1073, of limiting above-COA "NIL cash payments untethered to [students’] education expenses," id. at 1076.

The record in this case, by contrast, reflects no such concrete procompetitive effect of limiting non-cash, education-related benefits. Instead, the record supports a much narrower conception of amateurism that still gives rise to procompetitive effects: Not paying student-athletes "unlimited payments unrelated to education, akin to salaries seen in professional sports leagues" is what makes them "amateurs." Alston , 375 F. Supp. 3d at 1083. The district court credited NCAA testimony that college sports resonates with fans because they are not professionalized, and that "if the college game looks to be professional sports, [fewer] people will watch it." Id. at 1082 (internal citations omitted). But the court reasonably declined to adopt the Not One Penny standard based on considerable evidence that college sports have retained their distinctive popularity despite an increase in permissible forms of above-COA compensation and benefits.

In defense of its expansive conception of amateurism, the NCAA relies on its survey of 1,100 college sports fans, reflecting that 31.7 percent watch college sports because, inter alia , they "like the fact that college players are amateurs and/or are not paid." The NCAA claims that the district court rejected this survey on "baseless grounds." But it disregards the court’s primary and most compelling reason for dismissing this evidence: The survey results reflect, at most, a consumer preference for "amateurism," but do not capture the effects (if any) that the tested compensation scenarios would have on consumer behavior . See id. at 1079 ("Dr. Isaacson acknowledged that measuring consumer preferences is ‘not the same thing’ as measuring future consumer behavior, and that he did not do any work to measure any relationship between the two." (internal citation omitted)). The NCAA does not deny this flaw in its survey evidence.

The district court offered another sound reason to reject the NCAA’s survey evidence: The survey’s use of the phrase "amateurs and/or not paid" made its responses "hopelessly ambiguous." Id. at 1078. In so finding, the district court did not, as the NCAA complains, "inject ambiguity into a commonplace term." Amateurism does not have a fixed definition, as NCAA officials themselves have conceded. See, e.g. , id. at 1070–71 ("Mike Slive, who served as commissioner of the SEC, one of the Power Five, ... testified that amateurism is ‘just a concept that I don’t even know what it means. I really don’t.’ " (internal citation omitted)); see also O’Bannon II , 802 F.3d at 1083 (Thomas, C.J., dissenting) (referring to amateurism as a "nebulous concept prone to ever-changing definition"). Survey respondents who selected "amateurs and/or not paid" may have very well equated amateurism with student status, irrespective of whether those students receive compensation for athletics. See Alston , 375 F. Supp. 3d at 1082 (acknowledging defense witness testimony that "consumers’ perception that student-athletes are, in fact, students" drives consumer demand for D1 basketball and FBS football). Given this lack of clarity, the district court reasonably concluded that the NCAA’s survey results were of limited evidentiary value.

Finally, the district court properly considered whether the challenged rules themselves, rather than hypothetical alternatives, have procompetitive benefits. As both parties recognize, the proper "inquiry under the Rule of Reason is: What procompetitive benefits are served by the NCAA’s [challenged] rule[s]?" See O’Bannon II , 802 F.3d at 1073 n.17. As we have recounted, the district court gave reasoned consideration to the procompetitive effects achieved by each type of challenged rule, ultimately concluding that the NCAA "sufficiently show[ed] a procompetitive effect of some aspects of the challenged compensation scheme ," but not all. Alston , 375 F. Supp. 3d at 1103 (emphasis added). By contrast, in O’Bannon , the district court erred at step two because it considered the procompetitive benefits of hypothetical limits on large amounts of compensation. See O’Bannon II , 802 F.3d at 1073 n.17 ("During the second step, the district court could only consider the benefits of the NCAA’s existing rule prohibiting NIL payments—it could not consider the potential benefits of an alternative rule (such as capping large payments)."). Here, the NCAA has conceded that its rules, in part, "prevent the receipt of unlimited pay" unrelated to education. Dr. Isaacson also acknowledged that the challenged rules prohibit unlimited pay. Thus, the court did not err in assessing whether such rules have procompetitive effects.

In short, the district court fairly found that NCAA compensation limits preserve demand to the extent they prevent unlimited cash payments akin to professional salaries, but not insofar as they restrict certain education-related benefits.

C

At the Rule of Reason’s third step, it is Student-Athletes’ burden to "make a strong evidentiary showing" that their proposed LRAs to the challenged scheme "are viable." Id. at 1074. "[T]o be viable," an alternative "must be ‘virtually as effective’ in serving the procompetitive purposes of the NCAA’s current rules, and ‘without significantly increased cost.’ " Id. (quoting Cty. of Tuolumne v. Sonora Cmty. Hosp. , 236 F.3d 1148, 1159 (9th Cir. 2001) ). Where "a restraint is patently and inexplicably stricter than is necessary to accomplish all of its procompetitive objectives, an antitrust court can and should invalidate it and order it replaced with [an LRA]." Id. at 1075.

The LRA identified by the district court would prohibit the NCAA from (i) capping certain education-related benefits and (ii) limiting academic or graduation awards or incentives below the maximum amount that an individual athlete may receive in athletic participation awards, while (iii) permitting individual conferences to set limits on education-related benefits. See Alston , 375 F. Supp. 3d at 1087. The district court did not clearly err in determining that this LRA would be " ‘virtually as effective’ in serving the procompetitive purposes of the NCAA’s current rules," and may be implemented without "significantly increased cost." See O’Bannon II , 802 F.3d at 1074 (internal citation omitted).

1

The district court reasonably concluded that uncapping certain education-related benefits would preserve consumer demand for college athletics just as well as the challenged rules do. Such benefits are easily distinguishable from professional salaries, as they are "connect[ed] to education"; "their value is inherently limited to their actual costs"; and "they can be provided in kind, not in cash." Alston , 375 F. Supp. 3d at 1102. And, as already detailed, the record furnishes ample support for the district court’s finding that the provision of education-related benefits has not and will not repel college sports fans.

The district court drew an apt analogy between the LRA upheld in O’Bannon II and the LRA it identified here: Both athletic scholarships for the COA and education-related benefits "cover legitimate education-related costs." Id. at 1105. Indeed, in affirming the district court’s order insofar as it raised the grant-in-aid cap to the COA, the O’Bannon II panel noted Dr. Emmert’s testimony that this alternative would not harm demand "because all the money given to students would be going to cover their ‘legitimate costs’ to attend school." O’Bannon II , 802 F.3d at 1075. In reference to this litigation, Dr. Emmert similarly announced the NCAA’s approval of the court’s order to the extent that it would foster competition among conferences and schools "over who can provide the best educational experience"—"an inherently good thing." Associated Press, Emmert: Ruling reinforced fundamentals of NCAA , ESPN, Apr. 4, 2019, available at https://tinyurl.com/emmert-NCAA/.

Dr. Emmert’s comment is consistent with the record. As in O’Bannon II , the NCAA presented no evidence that demand will suffer if schools are free to reimburse education-related expenses of inherently limited value. Indeed, its evidence was to the contrary. For instance, in testifying about a University of Nebraska program that permits student-athletes to receive up to $7,500 in post-eligibility aid (for study-abroad expenses, scholarships, and internships), the University’s former chancellor conceded that such benefits "relate to the educational enterprise" and, thus, do not erode demand. When asked about the propriety of above-COA compensation, the current MAC commissioner similarly testified that the "key" is "linking" payments to the "pursuit of the educational opportunities of the individual involved." The LRA fashioned by the district court achieves that link.

In light of this evidence, the district court reasonably concluded that market competition in connection with education-related benefits will only reinforce consumers’ perception of student-athletes as students, thereby preserving demand. See Alston , 375 F. Supp. 3d at 1089 (observing that NCAA’s "own witnesses" testified that "consumer demand for [D1] basketball and FBS football is driven largely by consumers’ perception that student-athletes are, in fact, students").

Moreover, no evidence in the record substantiates the NCAA’s concerns that certain benefits permissible under the LRA, if uncapped, will become vehicles for payments that are virtually indistinguishable from a professional’s salary. These concerns are premised on an unreasonably expansive reading of the injunction, including its requirement that the NCAA permit reimbursement for "tangible items not included in the [COA] calculation but nonetheless related to the pursuit of academic studies." In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. , 2019 WL 1593939, at *1. We construe injunctions in "context" and "so as to avoid ... absurd result[s]." Gathright v. City of Portland , 439 F.3d 573, 581 (9th Cir. 2006). The context here makes plain that it "cannot have been the district court’s intent," id. , for uncapped benefits to be vehicles for unlimited cash payments. Instead, it expressly envisioned "non-cash education-related benefits" for "legitimate education-related costs," not luxury cars or expensive musical instruments for students who are not studying music. Alston , 375 F. Supp. 3d at 1105 (emphasis added). Thus, properly construed, the injunction does not permit the type of unlimited cash payments asserted by the NCAA. Further, as the district court properly concluded, it is doubtful that a consumer could mistake a post-eligibility internship for a professional athlete’s salary, where the former is necessarily divorced from participation in college athletics. The NCAA’s challenges to the evidence underlying this LRA are likewise unavailing. To be sure, neither the survey nor Dr. Rascher’s observations regarding the Nebraska program purport to reflect the effect that nationwide education-related benefits, implemented in the aggregate , would have on consumer demand. But the district court did not rely exclusively on this evidence. Under the deferential standard of review required here, we must examine the record "in its entirety." Alexander , 106 F.3d at 877. The NCAA fails to explain why the cumulative evidence, which included demand analyses regarding the growth of NCAA revenue alongside the expansion of SAF and AEF payments for education-related expenses, was insufficient.

And though the record does not reflect whether an athlete has ever received $5,600 in aggregate athletic participation awards, the district court reasonably concluded that permitting student-athletes to receive up to that amount in academic or graduation awards and incentives will not erode consumer demand. See Alston , 375 F. Supp. 3d at 1072 (citing Dr. Elzinga’s testimony that a player on a successful team could obtain $5,600 in cumulative awards under existing rules). The district court had before it (and fairly credited) evidence that demand would withstand even higher caps on such awards and incentives. See id. at 1080 (discussing Student-Athletes’ survey, which indicated that consumers would continue to view and attend college sports events even if student-athletes received academic or graduation incentive payments of up to $10,000); see also id. at 1074, 1102, n.42 (observing that NCAA’s 30(b)(6) witness was unable to explain the NCAA’s reason for limiting Senior Scholar-Athlete Awards to two students per year and a value of $10,000). The NCAA’s objection to the $5,600 cap rings especially hollow considering that it does not cap individual academic or graduation awards drawn from the AEF or SAF. See id. at 1072 n.15.

Finally, the NCAA contends that the district court engaged in improper judicial price setting by tying the cap on academic and graduation awards and incentives to the cap on aggregate athletic participation awards. The Supreme Court has remarked that courts are "ill suited" to identify terms of dealing between competitors, including a product’s "proper price." Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP , 540 U.S. 398, 408, 124 S.Ct. 872, 157 L.Ed.2d 823 (2004). But the district court did not fix the value of these academic awards: The task of setting their value to protect demand, by adjusting the aggregate value of athletic participation awards, remains in the NCAA’s court. See Alston , 375 F. Supp. 3d at 1107.

2

The district court did not clearly err in finding that this LRA will not result in significantly increased costs. The district court reasoned that enjoining NCAA caps on most education-related benefits will actually save the NCAA resources that it would have otherwise spent on enforcing those caps. Id. at 1090. Commonsense supports that determination, as does the record.

Moreover, though the injunction permits the NCAA to regulate, to an extent, academic and graduation awards and incentives, and conferences to regulate all education-related benefits, there is no reason to believe that such regulation, if pursued, will result in significantly increased costs. The NCAA does not dispute that it and its conferences have existing rulemaking and enforcement infrastructure to achieve such regulation. See id. at 1090 n.32 (noting NCAA’s recent creation of enforcement body to adjudicate violations of "complex" NCAA rules, including the "prioritiz[ation of] academics and the well-being of college athletes" (internal citation omitted); see also id. (noting that conferences are legislative bodies under the Bylaws).

The court’s findings at step three are supported by the record, and certainly not clearly erroneous.

V

The final question remaining is whether the district court’s injunction goes too far or not far enough in enjoining the NCAA’s unlawful conduct. In the NCAA’s view, the injunction is impermissibly vague, in violation of Federal Rule of Civil Procedure 65(d) (" Rule 65(d)"), and usurps the association’s role as the "superintend[ent]" of college sports, O’Bannon II , 802 F.3d at 1074. On cross-appeal, Student-Athletes urge that the district court should have enjoined all NCAA compensation limits, including those on payments untethered to education. In our view, the district court struck the right balance in crafting a remedy that both prevents anticompetitive harm to Student-Athletes while serving the procompetitive purpose of preserving the popularity of college sports. Thus, we neither vacate nor broaden the injunction, but affirm.

A

Rule 65(d) reflects the "basic principle" that "those against whom an injunction is issued should receive fair and precisely drawn notice of what the injunction actually prohibits." Fortyune v. Am. Multi-Cinema, Inc. , 364 F.3d 1075, 1086–87 (9th Cir. 2004) (internal citation omitted). "[W]e will not set aside injunctions under Rule 65(d) ‘unless they are so vague that they have no reasonably specific meaning.’ " Id. at 1087 (internal citation omitted). The challenged injunction clears this hurdle.

The district court enjoined the NCAA from limiting enumerated "compensation and benefits related to education," In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. , 2019 WL 1593939, at *1 (listing computers, science equipment, musical instruments, etc.). The NCAA does not claim confusion as to the meaning of any of these items. Instead, it stakes its Rule 65(d) objection on the injunction’s reference to "other tangible items not included in the [COA] but nonetheless related to the pursuit of academic studies." Id . When read in context, following a list of specific types of education-related equipment, this language is reasonably specific. And unlike in Columbia Pictures Industries, Inc. v. Fung , a copyright infringement case on which the NCAA relies, the injunction here does not make cryptic reference to "general[ ]" or "wide[spread]" understanding and knowledge of technical terms. 710 F.3d 1020, 1048 (9th Cir. 2013).

Nor did the district court impermissibly wrest control of college sports from the NCAA by empowering itself to determine the types of benefits that qualify as "related to" education, instead of "leaving th[at] task" to "the institutions experienced in and responsible for providing education." The NCAA does not (nor can it reasonably) dispute that the benefits enumerated in the injunction are plainly related to academics. What is more, the injunction invites the NCAA to promulgate a definition of "related to education," based on its institutional expertise, subject to the court’s approval. See In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. , 2019 WL 1593939, at *1. This allowance does not constitute judicial usurpation by a long shot.

In sum, we uphold the injunction against the NCAA’s challenges.

B

If the district court had concluded, as Student-Athletes contend, that NCAA limits on compensation unrelated to education unreasonably restrain trade, then it should have enjoined those limits. See 15 U.S.C. §§ 4, 25 (conferring jurisdiction on federal courts to "prevent and restrain violations" of antitrust law); see also Ford Motor Co. v. United States , 405 U.S. 562, 577–78, 92 S.Ct. 1142, 31 L.Ed.2d 492 (1972) ("Antitrust relief should unfetter a market from anticompetitive conduct ." (emphasis added)). The problem for Student-Athletes is that the court did not conclude as much; instead, it determined that NCAA limits on education -related compensation are the only challenged rules that flunk the Rule of Reason.

Although the district court found that all the challenged rules have an anticompetitive effect, Alston , 375 F. Supp. 3d at 1067–70, a finding of anticompetitive harm at step one does not end the inquiry. A defendant may escape antitrust liability despite inflicting harm if a court determines that the restraint has a procompetitive effect, and a proposed LRA eliminating that restraint is not viable. See, e.g. , O’Bannon II , 802 F.3d at 1070, 1076–79 (finding that rules prohibiting NIL compensation had significant anticompetitive effects, but vacating portion of injunction requiring deferred compensation for NILs after concluding that this alternative was not a viable LRA).

As previously stated, the district court concluded, at step two, that the NCAA satisfied its burden of showing that "[r]ules that prevent unlimited payments"—"unrelated to education" and "akin to salaries seen in professional sports leagues"—serve the procompetitive end of distinguishing college from professional sports. Alston , 375 F. Supp. 3d at 1083. And at step three, it rejected Student-Athletes proposed LRAs, which would have eliminated such limits, reasoning:

[A]t least some conferences would allow their schools to offer student-athletes unlimited cash payments that are unrelated to education. Such payments could be akin to those observed in professional sports leagues. Payments of that nature could diminish the popularity of college sports as a product distinct from professional sports.

Id. at 1087. Contrary to Student-Athletes’ understanding, this analysis reflects the judgment that limits on cash compensation unrelated to education do not , on this record, constitute anticompetitive conduct and, thus, may not be enjoined.

This judgment was adequately reasoned and rests on neither factual nor legal error. The district court acknowledged the theoretical possibility that "conference officials, as rational economic actors, would not act contrary to their members’ aggregate economic interests" by paying demand-reducing levels of compensation. Id. But it reasonably perceived a risk of "miscalculations" by conferences during an "inevitable trial-and-error phase." Id. The district court did not clearly err in declining to assume that conferences, in reality, would act rationally. The record indicates that the Power Five schools have exercised their autonomy in recent years to expand benefits unrelated to education and that conferences and schools have provided largely discretionary SAF and AEF payments for a wide range of expenses unrelated to education—both without harming consumer demand. But the district court reasonably concluded that this evidence may not reliably indicate that individual conferences would regulate payments in a demand-preserving manner absent any restrictions: The autonomy structure permits the Power Five to collectively adopt compensation-related legislation, in line with O’Bannon II ’s guidance that some degree of "mutual agreement" is necessary to make the college sports product available. See 802 F.3d at 1069 (quoting Bd. of Regents , 468 U.S. at 102, 104 S.Ct. 2948 ). And the NCAA currently limits the use of SAF funds to payments that are distinguishable from a professional’s salary in that they "meet[ ] financial needs that arise in conjunction with participation in intercollegiate athletics, enrollment in academic curriculum or to recognize academic achievement."

Student-Athletes’ claims of legal error are likewise unpersuasive. They cite no support for their position that a court "should not simply import the [LRA] as its injunction." Indeed, O’Bannon II holds otherwise: "Where, as here, a restraint is patently and inexplicably stricter than is necessary to accomplish all of its procompetitive objectives, an antitrust court can and should invalidate it and order it replaced with a [viable LRA] ." Id. at 1075 (emphasis added).

Finally, Student-Athletes argue that the NCAA may no longer rely on O’Bannon II ’s conclusion that NCAA limits on cash payments untethered to education are critical to preserving the distinction between college and professional sports now that it has "endorse[d]" the very "same NIL benefits" at issue there. This argument is premature. As it stands, the NCAA has not endorsed cash compensation untethered to education; instead, it has undertaken to comply with the FPP Act in a manner that is consistent with O’Bannon II —that is, by loosening its restrictions to permit NIL benefits that are "tethered to education." Fed. and State Leg. Working Grp. Report 4 (Oct. 23, 2019), available at https://tinyurl.com/working-grp-report; see also Test. of Dr. Mark Emmert 6 (Feb. 11, 2020), available at https://tinyurl.com/Emmert-Test-y. Accordingly, we disagree that the NCAA’s response to the FPP Act militates in favor of enjoining all NCAA compensation limits.

VI

To repeat my observation in O’Bannon II : "The national debate about amateurism in college sports is important. But our task as appellate judges is not to resolve it. Nor could we. Our task is simply to review the district court judgment through the appropriate lens of antitrust law and under the appropriate standard of review." O’Bannon II , 802 F.3d at 1083 (Thomas, C.J., concurring in part and dissenting in part).

For the foregoing reasons, we hold that the district court properly concluded that NCAA limits on education-related benefits do not "play by the Sherman Act’s rules." Id. at 1079. Accordingly, we affirm its liability determination and injunction in all respects.

AFFIRMED.

M. SMITH, Circuit Judge, concurring:

Because I am bound by our decision in O’Bannon v. NCAA (O’Bannon II ), 802 F.3d 1049 (9th Cir. 2015), I join the panel opinion in full. I write separately to express concern that the current state of our antitrust law reflects an unwitting expansion of the Rule of Reason inquiry in a way that deprives the young athletes in this case (Student-Athletes) of the fundamental protections that our antitrust laws were meant to provide them.

Student-Athletes are talented, hardworking individuals who have dedicated their young lives to excelling in specific sports. As amici describe, Student-Athletes work an average of 35–40 hours per week on athletic duties during their months-long athletic seasons, and most work similar hours during the off-season to stay competitive. At the same time, most of them do their best to succeed academically, managing to devote on average another 40 hours per week to classes and study. Nevertheless, their coaches and others in the Division 1 ecosystem make sure that Student-Athletes put athletics first, which makes it difficult for them to compete for academic success with students more focused on academics. They are often forced to miss class, to neglect their studies, and to forego courses whose schedules conflict with the sports in which they participate. In addition to lessening their chances at academic success because of the time they must devote to their sports obligations, Student-Athletes are often prevented from obtaining internships or part-time paying jobs, and, as a result, often lack both income and marketable work experience. Meanwhile, the grueling hours and physical demands of college sports carry significant health risks, such as sleep deprivation, stress, broken bones, and even potential brain damage. Despite their best efforts, however, fewer than 5% of Student-Athletes will ever play at a professional level, and most of those lucky few will stay in the pros only a few short years. In short, the college years are likely the only years when young Student-Athletes have any realistic chance of earning a significant amount of money or achieving fame as a result of their athletic skills.

For all their dedication, labor, talent, and personal sacrifice, Student-Athletes go largely uncompensated. They may receive tuition for an academic experience that they cannot take full advantage of, minimal living expenses, and some lavish perks that do nothing for their present or future financial security. However, that is not because their athletic services have little value. On the contrary, the NCAA and Division 1 universities make billions of dollars from ticket sales, television contracts, merchandise, and other fruits that directly flow from the labors of Student-Athletes. A number of Division 1 head football coaches take home multimillion-dollar salaries that exceed those of many NFL coaches. Moreover, contrary to the NCAA’s representations about the importance of "amateurism," the evidence in this case shows that college sports viewership has only increased since we reduced some limitations on student-athlete compensation in O’Bannon II . See Panel Op. at 11–13.

My reaction to our application of federal antitrust law to the case of the Student-Athletes is similar Justice Alito’s reaction to the majority’s view in Collins v. Virginia , 584 U.S. ––––, 138 S. Ct. 1663, 201 L.Ed.2d 9 (2018). Said he: "An ordinary person of common sense would react to the Court’s decision the way Mr. Bumble famously responded when told about a legal rule that did not comport with the reality of everyday life. If that is the law, he exclaimed, ‘the law is a ass—a idiot.’ " Id. at 1681 (Alito, J., dissenting) (quoting C. Dickens, Oliver Twist 277 (1867)).

The treatment of Student-Athletes is not the result of free market competition. To the contrary, it is the result of a cartel of buyers acting in concert to artificially depress the price that sellers could otherwise receive for their services. Our antitrust laws were originally meant to prohibit exactly this sort of distortion.

The Sherman Act and related antitrust laws were designed to preserve our economic freedom. United States v. Topco Assocs., Inc. , 405 U.S. 596, 610, 92 S.Ct. 1126, 31 L.Ed.2d 515 (1972). Under those laws,

the freedom guaranteed each and every business, no matter how small, is the freedom to compete—to assert with vigor, imagination, devotion, and ingenuity whatever economic muscle it can muster. Implicit in such freedom is the notion that it cannot be foreclosed with respect to one sector for the economy because certain private citizens or groups believe that such foreclosure might promote greater competition in a more important sector of the economy.

Id. The Sherman Act thus "protect[s] the economic freedom of participants in the relevant market." Am. Ad Mgmt., Inc. v. Gen. Tel. Co. of Cal. , 190 F.3d 1051, 1057 (9th Cir. 1999) (quoting Associated Gen. Contractors of Cal., Inc. v. Cal. State Council of Carpenters , 459 U.S. 519, 538, 103 S.Ct. 897, 74 L.Ed.2d 723 (1983) ). Those protections extend to sellers of goods and services—such as Student-Athletes—to the same extent they do buyers, consumers, or competitors. Mandeville Island Farms, Inc. v. Am. Crystal Sugar Co. , 334 U.S. 219, 235, 68 S.Ct. 996, 92 L.Ed. 1328 (1948). "The Act is comprehensive in its terms and coverage, protecting all who are made victims of the forbidden practices by whomever they may be perpetrated." Id. (emphasis added).

Section 1 of the Sherman Act, at issue here, prohibits agreements that unreasonably restrain trade. 15 U.S.C. § 1 ; Standard Oil Co. of N.J. v. United States , 221 U.S. 1, 58, 31 S.Ct. 502, 55 L.Ed. 619 (1911). In evaluating alleged violations of Section 1 that fall outside the bounds of several now-established per se rules, courts apply the Rule of Reason to determine the effect of a given restraint on competition. "[T]he inquiry mandated by the Rule of Reason is whether the challenged agreement is one that promotes competition or one that suppresses competition." Nat’l Soc’y of Prof’l Eng’rs v. United States , 435 U.S. 679, 691, 98 S.Ct. 1355, 55 L.Ed.2d 637 (1978). Importantly, it is not the purpose of the Rule of Reason analysis "to decide whether a policy favoring competition is in the public interest, or in the interest of the members of an industry. Subject to exceptions defined by statute, that policy decision has been made by the Congress." Id. at 692, 98 S.Ct. 1355.

The Rule of Reason entails a three-step analysis, of which the starting point is to identify the market in which the restraint occurs. See Big Bear Lodging Ass’n v. Snow Summit, Inc. , 182 F.3d 1096, 1104–05 (9th Cir. 1999). At Step One, the "plaintiff bears the initial burden of showing that the restraint produces significant anticompetitive effects within" that market. O’Bannon II , 802 F.3d at 1070 (quoting Tanaka v. Univ. of S. Cal. , 252 F.3d 1059, 1063 (9th Cir. 2001) ). If the plaintiff meets that burden, at Step Two, "the defendant must come forward with evidence of the restraint’s procompetitive effects." Id. (quoting Tanaka , 252 F.3d at 1063 ). Finally, at Step Three, "the plaintiff must ... show that any legitimate objectives can be achieved in a substantially less restrictive manner." Id. (quoting Tanaka , 252 F.3d at 1063 ).

Despite confining the Step One analysis of anticompetitive effects to the defined market, courts have not consistently limited the scope of the Step Two analysis in the same way. Some, including our court, have permitted defendants to offer procompetitive effects in a collateral market as justification for anticompetitive effects in the defined market. In NCAA v. Board of Regents of Univ. of Oklahoma (Board of Regents ), 468 U.S. 85, 104 S.Ct. 2948, 82 L.Ed.2d 70 (1984), for example, the Supreme Court considered whether preserving demand for tickets to live college football games could justify anticompetitive restraints in the market for live college football television. Id. at 95–96, 115–17, 104 S.Ct. 2948. The district court defined the relevant market at Step One as "live college football television." Id. at 95, 104 S.Ct. 2948. The NCAA had restrained competition in this market by fixing the price of telecasts, negotiating exclusive contracts with two television networks, and artificially limiting the number of televised games. Id. at 96, 104 S.Ct. 2948. Among other alleged procompetitive justifications, all of which the Court ultimately rejected, the NCAA argued that its television plan promoted consumer demand for live attendance at college football games. Id. at 115, 104 S.Ct. 2948. The Court rejected this argument for three reasons: (1) individual schools could protect live attendance at the specific game being televised by negotiating a regional blackout, without acting in concert with other schools; (2) no evidence supported the NCAA’s theory that limiting televised games actually promoted live attendance, especially since games would still be broadcast at all hours of the day; and (3) the NCAA’s live attendance theory was "not based on a desire to maintain the integrity of college football as a distinct and attractive product, but rather on a fear that ... ticket sales for most college games are unable to compete in a free market"—"a justification that is inconsistent with the basic policy of the Sherman Act." Id. at 115–17, 104 S.Ct. 2948. The Supreme Court did not, however, say that the live attendance justification failed because courts categorically cannot consider procompetitive benefits outside the defined market.

Our relevant precedents follow a similar analysis. In O’Bannon II , we held that preserving consumer demand for college sports was a legitimate procompetitive justification for anticompetitive restraints on compensation for student-athletes’ names, images, and likenesses in the market among colleges for student-athletes’ services. 802 F.3d at 1069–73. The district court had defined the relevant market at Step One as the "college education market," "wherein colleges compete for the services of athletic recruits by offering them scholarships and various amenities, such as coaching and facilities." Id. at 1070. The NCAA had restrained competition in this market by preventing member schools from paying student athletes for the use of their names, images, and likenesses. Id. Contrary to two of the NCAA’s proffered justifications, we accepted the district court’s factual determinations that the restraint did "not promote competitive balance," and did "not increase output in the college education market." Id. at 1072. We also rejected the NCAA’s argument that, by preserving the character of college sports, the restraint " ‘widen[ed]’ the choices ‘available to athletes.’ " Id. (quoting Board of Regents , 468 U.S. at 102, 104 S.Ct. 2948 ). "As the district court found, it is primarily ‘the opportunity to earn a higher education’ that attracts athletes to college sports rather than professional sports, and that opportunity would still be available to student-athletes if they were paid some compensation in addition to their athletic scholarships." Id. at 1073. Yet, without tying the Step 2 analysis to the "college education market," we held that the NCAA had demonstrated that the restraint served the procompetitive purpose of preserving "the amateur nature of collegiate sports [that] increases their appeal to consumers." Id. Accordingly, we proceeded to Rule of Reason Step Three, wherein we upheld the district court’s less restrictive alternative of allowing grant-in-aid up to the full cost of attendance, but we vacated the district court’s less restrictive alternative of allowing "small" amounts of deferred cash compensation as incompatible with amateurism. Id. at 1074–79.

Other courts, however, have rejected procompetitive justifications outside of the defined market. For example, in Smith v. Pro Football, Inc. , 593 F.2d 1173 (D.C. Cir. 1978), a former NFL player challenged rules governing the draft of graduating college players under which "no team was permitted to negotiate prior to the draft with any [eligible] player ... and no team could negotiate with (or sign) any player selected by another team in the draft." Id. at 1176. The D.C. Circuit affirmed the finding that the draft had anticompetitive effects. The draft eliminated competition by "inescapably forc[ing] each seller of football services to deal with one, and only one buyer, robbing the seller, as in any monopsonistic market, of any real bargaining power." Id. at 1185.

At Step Two of the Rule of Reason analysis, the NFL asserted that the draft rules were procompetitive because they promoted "competitive balance" among the league’s teams, in turn "producing better entertainment for the public, higher salaries for the players, and increased financial security for the clubs." Id. at 1186. The court rejected those justifications because they did not have procompetitive effects in the market for players’ services . "The draft is ‘procompetitive,’ if at all, in a very different sense from that in which it is anticompetitive." Id. "[W]hile [the draft] may heighten athletic competition and thus improve the entertainment product offered to the public, [it] does not increase competition in the economic sense of encouraging others to enter the market and to offer the product at lower cost." Id. The court concluded that the draft’s anticompetitive and procompetitive effects were "not comparable," and thus it was "impossible to ‘net them out’ in the usual rule-of-reason balancing." Id.

Despite its ruling in Board of Regents , the Supreme Court has not squarely addressed the proper scope of the Step Two analysis. And, although we conducted a similar analysis in O’Bannon II , neither have we. In my view, the underlying purpose of the Sherman Act—promoting competition—counsels in favor of conducting a more limited Rule of Reason analysis, as the court in Smith did. Realistically, the Rule of Reason analysis is judicially administrable only if it is confined to the single market identified from the outset. If the purpose of the Rule of Reason is to determine whether a restraint is net procompetitive or net anticompetitive, accepting procompetitive effects in a collateral market disrupts that balancing. It weakens antitrust protections by permitting defendants to rely on a broader array of justifications that promote competition, if at all, in collateral markets where the restraint under analysis does not occur.

Jurists faced with weighing the anticompetitive effects in one market with the procompetitive effects in another cannot simply "net them out" mathematically. Smith , 593 F.2d at 1186. Rather, courts employing a cross-market analysis must—implicitly or explicitly—make value judgments by determining whether competition in the collateral market is more important than competition in the defined market. As the Supreme Court has warned, this is not what the antitrust laws invite courts to do. "If a decision is to be made to sacrifice competition in one portion of the economy for greater competition in another portion this too is a decision that must be made by Congress and not by private forces or by the courts. Private forces are too keenly aware of their own interests in making such decisions and courts are ill-equipped and ill-situated for such decisionmaking." Topco , 405 U.S. at 611, 92 S.Ct. 1126.

Consider this case. The district court accepted the relevant market as that for Student-Athletes’ "labor in the form of athletic services in men’s and women’s Division I basketball and FBS football," in which Student-Athletes "sell their athletic services to the schools that participate in Division I basketball and FBS football in exchange for grants-in-aid and other benefits and compensation permitted by NCAA rules." In re NCAA Athletic Grant-In-Aid Cap Antitrust Litig. (Alston ), 375 F. Supp. 3d 1058, 1067 (N.D. Cal. 2019). At Step One, the district court found that Student-Athletes had established significant anticompetitive effects in the market for their athletic services. The court concluded that the NCAA rules "have the effect of artificially compressing and capping student-athlete compensation and reducing competition for student-athlete recruits by limiting the compensation offered in exchange for their athletic services." Id. at 1068.

At Step Two, the court did not limit its consideration to the procompetitive effects of the compensation limits in the market for Student-Athletes’ athletic services. Rather, it found that certain of the compensation limits are procompetitive because they drive consumer demand for college sports by distinguishing collegiate from professional athletics. Id. at 1083. In other words, the court found that limiting Student-Athletes’ pay in the market for their services was justified because that restraint drove demand for the distinct product of college sports in the consumer market for sports entertainment. The court did not require that the NCAA prove that this impact on consumer demand had a corollary procompetitive impact on the market for Student-Athletes’ services, that it "increase[d] output" or " ‘widen[ed]’ the choices ‘available to athletes.’ " O’Bannon II , 802 F.3d at 1072 (quoting Board of Regents , 468 U.S. at 102, 104 S.Ct. 2948 ). The court did not require that the NCAA prove its compensation rules, within the defined market, "increase competition in the economic sense of encouraging others to enter the market to offer the product at lower cost." Smith , 593 F.2d at 1186. It was enough for the NCAA to meet its Step Two burden that it could show (however feebly) a procompetitive effect in a collateral market.

Although the district court correctly applied our precedents, the result of this analysis seems to erode the very protections a Sherman Act plaintiff has the right to enforce. Here, Student-Athletes are quite clearly deprived of the fair value of their services. Alston , 375 F. Supp. 3d at 1068. As the district court found, while the NCAA and its conferences generate billions in revenue from college sports, they "have monopsony power to restrain student-athlete compensation in any way and at any time they wish, without any meaningful risk of diminishing their market dominance." Id. at 1063, 1070. Under the Rule of Reason analysis we affirm today, so long as the NCAA cites consumer demand for college sports, we allow it to artificially suppress competition for collegiate athletes’ services by limiting their compensation. Instead of requiring the NCAA to explain how those limits promote schools’ competition for athletes, we leave Student-Athletes with little recourse under the antitrust laws. Student-Athletes are thus denied the freedom to compete and, in turn, "of compensation they would receive in the absence of the restraints." Id. at 1068.

Our Rule of Reason framework has shifted toward this cross-market analysis without direct consideration or a robust justification. It may be that scholars or litigants can develop a purely economic, mathematically-defensible method for cross-market analysis that does not depend on policy judgments that our antitrust laws never meant to delegate to the courts. But we do not currently have such a method, and it may equally be the case that no such method is possible or desirable.

Lacking a robust justification, I fear that our cross-market Rule of Reason analysis frustrates the very purpose of the antitrust laws, in this case to the great detriment of Student-Athletes. I hope our court will reconsider this issue in a case that squarely raises it.