Showing 9 of 13 Publications by Daniel Lyons

Amicus of ICLE and Law & Economics Scholars to the 2nd Circuit in Giordano v Saks

Amicus Brief INTEREST OF AMICI CURIAE[1] The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the . . .


The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy.  ICLE promotes the use of law and economics methodologies to inform public policy debates and has longstanding expertise in antitrust law.  ICLE has an interest in ensuring that antitrust promotes the public interest by remaining grounded in sensible legal rules informed by sound economic analysis.

Amici also include fifteen scholars of antitrust, law, and economics at leading universities and research institutions across the United States.  Their names, titles, and academic affiliations are listed in Appendix A.  All have longstanding expertise in antitrust law and economics.

Amici respectfully submit that this brief will aid the Court in reviewing the order of dismissal by explaining that the district court properly held, on the pleadings, that the restraint at issue is ancillary and thus that per se treatment is inappropriate.  The restraint furthers Saks’s procompetitive goal of creating a strong and stable luxury brand through collaboration with the Brand Defendants.  Treating such a restraint as per se unlawful, as Plaintiffs ask this Court to do, would stifle the type of legitimate cooperation that facilitates output and would ultimately harm consumers.  Amici also explain why Plaintiffs and several of their amici, including the United States, make foundational errors of law and economics in arguing that ancillarity is an affirmative defense that may not be resolved on the pleadings.


Saks and the Brand Defendants are well-known luxury retail brands.  As luxury retailers, their business models depend on developing and maintaining a distinct, exclusive brand to differentiate their products from the lower-priced goods sold by mass-market retailers.  A primary way in which they define and protect their brands is by cultivating a premium shopping experience for customers that promotes “an atmosphere of exclusivity and opulence surrounding . . . luxury products.”  Compl. ¶ 33.  To that end, Saks and the Brand Defendants have for years collaborated through “store-within-a-store” arrangements: Saks allows the Brand Defendants to set up boutiques and concessions within Saks’s stores, which in turn helps all involved grow their customer base, augment their luxury brand status, and sell more products.  This “store-within-a-store” model not only expands customer product choice within a single retail establishment, resulting in a better shopping experience, but also creates additional jobs at Brand Defendants’ concessions in Saks’s stores.

Plaintiffs allege that the Brand Defendants agree, as part of their respective partnerships with Saks, not to hire Saks’s own luxury retail employees without the approval of a Saks manager or until six months after the employee leaves Saks.  Plaintiffs argue that these alleged no-hire provisions violate Section 1 of the Sherman Act.  The district court disagreed, concluding that the per se rule could not apply because the no-hire provisions were “ancillary” to a broader procompetitive collaboration between Saks and each of the Brand Defendants, and that Plaintiffs failed to plead a plausible claim under the rule of reason.  That decision is correct and should be affirmed.

First, the alleged no-hire agreements are ancillary to the arrangements between Saks and the Brand Defendants.  Saks invests heavily in its employees.  But without the no-hire provision, Saks would stand to lose those investments as the Brand Defendants could take advantage of their co-location within Saks’s stores to hire away Saks’s best workers, thereby free-riding on Saks’s training.  The alleged no-hire provisions eliminate that powerful economic disincentive and thereby facilitate brand-enhancing, procompetitive store-within-a-store arrangements.  That is all that is required for the agreements to be “ancillary.”  Plaintiffs’ (and their amici’s) insistence on a rigid two-prong test for ancillarity is not only at odds with economic logic but also out of step with this Circuit’s precedent—and, in any event, would not change the result here.

Second, the district court properly resolved ancillarity on the pleadings.  Ancillarity is a threshold inquiry decided at the earliest possible stage of a Section 1 case to determine whether the alleged facts justify departing from the default rule of reason standard.  That is precisely what the district court did here: based on Plaintiffs’ own allegations—including those regarding “a continual risk that the Brand Defendants would use their concessions in Saks stores to recruit employees” (Op. 32)—the district court ruled that the alleged restraints were ancillary and thus incompatible with per se condemnation.  Contrary to Plaintiffs’ argument, the district court did not improperly resolve any factual inferences.  The court considered the Complaint in its entirety and determined that Plaintiffs did not state a plausible per se claim, just as it was supposed to do before requiring the enormous expense that would result should this kind of “potentially massive factual controversy . . . proceed.”  Bell Atl. Corp. v. Twombly, 550 U.S. 544, 558 (2007).


I.              The Alleged Restraints Are Ancillary To Procompetitive Collaboration

The per se rule is reserved for the most pernicious and anticompetitive restraints.  Before condemning a restraint as per se unlawful, therefore, courts must “have amassed considerable experience with the type of restraint at issue” and be able to “predict with confidence that it would be invalidated in all or almost all instances.”  NCAA v. Alston, 141 S. Ct. 2141, 2156 (2021).  Reserving per se condemnation for that small category of restraints ensures that the antitrust laws do not inadvertently chill procompetitive conduct.  Ancillary restraints do not fit the per se mold because they have a “reasonable procompetitive justification, related to the efficiency-enhancing purposes of [a] joint venture.”  MLB Props., Inc. v. Salvino, Inc., 542 F.3d 290, 339 (2d Cir. 2008) (Sotomayor, J., concurring in the judgment).

Here, any purported no-hire agreements form a key plank of the broader leasing, concession, and distribution arrangements between Saks and the Brand Defendants.  Op. 30-32.  It is beyond dispute that these agreements are procompetitive.  They not only enhance Saks’s and the Brand Defendants’ ability to vigorously compete against other retailers and luxury brands (i.e., increasing output in markets for luxury products) but also create jobs (i.e., increase output in labor markets).  That places the restraint far beyond the per se rule, MLB, 542 F.3d at 339 (Sotomayor, J., concurring in the judgment); only the rule of reason can be used to determine whether the restraint “stimulat[es] competition that [is] in the consumer’s best interest” or has “anticompetitive effect[s] that are harmful to the consumer.”  Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877, 886 (2007).[2]

                  A.            The Alleged No-Hire Agreements Are Facially Procompetitive

A restraint is ancillary where it “could have a procompetitive impact related to the efficiency-enhancing purposes” of a cooperative venture.  MLB, 542 F.3d at 340 (Sotomayor, J., concurring in the judgment); see Polk Bros., Inc. v. Forest City Enters., Inc., 776 F.2d 185, 188-89 (7th Cir. 1985) (restraint is ancillary if it “may contribute to the success of a cooperative venture that promises greater productivity and output”).  Where a restraint is deemed “ancillary to the legitimate and competitive purposes” of a venture, the restraint is presumptively “valid” and must be assessed under the rule of reason.  Texaco Inc. v. Dagher, 547 U.S. 1, 7 (2006).  There is a clear procompetitive rationale for the collaboration arrangement between Saks and the Brand Defendants: the arrangement allows customers to expand their choice in one-stop shopping, and the retailers to offer a wider range of high-end luxury goods.  And it creates a halo effect across the store-within-a-store through proximity and availability of multiple luxury brands.  All of this in turn promotes and enhances the luxury status of Saks and the Brand Defendants alike.  The alleged no-hire restraints enable and are ancillary to that larger endeavor.

As Plaintiffs allege, Defendants each derive much of their respective brand value from their ability to project a “luxury brand[] aura[],” which both entices customers and creates demand for Defendants’ goods “over other, lower-priced goods.”  Compl. ¶¶ 23, 26, 28.  For this reason, Defendants “go[] to great lengths to market” and otherwise “maintain[] their luxury brands’ auras.”  Id. ¶¶ 23, 26.  They “accomplish this feat,” in part, by ensuring that their brick-and-mortar stores provide a “luxury shopping experience[].”  Id. ¶ 27.  Defendants do that with sophisticated “décor and design” and premium “customer service” from skilled employees “who reflect their respective brand images and cultures.”  Id. ¶¶ 27-29.

Store-within-a-store arrangements further enhance the luxury brand shopping experience for both consumers and retailers.  In these arrangements, Saks allows the Brand Defendants to set up mini-stores or concessions within Saks’s large stores.  These arrangements, similar to those used by “[a]lmost all department store chains,” Kinshuk Jerath & Z. John Zhang, Store Within a Store, 47 J. Mktg. Rsch. 748, 748 (2010), are mutually beneficial and procompetitive.  The presence of the popular luxury brands helps draw brand-loyal customers into Saks, thus increasing foot traffic and broadening Saks’s customer reach—directly boosting sales output.  Compl. ¶ 28; see Jerath & Zhang, supra, at 756-57 (“The introduction of new products through stores within a store can bring new consumers to the store who want to purchase the focal product and also purchase other products.”).  The Brand Defendants benefit from access to Saks’s considerable customer base, Compl. ¶ 28, and their presence also makes possible cross-brand marketing opportunities.  Consumers benefit as well: they have access to a wider array of products, and have it all at hand in a single store.  And they have the benefit of workers highly trained with respect to the luxury goods they sell.  Id. ¶¶ 27-29, 32-34.

But there is a significant practical impediment to allowing stores-within-stores: employee raiding.  Saks invests heavily in its luxury retail employees, providing them with the “extensive training on service, selling, and product-knowledge” required to ensure that they are “knowledgeable about the particular products” for sale “as well as current trends.”  Compl. ¶¶ 32, 34.  Permitting the Brand Defendants to operate inside of Saks stores without restriction would put that investment in immediate danger.  The Brand Defendants would have every incentive to free-ride off of Saks’s investment, observing and hiring Saks’s highly trained luxury retail employees, thereby “tak[ing] advantage of the efforts [Saks] has expended in soliciting, interviewing, and training skilled labor” and “simultaneously inflicting a cost on [Saks] by removing an employee on whom [Saks] may depend.”  Id. ¶ 62.  This risk—and the mistrust it can create—disincentivizes the formation and maintenance of store-within-a-store agreements.

No-hire restraints solve this problem.  By imposing a narrow, time-limited, waivable restriction on the Brand Defendants’ ability to hire Saks employees, Compl. ¶ 92, the alleged no-hire agreements remove a roadblock from the “cooperation underlying the restraint,” which “has the potential to create the efficient production that consumers value,” Premier Elec. Constr. Co. v. Nat’l Elec. Contractors Ass’n, Inc., 814 F.2d 358, 370 (7th Cir. 1987).  In particular, the alleged no-hire restrictions help prevent free-riding by Brand Defendants on Saks’s training.  The agreement encourages Saks to invest in employee development, including by providing specific training on Brand Defendants’ products, and that investment enhances Saks’s ability to sell products from and compete against Brand Defendants’ stand-alone brick and mortar and online stores.  See, e.g., Gregory J. Werden, The Ancillary Restraints Doctrine After Dagher, 8 Sedona Conf. J. 17, 21 (2007).  “[W]ith the restraint,” Saks may “collaborate” with the Brand Defendants “for the benefit of its [customers] without ‘cutting [its] own throat.’”  Aya Healthcare Servs., Inc. v. AMN Healthcare, Inc., 9 F.4th 1102, 1110-11 (9th Cir. 2021) (quoting Polk Bros., 776 F.2d at 189).  As a result, the alleged no-hire restraints are “at least potentially reasonably ancillary to joint, efficiency-creating economic activities.”  Phillips v. Vandygriff, 711 F.2d 1217, 1229 (5th Cir. 1983); cf. Eichorn v. AT&T Corp., 248 F.3d 131, 146-47 (3d Cir. 2001) (“As an ancillary covenant not to compete, the no-hire agreement was reasonable in its restrictions on the plaintiffs’ ability to seek employment elsewhere.”).

The contrary conclusion—that the alleged no-hire restraints are not ancillary—risks stifling competition across the retail economy.  No-hire agreements are merely one of the many ancillary contractual restraints commonly used in store-within-a-store partnerships (exemplified by, for instance, the well-known collaborations between Target and Starbucks or Best Buy and Samsung) to preserve brand integrity, guard against misuse of store space, and safeguard investments in specialized training.  By solving for risks such as employee raiding or damage to property, these restrictions instill confidence in both parties, facilitating the creation of these cooperative ventures in the first place.  Categorizing the alleged no-hire provisions here as per se unlawful could chill a whole spectrum of reasonable ancillary restraints, undermining the careful balance that store-within-a-store arrangements aim to maintain and inhibiting market innovation.  That would be bad for potential employees, who would lose the opportunity to work at stores-within-stores, as well as for consumers, who would lose the convenient access to goods in-store concessions provide.

                   B.            The Rigid Two-Prong Test Advanced By Plaintiffs And Their Amici Is Not The Law, And The Alleged Restraints Here Satisfy It In Any Event

Plaintiffs and their amici resist ancillarity by, in part, insisting upon application of a strict and formalistic test not found in the law of this Circuit or any other.  In their view, an ancillary restraint must be both (1) “subordinate and collateral to a separate legitimate transaction” and (2) “reasonably necessary to achiev[e] that transaction’s procompetitive purpose.”  AOB 34-35.  This rigid two-step test is not the law in this Circuit.  But even if it were, Plaintiffs and amici misconstrue the second prong, improperly transforming it into a strict necessity standard that no circuit has adopted.  Consistent with their evident procompetitive potential, the alleged restraints here amply satisfy the actual test.

Although some courts have moved toward a delineated two-prong standard, this Court has not.  This Court’s leading opinion on ancillarity is then-Judge Sotomayor’s influential concurrence in MLB, in which she observed that a restraint is ancillary where it is “reasonably necessary to achieve any of the efficiency-enhancing benefits of a joint venture.”  542 F.3d at 338 (Sotomayor, J., concurring in the judgment).  She noted no other requirements, invoking Judge Easterbrook’s similar formulation in Polk Bros. that a restraint is ancillary where it “may contribute to the success of a cooperative venture that promises greater productivity and output.”  Id.; Polk Bros., 776 F.2d at 189; see Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210, 229 (D.C. Cir. 1986) (restraint is ancillary when it “appears capable of enhancing the group’s efficiency”).  That approach in turn traces all the way to then-Judge Taft’s seminal United States v. Addyston Pipe & Steel Co. decision, which assessed ancillarity using this same flexible formulation.  See 85 F. 271, 281 (6th Cir. 1898).

Even if the two-prong test advanced by Plaintiffs and their amici did apply, however, they misconstrue the second prong by paying only lip service to a “reasonably necessary” standard and in reality asking this Court to impose a “strictly necessary” test.  Instead of asking whether the restraint “promoted enterprise and productivity”—which is all that is required for a restraint to be “reasonably necessary,” Aya, 9 F.4th at 1110-11—Plaintiffs would require Defendants to show that the “restraint [is] necessary to achieve the business relationship,” AOB 36, such that in its absence, “Saks would terminate or . . . alter its purported collaborative relationships,” NY Br. 29.

No court of appeals has embraced this strict-necessity standard.  In Medical Center at Elizabeth Place, LLC v. Atrium Health System, for instance, the Sixth Circuit considered and rejected it, holding that requiring a defendant to show that a restraint “is necessary” is “too high a standard to determine what qualifies as ‘reasonable.’”  922 F.3d 713, 725 (6th Cir. 2019); see also id. at 726 (observing Judge Sotomayor’s MLB concurrence “categorically rejected” a strict necessity test).  Rather, an ancillary restraint “need not be essential, but rather only reasonably ancillary to the legitimate cooperative aspects of the venture” because “there exists a plausible procompetitive rationale for the restraint.”  Id. (quotation marks omitted).  The Ninth Circuit similarly rejected the United States’ attempt to advance this standard, and instead held in Aya that a no-hire restraint was “properly characterized as ancillary” where it “promoted enterprise and productivity at the time it was adopted.”  9 F.4th at 1111.  And the United States and a different set of plaintiffs recently argued for a strict-necessity test in the Seventh Circuit.  See Br. for the U.S. and the FTC as Amici Curiae Supporting Neither Party at 26, Deslandes v. McDonald’s USA, LLC, Nos. 22-2333 & 22-2334 (7th Cir. Nov. 9, 2022) (arguing no-hire agreement was not ancillary because it “was not necessary to encourage franchisees to sign” franchising agreements).  The panel declined to adopt it, adhering instead to the Polk Bros. test.  See Deslandes v. McDonald’s USA, LLC, 81 F.4th 699, 703 (7th Cir. 2023).

All of these decisions make sense.  The per se rule applies only when a challenged restraint is obviously and clearly anticompetitive, and a restraint that is plausibly part of a procompetitive venture should be judged by “the facts peculiar to the business to which the restraint is applied.”  Bd. of Trade of Chi. v. United States, 246 U.S. 231, 238 (1918).  A contrary decision would discourage competition; strict necessity is not only an unrealistic requirement, as businesses make these decisions ex ante, but also would require them to constantly recalibrate their policies.  The result would be that firms forego potentially procompetitive collaborations, chilling innovative policies and business models.  See Werden, supra, at 23-24 (comprehensive analysis by DOJ economist rejecting strict-necessity test).

Nor is there any legal or logical basis for Plaintiffs’ made-up “tailor[ing]” prong—that a “restraint must be ‘tailored’ to a legitimate objective to qualify as ancillary.”  AOB 35.  Courts routinely reject any “reasonabl[e] tailor[ing]” requirement, because that phrase would not “carr[y] a materially different meaning than ‘reasonably necessary’” and because a restraint “need not satisfy a less-restrictive-means test.”  Aya, 9 F.4th at 1111 & n.5.  A tailoring analysis can be part of the rule-of-reason framework employed after a restraint is deemed ancillary, but it has no role in the ancillarity inquiry itself, which evaluates whether a restraint “should be reviewed under the rule of reason” in the first place.  MLB, 542 F.3d at 341 (Sotomayor, J., concurring in the judgment).  The flaw in Plaintiffs’ argument is underscored by the only case they cite to support their purported tailoring requirement, which did not even involve ancillarity, but instead analyzed whether there was a less restrictive alternative under the rule of reason.  See NCAA v. Bd. of Regents of Univ. of Okla., 468 U.S. 85, 117 (1984); see also Aya, 9 F.4th at 1111 (“[T]he less restrictive alternative analysis falls within the rule-of-reason analysis, not the ancillary restraint consideration.”).

Properly interpreted to require only “reasonable necessity,” the two-prong test is satisfied here on the face of the Complaint.  The alleged restraint is “subordinate and collateral” to a broader venture in which Saks permits the Brand Defendants to “sell their goods and apparel” with Saks’s stores.  Compl. ¶ 21.  Although the United States argues that the Complaint “contains no allegations of any connection . . . between the alleged conspiracy and those business relationships,” U.S. Br. 15-16, that is not correct: the Brand Defendants operate “concessions at Saks stores,” Compl. ¶ 21, and Saks employees receive brand-specific training, id. ¶ 160.  As the district court held, Op. 34 n.22, the alleged restraint prevented the Brand Defendants from hiring Saks employees who sold the Brand Defendants’ merchandise, thereby protecting Saks’s training investments, see Compl. ¶¶ 156-61, 187-91, increasing the attractiveness of the broader collaboration, and promoting mutual trust between the parties, see Rothery Storage, 792 F.2d at 224 (restraint that “serves to make the main transaction more effective in accomplishing its purpose” is “subordinate and collateral”).

Plaintiff Susan Giordano’s allegations about her own experience demonstrate that this alleged restraint is ancillary.  Giordano was a Saks employee at Saks’s Loro Piana boutique for “18 months,” during which time she became “familiar[] with Loro Piana’s . . . merchandise.”  Compl. ¶¶ 157, 160.  Giordano sought employment at a standalone Loro Piana boutique, explaining that she “would surely be an asset” because of her familiarity with Loro Piana’s product gained from Saks’s training.  Id. ¶¶ 156-61.  But the no-hire restraint allegedly prevented Loro Piana from hiring Giordano, id. ¶ 161, “ensur[ing] that [Saks] [did] not lose its personnel during the collaboration” with Brand Defendants, Aya, 9 F.4th at 1110.  Courts have found just these sorts of no-hire agreements to facilitate “procompetitive collaboration” to be “reasonably necessary.”  Id.; cf. Bogan v. Hodgkins, 166 F.3d 509, 515 (2d Cir. 1999) (rejecting per se treatment for no-hire agreements).

The United States’ arguments to the contrary are unavailing.  It argues that the alleged no-hire agreements go beyond solicitation at the concessions themselves, barring the Brand Defendants from hiring even Saks employees who independently apply or approach the Brand Defendants for a job.  U.S. Br. 19.  But the no-hire agreements’ purpose, to protect against risks that employees would leave for a collaborating brand located inside their own store, applies equally regardless of whether an employee is solicited by or independently approaches a competitor.  In both instances, Saks invested in brand-specific employee training, see Compl. ¶¶ 32, 34, 156, that the no-hire agreement protects from the unique exposure of a store-within-a-store.

The United States also suggests that the restraint is not reasonably necessary because it applies to “any brand [or designer company] carried by Saks” rather than just brands that maintain concession stands.  U.S. Br. 16, 19.  But a restraint “need not satisfy a less-restrictive-means test,” Aya, 9 F.4th at 1111; regardless, Saks employees receive detailed training on all luxury brands sold in the stores, even those that do not maintain concession stands, see Compl. ¶ 34.  The alleged no-hire agreement notably does not extend to the many luxury brands whose goods are not “carried by Saks,” id. ¶ 175, leaving Saks employees free to take their talents to those competing employers or to other retailers of luxury goods.  And the United States’ suggestion that the duration of the agreement is too long, U.S. Br. 19, ignores that employees receive continuous training to remain “knowledgeable about the particular products [sold] . . . as well as current trends,” Compl. ¶ 34 (emphasis added).  If employees could leave their employment with Saks and immediately join the competitor, then the alleged restraint would have no effect at all, and Saks would lose the incentive to invest in ongoing specialized training regarding competitor brands.

II.            The District Court Properly Decided Ancillarity On The Pleadings

Nothing in the antitrust laws prohibits a district court from resolving ancillarity on the pleadings, and the court’s decision to do so here was procedurally proper and analytically sound.  Determining whether a challenged restraint is “naked” or “ancillary” is a threshold inquiry for a Section 1 claim because “[t]his all-important classification largely determines the course of subsequent legal evaluation of [the] restraint.”  Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application, ¶ 1904 (5th ed., 2023 Cum. Supp.).  Put another way, resolving ancillarity at the outset of the case dictates the mode of analysis employed by the court: naked restraints are subject to per se treatment, while ancillary restraints are analyzed under the rule of reason.

This does not mean that ancillarity must be resolved at the pleadings—depending on the circumstances, it may be resolved after the pleadings but before summary judgment, at summary judgment, or even at trial.  See In re HIV Antitrust Litig., 2023 WL 3088218, at *23 (N.D. Cal. Feb. 17, 2023) (summary judgment); N. Jackson Pharmacy, Inc. v. Caremark RX, Inc., 385 F. Supp. 2d 740, 743 (N.D. Ill. 2005) (pre-summary judgment Rule 16 motion).  Rather, ancillarity is a threshold issue that sets the stage for the analysis that follows, and deciding it at the pleadings stage permits defendants to defeat meritless claims before undergoing costly discovery.

The district court properly resolved the question on a motion to dismiss here because Plaintiffs’ own allegations made clear that the alleged no-hire agreements were ancillary.  Plaintiffs and their amici make two arguments: first, that ancillarity cannot be resolved on the pleadings, and second, that the district court improperly resolved facts in Defendants’ favor.  Neither argument persuades.

                  A.            Ancillarity Is A Threshold Inquiry, Not An Affirmative Defense

Courts analyzing Section 1 claims must first determine the proper framework to apply: the per se rule or the rule of reason (or, in some cases, an abbreviated “quick look” analysis).  See Leegin, 551 U.S. at 886-87.  To make that determination, “[a] court must distinguish between ‘naked’ restraints, those in which the restriction on competition is unaccompanied by new production or products, and ‘ancillary’ restraints, those that are part of a larger endeavor whose success they promote.”  MLB, 542 F.3d at 339 (Sotomayor, J., concurring in the judgment) (quoting Polk Bros., 776 F.2d at 188).  “This all-important classification largely determines the course of subsequent legal evaluation of any restraint.”  Areeda & Hovenkamp, supra, ¶ 1904; see Thomas B. Nachbar, Less Restrictive Alternatives and the Ancillary Restraints Doctrine, 45 Seattle U. L. Rev. 587, 634 (2022) (“In order to do any real work, the ancillary restraints doctrine has to precede the rule of reason.”); Herbert Hovenkamp, The Rule of Reason, 70 Fla. L. Rev. 81, 140 (2018) (“The ancillary restraints doctrine is not a comprehensive method for applying the rule of reason, but rather an early stage decision about which mode of analysis should be applied.”).  Thus, ancillarity is a gating inquiry.  By determining at the outset of the case whether a challenged restraint is naked or ancillary, the court ensures it applies the proper analytical framework.

Because this determination guides how the parties conduct discovery and try the case, it is important to decide ancillarity at the earliest possible stage.  This avoids “expensive pretrial discovery” on the wrong questions and issues.  And it avoids discovery altogether in cases that do not state a claim and should never proceed past the pleadings.  Limestone Dev. Corp. v. Vill. of Lemont, 520 F.3d 797, 803 (7th Cir. 2008) (noting importance of carefully evaluating antitrust claims at pleading stage “lest a defendant be forced to conduct expensive pretrial discovery in order to demonstrate the groundlessness of the plaintiff’s claim” (citing Bell Atl. Corp. v. Twombly, 550 U.S. 544, 558-59 (2007))).

Treating ancillarity as a gating inquiry also is consistent with the Supreme Court’s admonition that per se treatment must be confined to a narrow class of cases.  As the Court has explained, “the per se rule is appropriate only after courts have had considerable experience with the type of restraint at issue and only if courts can predict with confidence that it would be invalidated” under the rule of reason.  Leegin, 551 U.S. at 886-87; Dagher, 547 U.S. at 5 (“Per se liability is reserved for only those agreements that are so plainly anticompetitive that no elaborate study of the industry is needed to establish their illegality.” (quotation marks omitted)).  That predictive confidence must be rooted in the “demonstrable economic effect” of the restraint at issue, not a plaintiff’s suspicion that the restraint is harmful.  Leegin, 551 U.S. at 887.  This is a high bar.  Only when a restraint is “so obviously lacking in any redeeming pro-competitive values” may courts apply the per se rule.  Cap. Imaging Assocs., P.C. v. Mohawk Valley Med. Assocs., Inc., 996 F.2d 537, 542 (2d Cir. 1993).

Because per se analysis is warranted only when justified by “demonstrable economic effect,” resolving the issue of ancillarity on the pleadings ensures that plaintiffs cannot invoke per se treatment on mere say-so.  The ancillarity inquiry, by definition, considers the relationship of the challenged restraint to the parties’ business collaboration—that is, the inquiry explores the likely “economic effect” of the restraint within the context of commercial realities.  That is precisely what the Supreme Court requires before expanding the per se rule into new frontiers.  Broad. Music, Inc. v. Columbia Broad. Sys., Inc., 441 U.S. 1, 19 n.33 (1979) (“[T]he per se rule is not employed until after considerable experience with the type of challenged restraint.”); Bogan, 166 F.3d at 514 (“The Supreme Court is slow to . . . extend per se analysis to restraints imposed in the context of business relationships where the economic impact of certain practices is not immediately obvious.” (quotation marks omitted)).

If ancillarity could be resolved only after the pleadings stage, as Plaintiffs and their amici urge, then a Section 1 plaintiff could survive dismissal simply by invoking the per se rule without regard for the restraint’s “economic effect” or the courts’ ability to “predict with confidence that [the restraint] would be invalidated.”  Leegin, 551 U.S. at 886-87.  A simple example underscores the absurdity of that rule: ever since they were recognized in Addyston Pipe as axiomatic ancillary restraints, no-hire provisions are commonly included in agreements for the sale of a business.  The approach proposed would require litigation through discovery to decide if such a provision were ancillary.

Moreover, neither the federal courts nor the academy have amassed sufficient experience with this subject to allow default per se treatment.  Indeed, the only study that attempted to analyze the relevant economic considerations in a systematic way concluded that eliminating no-hire provisions “causes minimal reductions in job concentration and no increase in wages.”  Daniel S. Levy et al., No-Poaching Clauses, Job Concentration and Wages: A Natural Experiment Generated by a State Attorney General, Advanced Analytical Consulting Group, Inc., at 1 (Jan. 23, 2020).  That inconclusive literature falls far short of justifying a rule that would effectively extend per se treatment to all no-hire agreements.

If anything, the economic incentives weigh strongly in favor of deciding ancillarity at the earliest possible stage allowed by the record.  This is because a rule prohibiting courts from deciding ancillarity at the pleadings stage would be a free pass to discovery (and the “potentially enormous expense” associated with it), which would “push cost-conscious defendants to settle even anemic [Section 1] cases.”  Twombly, 550 U.S. at 559.  That pressure, in turn, would distort normal business incentives—faced with the prospect of huge discovery costs from meritless claims, rational businesses would understandably refrain from entering into legitimate, procompetitive collaborations.  Plaintiffs and their amici offer no good reason for adopting a rule that would undercut the very efficiency-enhancing purposes antitrust law is meant to advance.  See Morrison v. Murray Biscuit Co., 797 F.2d 1430, 1437 (7th Cir. 1986) (“The purpose of antitrust law, at least as articulated in the modern cases, is to protect the competitive process as a means of promoting economic efficiency.”); see also MLB, 542 F.3d at 339 (Sotomayor, J., concurring in the judgment) (restraints do not receive per se treatment when they have a “reasonable procompetitive justification, related to the efficiency-enhancing purposes of [a] joint venture”).

The United States asserts that ancillarity is only a “defense” to per se illegality, rather than a threshold inquiry to determine whether a case calls for departing from the rule of reason.  U.S. Br. 12-13.  None of the United States’ cases, however, limit the ancillarity restraints doctrine in this way.  The lone Second Circuit case the United States cites was a criminal matter where the standard applied to motions to dismiss is far more lenient and deferential to the United States than that mandated for civil cases in Twombly.  In such cases, courts treat the government’s characterization of conduct as within the four corners of a recognized per se theory as sufficient for indictment purposes.  See United States v. Aiyer, 33 F.4th 97, 116 (2d Cir. 2022) (indictments need only “contain[] the elements of the offense charged” and enable defendant to enter plea).  Moreover, in that case, the defendant had not even challenged on appeal the district court’s conclusion that the indictment at issue adequately alleged a per se antitrust violation.  See id. at 116-23.  The panel never characterized ancillarity as a “defense.”  See id.

The same goes for Blackburn and Board of Regents.  Although the courts in those cases ultimately concluded the restraints at issue were not ancillary, neither case held that ancillarity was only a defense.  Blackburn v. Sweeney, 53 F.3d 825, 828-29 (7th Cir. 1995); Bd. of Regents of Univ. of Okla. v. NCAA, 707 F.2d 1147, 1153-56 (10th Cir. 1983).  Freeman is similarly off base.  While the court there offhandedly referred to the defendant’s overall argument against the antitrust claim as a “defense,” it did so after the ancillarity discussion.  Freeman v. San Diego Ass’n of Realtors, 322 F.3d 1133, 1151-52 (9th Cir. 2003).  The court did not use the term with specific reference to ancillarity, and in any event its use of “defense” was not meant in the same way that Plaintiffs and their amici use it—that is, as an issue that cannot be resolved at the outset of the case.  AOB 39; U.S. Br. 12-13, 15.  In short, none of the government’s cases hold that ancillarity is strictly a defense or is otherwise immune from resolution on the pleadings.

The Seventh Circuit’s recent decision in Deslandes doesn’t advance the government’s cause either.  Although the court in Deslandes summarily stated that “the classification of a restraint as ancillary is a defense,” 81 F.4th at 705, plaintiffs can plead themselves out of court, Hadid v. City of New York, 730 F. App’x 68, 71 (2d Cir. 2018), which is what Plaintiffs have done here.  Nor should it be followed: the Seventh Circuit cited no case law and offered no analysis to support its bald assertion.  Deslandes, 81 F.4th at 705.  And, as explained, any suggestion that ancillarity can be treated only as a defense would undo the clear demarcation between the rule of reason and per se treatment.  If courts can’t evaluate ancillarity at the outset, restraints that should be presumptively analyzed under the rule of reason would instead be presumptively treated as per se illegal.  That result is plainly inconsistent with the Supreme Court’s antitrust precedents.

In a related argument, Plaintiffs contend that ancillarity cannot be decided on the pleadings, but instead “requires discovery.”  AOB 39.  But that also is wrong.  “In considering a motion to dismiss, the court is not required to don blinders and to ignore commercial reality.”  Car Carriers, Inc. v. Ford Motor Co., 745 F.2d 1101, 1110 (7th Cir. 1984), abrogated on other grounds by Schmees v. HC1.COM, Inc., 77 F.4th 483 (7th Cir. 2023).  Consistent with this principle, courts routinely resolve ancillarity on the pleadings where it is clear from the complaint that the restraint may be procompetitive.  For example, in Helmerich & Payne International Drilling Co. v. Schlumberger Technology Corp., the court dismissed a restraint of trade claim at the pleading stage where “the pleadings in [the] case [made] clear” that the challenged non-solicitation provision was “ancillary” to “a larger business transaction between two independent parties.”  2017 WL 6597512, at *4 (N.D. Okla. Dec. 26, 2017).  Similarly, the court in Gerlinger v. Amazon.Com, Inc. determined that a purported price-fixing arrangement between Borders and Amazon was “ancillary” to the companies’ broader website hosting agreement, in part because the “context in which the agreement was entered into” confirmed its procompetitive potential.  311 F. Supp. 2d 838, 848-49 (N.D. Cal. 2004).  The court reached this conclusion on a motion for judgment on the pleadings.  Id.  Other courts have similarly decided ancillarity on the pleadings alone.  See Kelsey K. v. NFL Enters. LLC, 2017 WL 3115169, at *4 (N.D. Cal. July 21, 2017) (motion to amend), aff’d, 757 F. App’x 524, 526 (9th Cir. 2018); Hanger v. Berkley Grp., Inc., 2015 WL 3439255, at *5 (W.D. Va. May 28, 2015) (motion to dismiss); Caudill v. Lancaster Bingo Co., 2005 WL 2738930, at *3-6 (S.D. Ohio Oct. 4, 2005) (motion for judgment on the pleadings).  Contrary to Plaintiffs’ argument, the district court’s pleading-stage ancillarity ruling was entirely proper.

                   B.            The District Court Did Not Reach Past Plaintiffs’ Allegations

Ancillarity can support dismissal when it is “apparent from the allegations in the complaint,” as even the United States acknowledges.  U.S. Br. 15.  Here, the district court’s ancillarity ruling was amply supported by Plaintiffs’ own allegations.  Plaintiffs allege that Saks and the Brand Defendants collaborate in the sale of luxury goods by partnering to sell the Brand Defendants’ goods both directly at Saks stores and through concessions within them.  Compl. ¶¶ 21, 28; see supra, at 4-10.  By cooperating in this way, Saks and the Brand Defendants can leverage each other’s employees and brands to create a distinct “shopping experience for customers”—that is, the “atmosphere of exclusivity and opulence surrounding . . . luxury products,” Compl. ¶ 33, needed to promote “demand for[] luxury goods over other, lower-priced goods,” id. ¶ 23.  The upshot is a procompetitive collaboration that, in the words of Polk Bros., “promises greater productivity and output.”  776 F.2d at 189.

The district court also properly relied on the Complaint to conclude that “absent the no-hire agreement, there would be a continual risk that the Brand Defendants would use their concessions in Saks stores to recruit [Saks] employees.”  Op. 32 (citing Compl. ¶¶ 56-57, 83).  Minimizing the risk of such “free rid[ing]” is a common, efficiency-enhancing feature of ancillary restraints.  Rothery Storage, 792 F.2d at 229 (restraints were ancillary where they “preserve[d] the efficiencies of the [collaboration] by eliminating the problem of the free ride”); Polk Bros., 776 F.2d at 190 (agreement was ancillary to a joint sales venture where it limited the potential that one retailer would free ride on the sales efforts of another).  That includes procompetitive restraints on employee movement.  Aya, 9 F.4th at 1110 (restraint was ancillary to business collaboration where it guarded against risk of one party “proactively raiding . . . employees” of another party).

Notably, the risk of free riding wasn’t hypothetical: as the district court pointed out, the Complaint specifically alleges that Plaintiff Giordano sought to leverage the experience she acquired while working at the Loro Piana boutique as a Saks employee to seek employment with Loro Piana.  Op. 34 n.22.  The district court also highlighted Plaintiffs’ allegations that without the no-hire agreements, Brand Defendants such as Louis Vuitton could “take advantage” of Saks’s hiring efforts by recruiting Saks employees away from Saks after that company had already invested time and money to recruit and train its personnel.  Op. 32; Compl. ¶¶ 62-63; see Compl. ¶ 53 (alleging that “a Defendant would save on training costs and receive the immediate benefit of a well-trained, motivated salesperson” by hiring “from one of its rivals”).  This poaching, according to Plaintiffs, would “inflict[] a cost on [Saks] by removing an employee on whom [Saks] may depend.”  Compl. ¶ 62.  Thus, Plaintiffs’ own allegations demonstrate the alleged no-hire agreement is ancillary.  By addressing the free-rider problem, the agreement eliminates an externality “that may otherwise distort the incentives of [the Brand Defendants] and limit the potential efficiency gains of [the collaboration].”  MLB, 542 F.3d at 340 (Sotomayor, J., concurring in the judgment).  Nothing more was required to resolve ancillarity on the pleadings.

Plaintiffs and their amici argue the district court erred by drawing factual inferences in favor of Saks, rather than Plaintiffs.  AOB 37-40; N.Y. Br. 26-27.  According to Plaintiffs, ancillarity was a “contested factual issue” that could be resolved in Saks’s favor only by improperly rejecting Plaintiffs’ allegations.  AOB 37-38.  Plaintiffs’ argument is misplaced.

“Determining whether a complaint states a plausible claim for relief [is] a context-specific task that requires the reviewing court to draw on its judicial experience and common sense.”  Jessani v. Monini N. Am., Inc., 744 F. App’x 18, 19 (2d Cir. 2018) (quoting Ashcroft v. Iqbal, 556 U.S. 662, 679 (2009)).  As part of that exercise, courts consider “a host of considerations: the full factual picture presented by the complaint, the particular cause of action and its elements, and the existence of alternative explanations so obvious that they render plaintiff’s inferences unreasonable.”  Fink v. Time Warner Cable, 714 F.3d 739, 741 (2d Cir. 2013); see Boca Raton Firefighters & Police Pension Fund v. Bahash, 506 F. App’x 32, 35 (2d Cir. 2012).

That is precisely what the district court did here.  It considered the “full factual picture presented by the complaint”—including the nature of the Defendants’ business relationship and the role of the no-hire agreement in the context of that relationship—to conclude that the alleged no-hire agreement was ancillary to a procompetitive collaboration.  Fink, 714 F.3d at 741 (emphasis added); Op. 28-34.  And in doing so, the court properly demonstrated that Plaintiffs’ own allegations precluded per se treatment.  See Weisbuch v. Cnty. of Los Angeles, 119 F.3d 778, 783 n.1 (9th Cir. 1997) (“Whether [a] case can be dismissed on the pleadings depends on what the pleadings say.”).  Plaintiffs can’t avoid the consequences of their allegations by truncating the court’s properly holistic review of the pleadings—indeed, “[i]f the pleadings establish facts compelling a decision one way, that is as good as if depositions and other expensively obtained evidence on summary judgment establishes the identical facts.”  Id.

The district court’s ancillarity ruling was sound.


For the foregoing reasons, this Court should affirm.

[1] All parties have consented to the filing of this brief.  Pursuant to Federal Rule of Appellate Procedure 29(a)(4)(E), counsel for ICLE represents that no counsel for any of the parties authored any portion of this brief and that no entity, other than amici curiae or their counsel, monetarily contributed to the preparation or submission of this brief.

[2] The alleged no-hire agreements also do not fit the per se mold because they are part of a dual-distribution relation in which the Brand Defendants sell their products to end consumers through “their own standalone boutiques” as well as through distributors, “including Saks.”  Compl. ¶ 21; see Beyer Farms, Inc. v. Elmhurst Dairy, Inc., 35 F. App’x 29, 29-30 (2d Cir. 2002) (holding that a restraint was “subject to scrutiny under the ‘rule of reason’” because the complaint alleged a “dual-distributorship relationship”); Elecs. Commc’ns Corp. v. Toshiba Am. Consumer Prods., Inc., 129 F.3d 240, 243 (2d Cir. 1997) (similar).

Continue reading
Antitrust & Consumer Protection

ICLE Amicus Brief in Illumina & Grail v FTC

Amicus Brief IDENTITY AND INTEREST OF AMICUS CURIAE AND SOURCE OF AUTHORITY TO FILE BRIEF The International Center for Law & Economics (“ICLE”) is a nonprofit, nonpartisan, . . .


The International Center for Law & Economics (“ICLE”) is a nonprofit, nonpartisan, global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law and economics methodologies, and economic findings, to inform public policy, and has longstanding expertise in antitrust law.

Amici also include 28 scholars of antitrust, law, and economics at leading universities and research institutions across the United States. Their names, titles, and academic affiliations are listed in Appendix A. All amici have extensive expertise in antitrust law and economics, and several served in senior positions at the Federal Trade Commission or the Antitrust Division of the Department of Justice.

Amici have an interest in ensuring that courts and agencies correctly apply the standards for evaluating horizontal and vertical mergers, and take into account the benefits commonly associated with vertical mergers.        

Amici are authorized to file this brief by Fed. R. App. P. 29(a)(2) because all parties have consented to its filing.

RULE 29(a)(4)(e) STATEMENT

Amici hereby state that no party’s counsel authored this brief in whole or in part; that no party or party’s counsel contributed money that was intended to fund the preparation or submission of the brief; and that no person other than amicus or its counsel contributed money that was intended to fund the preparation or submission of the brief.


The FTC’s decision to require Illumina to divest Grail rests on at least two misguided premises. The first is that the same scrutiny applies to both horizontal and vertical mergers. The second is that benefits typically associated with vertical mergers do not apply here.

A horizontal merger combines firms that compete in the same relevant market, which necessarily reduces the number of firms engaged in head-to-head competition and may eliminate substitutes. That reduction inherently tends to increase prices, but the price effect may be trivial.  In addition, market responses (competitive repositioning or new entry) or other benefits of the merger (savings in transaction and other costs, enhanced investment incentives) may neutralize or offset the impetus to higher prices. But because those benefits are not automatic (and the reduction of direct competition is), they must be proven rather than assumed if the merger otherwise poses a significant risk of anticompetitive effects.

A vertical merger, in contrast, combines firms with an upstream-downstream (e.g., seller-buyer) relationship—that is, “firms or assets at different stages of the same supply chain.” Dep’t of Justice, Antitrust Division and FTC, Vertical Merger Guidelines 1 (2020). Examples include a manufacturer’s acquiring a distributor or a firm providing a manufacturing input.

The economic consequences of combining complements rather than substitutes are fundamentally different. Whereas the first-order effect of a horizontal merger is upward pricing pressure, the first-order effect of a vertical merger is downward pricing pressure. Vertical mergers typically entail the elimination of double marginalization (“EDM”), which is akin to downward pricing pressure (and often considered alongside efficiencies). David Reiffen & Michael Vita, Comment: Is There New Thinking on Vertical Mergers? 63 Antitrust L.J. 917, 920 (1995). Vertical integration also typically internalizes externalities in research and development, resulting in greater investment. Henry Ogden Armour & David J. Teece, Vertical Integration and Technological Innovation, 62 Rev. Econ. & Stat. 470 (1980). Like horizontal mergers, vertical mergers often confer other benefits such as operational and transactional efficiencies. Dennis W. Carlton, Transaction Costs and Competition Policy, 73 Int’l J. Indus. Org. 1 (2019); Oliver Williamson, The Economic Institutions of Capitalism 86 (1985).

Thus, while both types of mergers can create benefits from cost savings, their intrinsic effects move in opposite directions: higher prices and less investment with horizontal mergers, and lower prices and more investment with vertical mergers.

  1. The FTC’s conclusion that the same scrutiny applies to horizontal and vertical mergers, Opinion 75, conflicts with precedent (and long-standing economic research). Courts and economists alike recognize that vertical integration typically is procompetitive, and it is widely accepted that vertical mergers and horizontal mergers should be evaluated under different presumptions. As the leading antitrust treatise puts it, “[i]n the great majority of cases no anticompetitive consequences can be attached to [vertical integration], and injury to competition should never be inferred from the mere fact of vertical integration.” 3B Phillip Areeda & Herbert Hovenkamp, Antitrust Law ¶?755a (4th ed. 2017). That vertical mergers can be anticompetitive—under specific facts and circumstances—does not establish that vertical integration is likely to be anticompetitive (it is not) or that there is no useful antitrust distinction between vertical and horizontal mergers (there is).

The Commission did not simply presume that this vertical merger would be anticompetitive, however. It also discounted both the likelihood of efficiencies in vertical mergers and specific evidence of efficiencies associated with the already-consummated merger. As a result, the Commission did not properly assess the likely competitive effects of the merger.

  1. The Commission also disregarded evidence of a current and operative constraint on any potential anticompetitive effects of the merger. Illumina’s Open Offer appears to be contractually binding, and addresses the risk of foreclosure that is the primary competitive concern here. Proper consideration of the Open Offer should have shifted the Commission further away from presuming harm. Instead, the Commission gave it no weight.
  2. The existing standards for vertical merger scrutiny are informed by, and consistent with, economic research regarding vertical mergers and other forms of vertical integration. That research shows that the Commission was wrong to hold that vertical and horizontal mergers should be analyzed identically, and wrong to disregard the well-established benefits of vertical integration. While economic theory indicates that vertical mergers can be anticompetitive, the weight of the empirical evidence overwhelmingly indicates that they tend to be procompetitive or competitively neutral. Indeed, the large majority of vertical mergers that have been studied have been found to be procompetitive or benign. That suggests that case-specific evidence is paramount in assessing both potential anticompetitive effects and countervailing pro-consumer efficiencies.


I.    Vertical and Horizontal Mergers Should Be Scrutinized Differently.

A.  Prima Facie Standards and the Government’s “Ultimate Burden” Differ in Horizontal and Vertical Merger Cases.

Courts have long recognized that horizontal and vertical mergers are categorically different. “As horizontal agreements are generally more suspect than vertical agreements,” courts are “cautious about importing relaxed standards of proof into vertical agreement cases.” Republic Tobacco v. North Atlantic Trading Co., 381 F. 3d 717, 737 (7th Cir. 2004). Thus, for vertical mergers, “unlike horizontal mergers, the government cannot use a shortcut to establish a presumption of anticompetitive effect.…” United States v. AT&T, Inc., 916 F.3d 1029, 1032 (D.C. Cir. 2019) (“AT&T II”). In a vertical merger case, “the government must make a fact-specific showing that the proposed merger is likely to be anticompetitive,” and the “ultimate burden of persuasion… remains with the government at all times.” Id. (emphasis added; cleaned up).

As the ALJ’s Initial Decision (ID) recognized (at 132), the burden-shifting approach is not bound by any specific, sequential form. Then-Judge Thomas stressed in United States v. Baker Hughes, 908 F.2d 981, 984 (D.C. Cir. 1990), that “[t]he Supreme Court has adopted a totality-of-the-circumstances approach…, weighing a variety of factors to determine the effects of particular transactions on competition.” As the ALJ aptly put it, the Baker Hughes “‘burden-shifting language’” provides “‘a flexible framework rather than an air-tight rule’”; “in practice, evidence is often considered all at once and the burdens are often analyzed together.” ID 132 (quoting Chicago Bridge & Iron Co. v. FTC, 534 F.3d 410, 424-24 (5th Cir. 2008)).

The differential treatment of vertical and horizontal mergers parallels the Supreme Court’s vertical restraints jurisprudence. The potential anticompetitive effects of vertical restraints are similar to those posed by vertical mergers, as both obtain between firms at different levels of the supply chain.

Over time, the Supreme Court has eliminated per se condemnation for vertical restraints. In 1977, the Court rejected per se illegality for vertical non-price restraints, Cont’l T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 49, 52 n.19, 58 (1977) (overruling United States v. Arnold, Schwinn & Co., 388 U.S. 365 (1967)), later confirming that “a vertical restraint is not illegal per se unless it includes some agreement on price or price levels.” Bus. Elecs. Corp. v. Sharp Elecs. Corp., 485 U.S. 717, 735-36 (1988). Eventually, the Court repudiated the last vertical per se prohibition—of vertical minimum price restraints. Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877 (2007) (overruling Dr. Miles Med. Co. v. John D. Park & Sons Co., 220 U.S. 373 (1911)). In these decisions, the Court emphasized that any departure from the evidence-specific rule of reason “must be based on demonstrable economic effect, rather than . . . upon formalistic line drawing.” Bus. Elecs., 485 U.S. at 724.

These decisions reflect a nearly categorical repudiation of presumptions of illegality in dealings involving entities at different levels of the supply chain. Here, however, the Commission took the opposite approach, presuming anticompetitive effect while rejecting the significance of rigorously established benefits in a way that approaches a per se standard. This Court should reject that departure from sound law and economics.

B.  The FTC Did Not Undertake the Necessary Fact-Specific Examination of the Merged Firm’s Incentives Given the Merger’s Efficiencies.

The FTC had to show that Illumina has a greater incentive to foreclose rivals following—and because of—the merger. Instead, the Commission adopted a standard of review that elides the requirement that, in a vertical merger case where there is “no presumption of harm in play,” “the government must make a fact-specific showing that the proposed merger is likely to be anticompetitive” both at the prima facie stage and in the final analysis. United States v. AT&T Inc., 310 F. Supp. 3d 161, 192 (D.D.C. 2018) (“AT&T I”), aff’d, 916 F.3d 1029 (D.C. Cir. 2019); see AT&T II, 916 F.3d at 1032.

To be sure, there is little recent case law regarding the standard of review for vertical mergers because the federal antitrust agencies have rarely challenged, let alone litigated, vertical acquisitions. The Department of Justice challenge to the AT&T/Time Warner merger marked “the first time in 40 years that a court has heard a fully-litigated challenge to a vertical merger.” Joshua D. Wright & Jan M. Rybnicek, US v. AT&T Time Warner: A Triumph of Economic Analysis, 6 J. Antitrust Enforcement 3 (2018).

Nevertheless, up to now, the agencies have considered likely structural benefits, transactional efficiencies, and potential remedies, along with potential harms, in toto and on net, in assessing a merger’s likely competitive impact. Hence, the Vertical Merger Guidelines—jointly adopted by the FTC and the Antitrust Division of the Department of Justice in June 2020 (although withdrawn by the FTC while this case was pending)—state that “[t]he Agencies do not challenge a merger if cognizable efficiencies are of a character and magnitude such that the merger is unlikely to be anticompetitive in any relevant market.” Vertical Merger Guidelines at 11. Even under the Horizontal Merger Guidelines, “[t]he Agency will not challenge a merger if cognizable efficiencies are of a character and magnitude such that the merger is not likely to be anticompetitive.” Dep’t of Justice, Antitrust Division & FTC, Horizontal Merger Guidelines § 10 (Aug. 19, 2010).

But here the Commission did not seriously account for likely efficiencies or other benefits that may be derived from practices inconsistent with foreclosure. Put simply, Illumina’s ability to profit from the merger without foreclosing rivals reduces its incentive to foreclose. Although the foreclosure incentive may remain on some margins, the question of the greater incentive cannot be resolved without assessing the incentives against foreclosure as well as those for it.

Illumina’s post-merger incentive to foreclose rivals may be constrained by:

  • its interest in revenue realized from a broader array of sequencing clients than the relatively few engaged in multi-cancer early detection (MCED) research;
  • the procompetitive—and pro-consumer—cost advantages it is likely to realize from integration with Grail;
  • the relatively low risk of entry by close substitutes for Galleri in the near, or even foreseeable, future;
  • the Open Offer;
  • reputational or transactional harms that may result from refusing to deal with firms in its industry; and
  • the litigation and regulatory risks attending attempted foreclosure.

But the FTC presumed away these and other factors that could mitigate the risk of harm.

1.   Presumptions Suitable to Horizontal Mergers Are Not Fit for the Analysis of Vertical Mergers

The Commission maintains that the same scrutiny applies to efficiencies claims in vertical and horizontal transactions. To justify this conclusion, the Commission declined to “simply take managers’ word for efficiencies without independent verification, because then the efficiency defense ‘might well swallow the whole of Section 7,’ as managers could present large unsubstantiated efficiencies claims and courts would be hard pressed to find otherwise.” Opinion 75-76 (quoting United States v. H&R Block, Inc., 833 F. Supp. 2d 36, 91 (D.D.C. 2011). But this is a non sequitur, and wrong for three reasons.

First, the Commission need not (and the ID did not) “simply take managers’ word for efficiencies.” As the Initial Decision noted, courts and academic authorities both recognize procompetitive effects, including efficiencies, generally observed with vertical integration. ID 133-35, 196. See also ID 135 (noting case-specific evidence regarding research and development efficiencies, EDM, and the acceleration of access).

Second, to support its rejection of competitive benefits, the FTC continued to conflate the legal standards for horizontal and vertical mergers, relying only on serial string citations to horizontal merger cases. Opinion 75-76. Because a vertical merger puts direct, downward pressure on prices and upward pressure on complementary investments—the inverse of horizontal merger effects—the reliance on horizontal cases highlights the Commission’s failure to recognize the fundamental difference between horizontal and vertical integration. See AT&T II, 916 F.3d at 1032.

Third, ignoring that distinction, and the resulting need for a “fact-specific showing” of the likely anticompetitive effects of a vertical merger, id. at 1032, the Commission repeatedly relied upon H&R Block, which says nothing about standards of review for vertical mergers. H&R Block involved a horizontal merger that allegedly would have produced “an effective duopoly.” 833 F. Supp. 2d at 44.

The FTC’s Opinion also cites a horizontal merger decision, FTC v. H.J. Heinz Co., 246 F.3d 708, 713 (D.C. Cir. 2001), and AT&T II—a vertical merger case—for the proposition that the Baker Hughes framework applies to both horizontal and vertical mergers. That is misleading. Again, AT&T emphasizes that the distinction between horizontal and vertical mergers precludes similar presumptions of anticompetitive effects, and makes it easier to establish certain recognized efficiency and other benefits of vertical integration.  See AT&T II, 916 F.3d at 1032; Vertical Merger Guidelines at 5. Not incidentally, the Government lost its merger challenge in AT&T, both at trial and on appeal.

2.   The Commission Failed to Give Due Consideration to Evident Benefits

The Commission also discounted—or ignored—various efficiencies and other benefits on the ground that “efficiencies are ‘inherently difficult to verify and quantify.’” Opinion 75 (citing H&R Block, 833 F. Supp. 2d at 89). To justify this approach, the Commission cites five horizontal merger matters: H&R Block; H.J. Heinz; Otto Bock HealthCare N. Am., Inc., 168 F.T.C. 324 (2019); FTC v. Wilh. Wilhelmsen Holding ASA, 341 F. Supp. 3d 27 (D.D.C. 2018); and FTC v Penn State Hershey Medical Center, 838 F.3d 327 (3d Cir. 2016).

Although some claimed efficiencies from horizontal mergers can be hard to verify, many efficiencies from vertical mergers are inherent. Specifically, if upstream and downstream margins are positive, basic economic theory predicts that the merger will mitigate double marginalization. Empirical research confirms this. See, e.g., Gregory S. Crawford, et al., The Welfare Effects of Vertical Integration in Multi­channel Television Markets, 86 Econometrica 891 (2018). Similarly, when vertically related firms make complementary investments, theory predicts—and empirical research confirms—that vertical mergers will internalize investment spillovers in a way that tends to expand investment. See, e.g., Chenyu Yang, Vertical Structure and Innovation: A Study of the SoC and Smartphone Industries, 51 Rand J. Econ. 739 (2020). Meanwhile, operational and transactional efficiencies can be supported by both theoretical and empirical evidence, as well as case-specific evidence about the merging firms.

Here, the ALJ’s findings of fact detail ongoing innovation by Illumina, including improvements to its next generation sequencing (NGS) technologies ranging from the release of new reagents to software updates expected to result from the merger. ID 88-89. The Initial Decision also describes a complex process of integration between Illumina’s NGS technology and the requirements of different MCED testing programs. ID 89-91.

Given the Commission’s disregard of efficiencies, it is unclear when or how procompetitive benefits could ever offset the harm alleged to result if the consummated merger were left undisturbed—harm that the Commission did not quantify in either magnitude or likeli­hood.

3.   The Commission’s Speculative Prima Facie Case Fails to Account for the Likely Risk of Actual Harm

The efficiencies and competitive benefits here seem substantially easier to verify and quantify than the magnitude or likelihood of the supposed harm that the Commission neither quantified nor estimated. The Commission did not seriously try to quantify the effects of the merger on the timing and competitive significance of entry of complex clinical products, such as MCED tests, in early stages of development. Rather, the Commission simply asserted that “likely substantial harms to current, ongoing innovation competition in nascent markets are sufficiently probable and imminent to violate Section 7” of the Clayton Act, Opinion 60-61 (cleaned up). But the Commission identified no evidence to support this assertion, or to refute the ALJ’s determinations that MCED tests in development were not poised to enter into competition with Grail’s Galleri test, ID 143-144, that most of the research on possible MCED tests was relatively preliminary, ID 144-145, and that most of the tests being investigated appeared to be far from close substitutes for Galleri. ID 145-153; see also ID 27-28, 44-61.

Instead, the Commission disputed the legal relevance of those findings, stating that its analysis “rests on harm to current, ongoing R&D efforts, rather than the precise timing or nature of any firm’s commercialization of an MCED test.” Opinion 56 n. 38. But that harm, too, is assumed rather than observed, and is neither verified nor quantified.

Thus, the Commission’s prima facie case rests both on a peremptory dismissal of competitive benefits and efficiencies and an uncritical acceptance of speculative theories of harm. Pre-merger, Illumina maintained a substantial ownership interest in Grail of no less than 12%, ID 7-11, yet the Commission did not identify any attempts by Illumina or Grail to interfere with research and development of any MCED test that might enter to compete with Galleri. The only head-to-head R&D competition noted was between Grail and one firm with a pipeline MCED test (Exact/Thrive), on two dimensions: first, various “prelaunch” activities, such as “competing for mindshare with physicians, with health systems, with payers,” ID 34; second, competition for research scientists capable of contributing to the development of MCED tests, id. But there was neither allegation nor evidence that Illumina or Grail engaged in anticompetitive conduct in these areas, and no obvious way in which Illumina could exploit whatever market power it enjoys in NGS markets to foreclose access to “mindshare” or research scientists.

Given no past, present, or ongoing harm to third-party R&D efforts, there is no basis to ignore the likelihood of entry into the MCED test product market, the likely timing of entry, or the likely competitive significance of entry by particular MCED tests that might be relatively close or poor substitutes for Galleri.

Each of those factors is directly relevant to the present risk of potential harm to future competition. They determine whatever risk ongoing R&D into MCED tests would pose to Grail, and hence affect the merged firm’s foreclosure incentives. Equally relevant is the risk to Illumina’s core income stream from NGS sales and services should it prove unreliable or capricious in fulfilling its contracts. That core business includes diverse clinical testing well beyond the potential rivals at issue, ID 92-93, with clients including “leading genomic research centers, academic institutions, government laboratories and hospitals, as well as pharmaceutical, biotechnology, commercial molecular diagnostic laboratories, and consumer genomics companies.” ID 6.

4.   Evidence of Likely Procompetitive Effects Should Not Be Ignored at Any Stage of Analysis

Finally, the Commission contends that “[c]ourts have never held that efficiencies alone immunized an otherwise unlawful transaction.” Opinion 75. That puts the cart before the horse, as benefits from aligning incentives between producers of complements (what the Commission terms “efficiencies”) often determine whether a transaction—especially a vertical transaction—is “unlawful” in the first place.

Most important, the courts have never held that these benefits are irrelevant generally (as the FTC would have it), or to the question whether a transaction is unlawful in the first instance. To the contrary, analysis of a vertical merger must account for the procompetitive benefits and efficiencies it is likely to achieve. See AT&T I, 310 F. Supp. at 198 (noting need “to ‘balance’ whether the Government’s asserted harms outweigh the merger’s conceded consumer benefits.”). Even in horizontal mergers, sufficiently large efficiency benefits may prevent a merger from being illegal. New York v. Deutsche Telekom AG, 439 F. Supp. 3d 179, 207 (S.D.N.Y. 2020).

Because AT&T II made clear that no presumption of illegality applies to vertical mergers, the Commission properly faces a rigorous burden to prove on case-specific evidence that the proposed merger is likely to cause substantial, actual harm to competition and consumers—not a possibility of some degree of harm to competition that in theory could harm consumers.  The Commission has not carried that burden.

II.          The Open Offer Undercuts the Commission’s Prima Face Case and Its Disregard of Potential Remedies.

The Commission’s legal error went beyond its application of a misplaced presumption of illegality that is impervious to evidence of the benefits from combining complements. The Commission also failed to recognize key structural differences between horizontal and vertical mergers.

The primary source of potential anticompetitive harm from vertical integration is foreclosure. While foreclosure is not consistently defined, one passable definition is:

[A] dominant firm’s denial of proper access to an essential good it produces, with the intent of extending monopoly power from that segment of the market (the bottleneck segment) to an adjacent segment (the potentially competitive segment).

Patrick Rey & Jean Tirole, A Primer on Foreclosure, in 3 Handbook of Industrial Organization 2145, 2148 (Mark Armstrong & Robert H. Porter, eds.) (2007). Because denial of access is a crucial aspect of foreclosure, agreements (or remedies) granting access to essential goods or services can mitigate the risk of foreclosure.

Illumina’s “Open Offer” appears to grant such access, yet the Commission failed to give proper weight to its effect on the risk of anticompetitive conduct. In contrast, the ALJ examined the Open Offer in detail, see, e.g., ID 98-125, 182-189, finding that it “provides a compre­hensive set of protections for Illumina’s customers for all aspects of conduct and competition.” ID 120. The Commission rejected those findings, relying in part on a mischaracterization of the Open Offer as only a proposed remedy, and in part on an overbroad repudiation of behavioral remedies.

First, the record indicates that the Open Offer is binding under New York law, at least with respect to several firms engaged in MCED research, and that it will remain so through August 2033. ID 103-04. Firms that have accepted (or will accept) the Open Offer can enforce it whether or not the merger is blocked; and they would have every incentive to do so if Illumina interfered with their R&D efforts. That is not just a proposed remedy, but a fully operative constraint.  If accepted, the Open Offer will become part of the institutional framework within which Illumina operates, further reducing or eliminating the firm’s incentives and ability to raise its rivals’ costs. ID 103-04, 179. Cf. United States v. General Dynamics Corp., 415 U.S. 486, 501-02 (1974) (noting importance of existing contracts in assessing competitive landscape).

That constraint seems especially significant given how few firms might someday enter to compete with Grail’s MCED test, and the difficulty inherent in trying to forecast R&D competition so far in advance.

Second, the Commission strains credulity in disregarding the Open Offer on the grounds that behavioral remedies can be hard to monitor and tend to be disfavored. If the Open Offer were incorporated into a consent order, the FTC would have to monitor only a very few agreements. The affected parties would assist in monitoring compliance, well-funded would-be entrants would have every incentive to report any difficulty gaining access to Illumina’s sequencing technology, and the FTC could modify the order as needed. Illumina, for its part, would face both the risk of damages imposed under state law and the risk of statutory penalties, among other remedies, for violations of FTC consent orders.

Under the flexible Baker Hughes approach, the Commission should have accorded substantial weight to the Open Offer in assessing whether the Illumina-Grail transaction is truly likely to cause harm. This behavioral remedy is neither cumbersome nor ineffective. Given the Open Offer, the Commission does not appear to have established that harm to R&D competition is likely or imminent.

III.       The Economics of Vertical Integration Support the Differential Treatment of Vertical and Horizontal Mergers.

Economic and empirical research confirm that the Commission was wrong to conclude that vertical and horizontal mergers should be analyzed identically. Horizontal mergers, by definition, remove a competitor from a relevant market; vertical mergers do not. As the economics literature makes clear, that structural distinction is central to antitrust analysis.

A.           In Theory, The Competitive Implications of Vertical Mergers Are Ambiguous.

The Supreme Court’s modern vertical restraints decisions underscore the importance of developments in the economic literature for assessing how to evaluate any type of integration under the antitrust laws. The Court removed per se prohibitions on vertical restraints in part because “economics literature is replete with procompetitive justifi­ca­tions for” them. Leegin, 551 U.S. at 889.

The economics literature is equally “replete with procompetitive justifications” for vertical integration. Vertical integration typically confers benefits, such as eliminating double marginalization, Reiffen & Vita, supra, 63 Antitrust L.J. 917; increasing R&D investment, Armour & Teece, supra, 62 Rev. Econ. & Stat. 470; and creating operational and transactional efficiencies, Carlton, supra, 73 Int’l J. Indus. Org. 1.

The logic behind EDM is simple: Vertical mergers can increase welfare, even if the upstream or downstream firm has market power. When firms “markup” their products over their marginal cost of production, that reduces output and increases the (input or distribution) costs of their (downstream or upstream) rivals. In other words, independent upstream and downstream firms can exert negative externalities on each other that ultimately push prices upwards. When firms have no incentive to consider the effect of their price (and output) determinations on downstream firms’ profits, see, e.g., Michael A. Salinger, Vertical Mergers and Market Foreclosure, 103 Q.J. Econ. 345 (1988), there is an additional markup over the downstream firm’s marginal cost of production, or “double marginalization.” Vertical mergers enable firms to coordinate their pricing behavior, eliminating this externality without the negative effects that coordination would entail in horizontal merger cases. See Reiffen & Vita, supra, 63 Antitrust L. J. at 920.

In a vertical merger, EDM is likely automatic. Id. That is “precisely opposite of the outcome that arises under the frequently used Cournot oligopoly model of horizontal competition with substitute products. Under Cournot oligopoly, joint pricing raises price; under Cournot complements [as in a vertical merger], it lowers price.” Daniel O’Brien, The Antitrust Treatment of Vertical Restraint: Beyond the Possibility Theorems, in Konkurrensverket, Swedish Competition Authority, Report: The Pros and Cons of Vertical Restraints 22, 36 (2008).[1]

To be clear, vertical mergers are not necessarily procompetitive. An integrated firm may have an incentive to exclude rivals, see Steven C. Salop & David T. Scheffman, Cost-Raising Strategies, 36 J. Indus. Econ. 19 (1985), and a vertical merger can have an anticompetitive effect if the upstream firm has market power and the ability, post-acquisition, to foreclose its competitors’ access to a key input. See Janusz A. Ordover, Garth Saloner & Steven C. Salop, Equilibrium Vertical Foreclosure, 80 Am. Econ. Rev. 127 (1990). In that regard, raising rivals’ costs can “represent[] a credible theory of economic harm” if other conditions of exclusionary conduct are met. Malcom B. Coate & Andrew N. Kleit, Exclusion, Collusion, and Confusion: The Limits of Raising Rivals’ Costs, FTC Bureau of Economics Working Paper No. 179 (1990). But this is merely a possibility, not a likely conclusion without solid empirical evidence: “The circumstances… in which [raising rivals’ costs] can occur are usually so limited that [it] almost always represents a minimal threat to competition.” Id. at 3.

The implications of vertical mergers are thus theoretically ambiguous, not typically anticompetitive. But while the Commission now seeks to equate horizontal and vertical mergers,

[a] major difficulty in relying principally on theory to guide vertical enforcement policy is that the conditions necessary for vertical restraints to harm welfare generally are the same conditions under which the practices increase consumer welfare.

James C. Cooper, et al., Vertical Antitrust Policy as a Problem of Inference, 23 Int’l. J. Indus. Org. 639, 643 (2005).

This structural ambiguity weighs against any presumption against vertical mergers, and suggests the importance of empirical research in formulating standards to evaluate vertical transactions.

B.           Empirical Research Establishes that Vertical Mergers Tend to Be Procompetitive In Practice.

Empirical evidence supports the established legal distinctions between horizontal mergers and vertical mergers (as well as other forms of vertical integration), indicating that vertical integration tends to be procompetitive or benign.

A meta-analysis of more than seventy studies of vertical transactions analyzed groups of studies for their implications for various theories or models of vertical integration, and for the effects of vertical integration. From that analysis

a fairly clear empirical picture emerges. The data appear to be telling us that efficiency considerations overwhelm anticompetitive motives in most contexts. Furthermore, even when we limit attention to natural monopolies or tight oligopolies, the evidence of anticompetitive harm is not strong.

Francine Lafontaine & Margaret Slade, Vertical Integration and Firm Boundaries: The Evidence, 45 J. Econ. Lit. 629, 677 (2007).

On the contrary, “under most circumstances, profit-maximizing vertical integration decisions are efficient, not just from the firms’ but also from the consumers’ points of view.” Id. And “[a]lthough there are isolated studies that contradict this claim, the vast majority support it….” Id. Lafontaine and Slade accordingly concluded that “faced with a vertical arrangement, the burden of evidence should be placed on competition authorities to demonstrate that that arrangement is harmful before the practice is attacked.” Id.

Another study of vertical restraints finds that, “[e]mpirically, vertical restraints appear to reduce price and/or increase output. Thus, absent a good natural experiment to evaluate a particular restraint’s effect, an optimal policy places a heavy burden on plaintiffs to show that a restraint is anticompetitive.” Cooper, et al., supra, 23 J. Indus. Org. at 639.

Subsequent research has reinforced these findings. Reviewing the more recent literature from 2009-18, John Yun concluded “the weight of the empirical evidence continues to support the proposition that vertical mergers are less likely to generate competitive concerns than horizontal ones.” John M. Yun, Vertical Mergers and Integration in Digital Markets, in The GAI Report on the Digital Economy (Joshua D. Wright & Douglas H. Ginsburg, eds., 2020) at 245.

Leading contributors to the empirical literature, reviewing both new studies and critiques of the established view of vertical mergers, maintain a consistent view. For example, testifying at a 2018 FTC hearing, Francine Lafontaine, a former Director of the FTC’s Bureau of Economics, acknowledged that some of the early empirical evidence is less than ideal, in terms of data and methods, but reinforced the overall conclusions of her earlier research “that the empirical literature reveals consistent evidence of efficiencies associated with the use of vertical restraints (when chosen by market participants) and, similarly, with vertical integration decisions.” Francine Lafontaine, Vertical Mergers (Presentation Slides), in FTC, Competition and Consumer Protection in the 21st Century; FTC Hearing #5: Vertical Merger Analysis and the Role of the Consumer Welfare Standard in U.S. Antitrust Law, Presentation Slides 93 (Nov. 1, 2018) (“FTC Hearing #5”), available at See also Francine Lafontaine & Margaret E. Slade, Presumptions in Vertical Mergers: The Role of Evidence, 59 Rev. Indus. Org. 255 (2021).

In short, empirical research confirms that the law properly does not presume that vertical mergers have anticompetitive effects, but requires specific evidence of both harms and efficiencies.

C.           New Research Does Not Undermine the Prevailing View of Vertical Mergers.

Critics of prevailing legal standards and agency practice have pointed to a few studies that might cast doubt on the ubiquity of benefits associated with vertical mergers. We briefly review several of those studies, including those discussed at the FTC’s 2018 “Competition and Consumer Protection in the 21st Century” hearings that purported to suggest that the “econometric evidence does not support a stronger procompetitive presumption [for vertical mergers].” Steven C. Salop, Revising the Vertical Merger Guidelines (Presentation Slides), in FTC Hearing #5, supra, Presentation Slides 25. In fact, these studies do not undermine the longstanding economic literature. See Geoffrey A. Manne, Kristian Stout & Eric Fruits, The Fatal Economic Flaws of the Contemporary Campaign Against Vertical Integration, 69 Kansas L. Rev. 923 (2020). “[T]he newer literature is no different than the old in finding widely procompetitive results overall, intermixed with relatively few seemingly harmful results.” Id. at 951.

One oft-cited study examined Coca-Cola and PepsiCo’s acquisitions of some of their downstream bottlers. Fernando Luco & Guillermo Marshall, The Competitive Impact of Vertical Integration by Multiproduct Firms, 110 Am. Econ. Rev. 2041 (2020). The authors presented their results as finding that “vertical integration in the US carbonated-beverage industry caused anticompetitive price increases in products for which double margins were not eliminated.” Id. at 2062. But the authors actually found that, while such acquisitions were associated with price increases for independent Dr Pepper Snapple Group products, they were associated with price decreases for both Coca-Cola and PepsiCo products bottled by vertically integrated bottlers. Because the products associated with increased prices accounted for such a small market share, “vertical integration did not have a significant effect on the price index when considering the full set of products.” Id. at 2056. Overall, the consumer impact was either an efficiency gain or no significant change. As Francine Lafontaine characterized the study, “in total, consumers were better off given who was consuming how much of what.” FTC Hearing #5, supra, Transcript 88 (statement of Francine Lafontaine), available at

In another study often cited by skeptics of vertical integration, Justine Hastings and Richard Gilbert examined wholesale price changes charged by a vertically integrated refiner/retailer using data from 1996-98. Justine S. Hastings & Richard J. Gilbert, Market Power, Vertical Integration, and the Wholesale Price of Gasoline, 33 J. Indus. Econ. 469 (2005). They observed that the firm charged higher wholesale prices in cities where its retail outlets competed more with independent gas stations, and concluded that their observations were consistent with the theory of raising rivals’ costs. Id. at 471.

In subsequent research, however, three FTC economists publishing in the American Economic Review examined retail gasoline prices following the 1997 acquisition of an independent gasoline retailer by a vertically integrated refiner/retailer. Their estimates suggested that the merger was associated with minuscule—and economically insignificant—price increases. Christopher T. Taylor, et al., Vertical Relationships and Competition in Retail Gasoline Markets: Empirical Evidence from Contract Changes in Southern California: Comment, 100 Am. Econ. Rev. 1269 (2010).

Hastings explains the discrepancy with Taylor et al., by noting the challenges of evaluating vertical mergers with incomplete data or, simply, different data sets, as seemingly similar data can yield very different results. Justine Hastings, Vertical Relationships and Competition in Retail Gasoline Markets: Empirical Evidence from Contract Changes in Southern California: Reply, 100 Am. Econ. Rev. 1227 (2010). But that observation does not undercut Taylor et al.’s findings. Rather, it suggests caution in drawing general conclusions from this line of research, even with regard to gasoline/refiner integration, much less to vertical integration generally.

Other commonly cited studies are no more persuasive. For example, one study examined vertical mergers between cable-programming distributors and regional sports networks using counterfactual simulations that enforced program access rules. Crawford, et al., supra, 86 Econometrica 891. While some have characterized their findings as “mixed” (FTC Hearing #5, supra, Transcript 54 (statement of Margaret Slade))—suggesting that vertical integration could have some negative as well as positive effects—their overall results indicated “that vertical integration leads to significant gains in both consumer and aggregate welfare.” Crawford, et al., supra, 86 Econometrica at 893-894.

Harvard economist Robin Lee, a co-author of the study, concluded that the findings demonstrated that the consumer benefits of efficiency gains outweighed any harms from foreclosure. As he testified at the FTC’s 2018 hearings,

our key findings are that, on average, across channels and simulations, there is a net consumer welfare gain from integration. Don’t get me wrong, there are significant foreclosure effects, and rival distributors are harmed, but these negative effects are oftentimes offset by sizeable efficiency gains. Of course, this is an average. It masks considerable heterogeneity.

FTC, Competition and Consumer Protection in the 21st Century: FTC Hearing #3: Multi-Sided Platforms, Labor Markets, and Potential Competition, Transcript 101 (Oct. 17, 2018), available at

While these studies indicate that vertical mergers can sometimes lead to harm, that point was never disputed. What is important is that the studies do not support any general presumption against vertical mergers or, indeed, any revision to either the legal distinction between horizontal and vertical mergers or to what was, up to now, established agency practice in merger review. The weight of the empirical evidence plainly indicates that vertical integration tends to be procompetitive; hence, no presumption of anticompetitive effects or of illegality should apply, and none should have been applied here.


There is much at stake here. The potential for harm from the merger seems speculative, but the benefits seem conspicuous and substantial, not only reducing the risk of net competitive harm but promising significant enhancement to consumer welfare. As the Commission observed, “better screening methods to detect more cancers at an earlier stage … have the potential to extend and improve many human lives.” Opinion 3. Those benefits should not be forestalled by speculation about possible harms that ignores the differences between vertical and horizontal mergers.

The FTC’s decision should be reversed.

[1] Many discussions of the competitive effects of vertical mergers, including the Vertical Merger Guidelines, conflate EDM, investment benefits, and transactional efficiencies.

Continue reading
Antitrust & Consumer Protection

Comments of Scholars of Law & Economics and ICLE in the Matter of Non-Compete Clause Rulemaking

Regulatory Comments Introduction and Executive Summary We appreciate the opportunity to comment on the Commission’s Notice of Proposed Rulemaking regarding non-compete clauses, Matter No. P201200 (“NPRM”).[1] The . . .

Introduction and Executive Summary

We appreciate the opportunity to comment on the Commission’s Notice of Proposed Rulemaking regarding non-compete clauses, Matter No. P201200 (“NPRM”).[1] The authors and contributors to these comments are scholars of law and economics with an interest in ensuring the effective functioning of the antitrust laws and of the Federal Trade Commission. The full list of signatories can be found in Appendix A, infra.

The Commission’s interest in non-compete agreements, non-compete clauses, non-compete terms, or covenants not to compete (collectively, “NCAs”) is understandable and, at some level, laudable. NCAs have been prominent in recent public policy debates, and numerous NCAs may be overbroad, inefficient, or otherwise objectionable. While most policy concerns regarding NCAs are not antitrust concerns (and most NCA-focused litigation not antitrust litigation), a given employer might possess significant market power in one or more specific local labor markets, and might exploit that market power to, e.g., foreclose entry or expansion by would-be competitors. In that regard, a specific NCA, under specific facts and circumstances, might well prompt antitrust concern and, potentially, a finding of liability.

Nevertheless, as explained below, we cannot recommend that the Commission adopt the proposed Non-compete Clause Rule (“Proposed Rule”). It is not supported by the evidence—empirical and otherwise—that is reviewed in the NPRM; neither is it supported by the Commission’s experience, authority, or resources.

First, while the NPRM amply catalogs potential problems associated with non-competes, NCAs, like other vertical restrictions in labor agreements, are not necessarily inefficient, anticompetitive, or harmful to either labor or consumer welfare; they can be efficiency-enhancing and pro-competitive. NCAs can solve a range of potential hold-up problems in labor contracting.[2] For example, both firms and workers have incentives to invest in employee training, but employees often lack the resources required to acquire adequate training—especially, but not only—job-specific training on their own. Employers, for their part, may have resource advantages; at the same time, employers may reasonably worry about their likely return on investment in employee training: because experienced labor is alienable, firms may worry that competitors will free ride on their investments by poaching trained employees; employees, for their part, may walk out the door or renegotiate compensation before their employer has recouped its investment. Facing those prospects, firms may tend to under-invest in employee training. Appropriately tailored NCAs can mitigate employers’ investment risks, and thereby encourage additional employee training. Firms can face analogous hold-up concerns when it comes to sharing private or privileged information—such as trade secrets or client lists—with their employees.[3] NCAs can mitigate the risk (and risk of hold-up) that firms would face if there were no constraints on job switching. NCAs can also reduce search and training costs by reducing turnover; and the benefits of reduced search costs may be shared, at least to some extent, with employees.[4] As we explain below, these potential benefits find support in both the economic literature and common-law standards of “reasonable” restraints.[5]

Second, and most critically, the emerging body of economic literature regarding the effects of NCAs—or the effects of what is purported to be the relative “enforceability” of NCAs—does not support the categorical ban on NCA usage contemplated by the NPRM.[6] Although the Commission proposes to prohibit NCAs across the economy, there appear to be numerous and broad gaps in the literature. For many sectors, industries, and occupations, there appear to be no studies of NCA’s effects. Moreover, the Commission cites only a single study of the impact of NCA enforceability on downstream prices, and that regards a specific occupation (physicians) delivering heavily regulated services. There are studies investigating wage and mobility effects, but even partial equilibrium analyses of labor markets provide an incomplete picture of the total impact of NCAs on labor markets, even if existing studies are taken at face value. And while some studies do suggest the potential for NCAs to reduce wages or worker mobility under certain circumstances, findings are mixed rather than unidirectional, and many of the relevant studies suffer significant data and methodological limitations. As a working paper from the Commission’s Bureau of Economics notes, the “more credible empirical studies tend to be narrow in scope, focusing on a limited number of specific occupations . . . or potentially idiosyncratic policy changes with uncertain and hard-to-quantify generalizability.”[7] That is not to say that none of the research is useful, but rather that the literature is not comprehensive or settled, and that it cannot support the adoption of sweeping federal regulations that preempt the development of a more nuanced body of state labor and NCA law.

Part of the problem is that measuring this so-called “enforceability” is far from trivial. There is no objective measure of enforceability, and no proven metric for making such a measurement. Studies of enforceability employ similar measurement schema, but these vary in their implementation, and there is no evident benchmark by which to evaluate the alternatives. As we explain in some detail in Section I, below, most of the literature investigating the effects of NCA policy changes—nominally, changes in NCA enforceability—employs one or another version of a triply-subjective scoring rubric. The taxonomy of relevant legal markers, the relative import (that is, weighting) of those markers, and the coding of legal changes all depend on subjective assessments of specific judicial decisions and legislative acts against no specified baseline. None represents the universe of potential policy reforms. And none specifies a theory of enforceability that it seeks to implement. Collectively, the enforceability studies depend on what is, at best, an essentially soft, variable, and heavily coding-dependent method; at worst, it’s a black box. The problem might be avoided going forward. Given recent, clear, income-based restrictions on enforcement in nine states and the District of Columbia, the Commission might well collect data to enable event studies without the artifice of enforceability ordering.[8] But those data do not yet exist and have not been analyzed. The absence of such data, and of any objective enforceability metric, tend to undermine many of the results on which the Commission relies.

In addition, the Commission should consider that most labor markets are local, rather than statewide or national.[9] As a corollary, research suggesting that, e.g., certain wage changes associated with changes in NCA enforceability, on-average and state-wide, do not resolve the question whether observed effects obtain across all (or nearly all) labor markets in the state or, in the alternative, are dominated by effects in those local labor markets in which key employers enjoy heightened market power. Of direct relevance to the Proposed Rule, “[t]there is little evidence on the likely effects of broad prohibitions of non-compete agreements.”[10] Indeed, they do not resolve the question how wage changes are distributed across workers, or whether the observed effects are due to workers receiving raises or, rather, to firm efforts to mitigate hold-up problems by hiring more experienced or better trained workers at higher wages.[11] This remains a developing body of economic literature and—as a related matter—improved data sources. The FTC can and should foster the further development of pertinent economic research before adopting a general rule or, in the alternative, before advising Congress on potential statutory restrictions on NCAs.[12]

Third, the Commission has very little experience with NCAs, several very recent settlement agreements notwithstanding.[13] The three 2023 matters discussed in the NPRM were concluded with consent orders announced the day before the Commission’s announcement of the NPRM. The Commission’s decisions contained no finding or stipulation of an antitrust violation, whether under Section 5 or any other antitrust statute.[14] That does not, of course, establish that the Commission erred in its complaints. Still, the settlements established no legal precedents, and the complaints and orders do little to set forth guidance on the Commission’s applications of Section 5 to the specific facts and circumstances underlying the three matters.[15]

At common law, NCAs might be found reasonable or unreasonable restraints of trade based on their terms, under specific facts and circumstances.[16] Federal law[17] and state laws[18] have tended to hew to this common law tradition, even if state laws vary in their criteria of reasonability. And while some states impose significant limitations on the ability of employers to enforce NCAs in court, no state has adopted the general prohibition on NCA usage that the FTC has proposed. No state chiefly restricts NCAs via a regulatory ban; and no state has adopted the seemingly arbitrary 25% share restriction that the Commission has proposed for permitting certain NCAs in conjunction with the sale of a business.[19] In addition, as noted by several participants in the FTC’s 2020 NCA workshop,[20] courts have tended to find that NCAs do not violate the antitrust laws,[21] even if certain NCAs may violate some state labor or commercial laws. Yet the NPRM contemplates what would be tantamount to a per se prohibition of NCA usage.

The Commission’s view that NCAs are generally, or even typically, anticompetitive seems to lack any basis in antitrust jurisprudence. Looking beyond the Sherman Act jurisprudence, we have not found any decisions by a federal court holding that an NCA violates Section 5 of the FTC Act or, specifically, the Commission’s standalone Section 5 authority over unfair methods of competition. Importantly, while the complaints in the three settled matters identified specific NCA terms, as well as other facts and circumstances, under which the NCAs in question were alleged to violate Section 5, there is no reason to expect that those specific terms or circumstances are representative of the very diverse terms in NCAs, as they are employed across industries, firms, labor markets, and individual employees.

The Commission’s limited experience with NCAs—or any vertical restrictions in labor agreements—undercuts the rationale for a general prohibition of NCAs, but it also undercuts the proposal that the Commission serve as a federal regulator of NCAs generally. Specifically, it does not bode well for likely court challenges to the Proposed Rule or the Commission’s authority to issue it. And while the Commission notes hearings and workshops it has conducted to gather information about NCAs and other labor competition issues,[22] neither the Commission nor its staff has issued any report summarizing or synthesizing information gathered through those inquiries. Such reporting would be consistent with the FTC’s mission under Section 6 of the FTC Act,[23] and it would be an important prologue to any consideration of rulemaking.

Fourth, the Commission lacks the resources required for effective enforcement of the Proposed Rule.[24] According to some survey evidence and the NPRM, NCAs now apply to roughly one fifth of all employed persons in the U.S. labor force; that is, nearly 30 million workers.[25] Regulations are not self-enforcing. And while regulation may be, in certain regards, more streamlined than case-by-case enforcement, it still requires investigation of alleged infractions, administration, and, in addition to regulatory challenge mechanisms, the resources to defend at least some challenges to regulatory determinations and penalties in federal court. Detection alone may often be a challenge to the extent that many “workers are totally uninformed about the law.”[26]

Effective enforcement need not entail detecting, much less penalizing, every violation, but it does require sufficient enforcement activity to establish a credible threat that violations will be penalized, without raising concerns about selective enforcement.[27] Yet the NPRM contains no assessment of the resources required for adequate enforcement of the Proposed Rule or any alternative NCA regulation. Enforcement staff in the Commission’s Bureau of Competition (“BC”) have substantial antitrust expertise in mergers and diverse conduct matters, but little experience in labor matters and none in the enforcement of competition regulations. Moreover, the Commission has recently reported that BC staff are barely able to meet the Commission’s already established and important workload.[28] Adding an obligation to monitor restrictions in labor agreements across all industries and occupations in the U.S. would drain the staff’s ability to scrutinize mergers and conduct under settled antitrust law.

Fifth, it is not clear that the Commission has the authority to adopt the Proposed Rule.[29] There is a grant of some type of rulemaking authority in Section 6(g) of the FTC Act; And there is a 1973 D.C. Circuit opinion in which the court defers to the Commission’s interpretation of the scope of its own regulatory authority. But as participants in the FTC workshop and numerous administrative law scholars have recognized, contemporary courts are unlikely to uphold that degree of agency deference. Recent decisions of the Supreme Court have declined to recognize broad grants of regulatory authority without express statutory language that is both specific and cabined in its grant of authority, and the Court has read the plain language of FTC Act narrowly on the specific question of the FTC’s remedial powers.

Finally, the economic import and the sweep of the Proposed Rule amplify each of the concerns stated above. Subject to very limited exceptions,[30] the Commission proposes to ban the use of NCAs of any duration, and of any geographic or occupational scope, adopted under any business contexts, across the entire U.S. workforce. Moreover, the Commission proposes to ban the maintenance of any existing NCAs, no matter what compensation may have been negotiated or conferred conditional on acceptance of the terms of an NCA. The scope of the Proposed Rule poses a tremendous challenge to the Commission’s experience and resources; it greatly outstrips the evidentiary basis cited on behalf of the Proposed Rule; and it increases the very real legal risk the Commission faces, with regard to both the substance of a rule that the Commission might adopt and the Commission’s regulatory and enforcement authority.

The Commission’s interest in NCAs is laudable. And the Commission is well-positioned to contribute to the further development of economic research regarding NCAs and, specifically, to the further application of Industrial Organization economics to research on NCAs and labor market competition. New research, and a critical synthesis of the relevant hearings and FTC workshops cited in the NPRM, could contribute to case-by-case antitrust enforcement, and to policy debates involving NCAs in Congress and in the states.[31] The Commission is also well positioned to help develop the antitrust case law where NCAs and related vertical restrictions on labor agreements demonstrably harm competition and consumers. These tasks are potentially important; they are tractable, given the Commission’s resources, including its human capital; and they fit well within the Commission’s jurisdiction. They should precede, not follow, a proposed federal NCA regulation.

I.        Empirical Evidence on the Effects of NCAs and NCA “Enforceability” Does Not Support the Commission’s Proposed Federal Ban

There is a significant and developing body of literature investigating the economic import of NCAs, but it does not support the Commission’s Proposed Rule. Much of the NPRM is devoted to a review of the literature regarding NCA usage and the effects of NCAs (or, in many cases, the effects of the relative “enforceability” of NCAs under the laws of the various states). The Commission’s attention to the empirical literature is welcome, and many parts of the discussion comprise useful summaries of published studies or research in progress. Overall, however, the NPRM’s discussion of the literature seems uneven. Some acknowledged limitations in the literature are discussed at some length, and others obliquely or not at all. It is not always clear how reliable the Commission finds the relevant methods or how accurate it deems relevant findings. In addition, some of the NPRM’s extrapolations from the literature seemed strained.[32] The scope of the Proposed Rule—a sweeping federal ban on the use of NCAs, including those already in effect, even if bargained-for—would seem to demand a far more settled and comprehensive body of economic literature, and far less mixed results, than we see in evidence.

Some studies do suggest the potential for NCAs to reduce wages or worker mobility, at least under certain circumstances. But findings on the effect of NCAs on wages are mixed, rather than unidirectional, and many of the relevant studies evidence significant data and methodological limitations. Some of those limitations cast doubt on the extent to which certain findings may be generalized; others may impugn the findings themselves. Moreover, as discussed at the FTC 2020 NCA Workshop, available findings tend to address average effects rather than the distribution of those effects.[33] A substantial number of observations of workers’ wages might vary from the average not just in magnitude but in sign; that is, it may be that wages were observed to increase for a large number of workers, which would be of no small import to antitrust (or to contract law). And a key question turns on the local nature of most labor markets,[34] and is—or should be—of special relevance to merits of antitrust intervention: if a wage effect is observed on average, state-wide, is that effect ubiquitous or is it chiefly driven by local labor markets in which key employers enjoy outsize market power? Also, because these studies cannot distinguish the workers whose wages appear to increase with legal reform, they do not resolve the question whether the observed average wage effects are due to workers receiving raises or, rather, to substitute hiring practices, with some firms seeking to mitigate hold-up problems by hiring more experienced or better trained workers at higher wages.[35]

One notable omission from the NPRM’s substantial discussion of the literature is a 2016 paper by Bishara and Starr, leading contributors to the economic literature on NCAs. Bishara and Starr observed serious research challenges, as well as significant data and methodological limitations to the then-available body of research:

First… identifying the causal effects of noncompete enforceability is a challenging task. Cross-sectional studies must somehow disentangle the effect of noncompete policies across states from the myriad of other potential state policies or state differences that are correlated with noncompete policies. Similarly, studies that examine the before and after effects of a noncompete policy change within a state must separately identify the impact of the noncompete laws from other trends or state level changes that might be occurring simultaneously. These are challenging identification issues to overcome, especially given that very few states have significantly changed their noncompete policies in the last 30 years.[36]

Second, since not all policy changes equally affect the noncompete-signing population, the measurement of noncompete enforceability is necessarily error-ridden without data on who signs noncompetes.[37]

Third, because enforceability is the key variable, not noncompete signing status, assumptions about knowledge of noncompete policies among the various actors must be made.[38]

Fourth, analyses comparing outcomes in high-enforceability versus low-enforceability states cannot disentangle the impact of the potentially increased use of noncompetes in higher-enforceability states from the impact of the noncompete policy on those who do and do not sign noncompetes.[39]

Fifth, the aggregate perspective cannot directly identify the potential micro-mechanisms at work, and thus limits the potential policy options. For example, how exactly might noncompete enforceability reduce mobility?[40]

While the literature has grown since Bishara and Starr’s review, their concerns remain salient. Not incidentally, the NPRM cites at least ten papers cited in the Bishara and Starr critique. However, the Commission seems more confident than Bishara and Starr about the implications of the academic research. For example, when discussing Samila and Sorenson, Bishara and Starr say:

The authors ambitiously conclude that noncompete enforceability “significantly impedes entrepreneurship and employment growth.” Such a conclusion may be too strong, however… [I]t could be that the causal effect of noncompete enforceability on entrepreneurship is positive, but that it is diminished in high venture-capital areas.[41]

Unfortunately, the NPRM recognizes no such qualifications when discussing Samila and Sorenson’s results with respect to new business formation,[42] although the Commission is more reserved when discussing the paper’s results for innovation.[43]

Other research by Bishara and Starr—jointly and separately—is discussed at length, and cited liberally, throughout the NPRM,[44] and Professor Starr participated as a panelist at the FTC 21st C. Hearings, the FTC 2020 NCA Workshop, and the FTC/DOJ 2021 Labor Competition Workshop. Yet Bishara and Starr’s critical review, and the concerns raised therein, are neither cited nor discussed in the NPRM.

Another puzzling omission is a 2019 literature review conducted by staff in the FTC’s Bureau of Economics.[45] That literature review was much discussed in comments submitted to the FTC 2020 NCA Workshop, and at the workshop itself.[46] Yet the McAdams paper is not even mentioned in the NPRM. McAdams observes that economic research regarding NCAs “has made important strides.”[47] However, he also observes mixed results, and he describes numerous data and methodological limitations running throughout the body of literature. Overall, he finds that the “more credible empirical studies tend to be narrow in scope, focusing on a limited number of specific occupations… or potentially idiosyncratic policy changes with uncertain and hard-to-quantify generalizability.”[48] Of direct relevance to the Proposed Rule, “[t]here is little evidence on the likely effects of broad prohibitions of non-compete agreements.”[49]

Research on NCAs is ongoing. Still, most of the studies cited in the NPRM predate the FTC 2020 NCA Workshop and the BE review, and many predate the 2016 Bishara and Starr critique. Not a few of the shortcomings identified in that work were revisited by panelists at the 2020 workshop;[50] these discussions, too, are absent from the NPRM. As a general matter, citations to the records of the workshops and the hearings seem both sparse and highly selective. The NPRM strains to discount positive findings by, among other things, disfavoring research regarding the effects of NCAs themselves in favor of research regarding changes in NCA “enforceability,”[51] conspicuous limitations in the more supporting research notwithstanding. Ad hoc and uneven critical scrutiny aside, the implications of the “enforceability” studies are far less clear than they might seem. As we discuss below, there is no objective metric for “enforceability.” Instead, relative “enforceability” scores result from various—if related—means of scoring disparate provisions of state statutory and judge-made law on a subjective basis.[52] None of these means is authoritative. And even as soft measurement tools, they fail to account for, much less reliably order, the universe of policy options.

A.      The Existing Studies of NCAs Yield Mixed Results

1.        The evidence shows ambiguous effects of NCAs on wages and mobility and supports the argument that they provide procompetitive benefits

Evidence regarding the impact of NCAs on wages is neither definitive nor unidirectional. Rather, as McAdams observed, it “is mixed.” While the NPRM correctly observes that several studies report negative wage effects associated with increased “enforceability” of NCAs[53] or, inversely, positive wage effects associated with decreased or limited “enforceability,” other studies suggest positive wage effects, at least for certain categories of highly compensated workers.

Studies also suggest that the effects of NCAs (or enforceability) are context dependent. For example, Starr, Prescott, and Bishara exploit their 2014 survey on NCA usage to study the impact of signing an NCA on wages and other factors, such as training.[54] They find a significant positive association between NCAs and wages, although they also find that the wage differential depends when employees receive notice of their NCAs: their results suggest that employees who learn of their NCAs before accepting a job offer have 9.7% higher earnings, but employees who learn of their NCAs after accepting a job offer have “no observable boost in wages or training.”[55]

As Alan Meese notes, the top-line lesson of this study is that the typical NCA increases wages, and distinguishing between properly disclosed and improperly disclosed NCAs—and encouraging, not prohibiting, the former—could have significant positive wage effects:

[Starr, et al.] has also found that 61 percent of employee noncompete agreements are disclosed before employees accept employment. Moreover, when employers do disclose such agreements, employees bound by them earn significantly higher wages than similarly situated employees not bound by such agreements. Taken together and viewed in their entirety, these data suggest two distinct results. First, the average impact of employee noncompete agreements is to reduce wages, and this result is driven by a subset of atypical employee noncompete agreements, i.e., those not initially disclosed to employees. Second, where employee noncompete agreements are disclosed, and the typical agreement is disclosed, employees receive higher wages than they would have received had they not entered into such agreements. These higher wages presumably reflect the parties’ expectations—confirmed by the data—that such agreements will induce additional training and/or the production of information.[56]

Further research into the impact of timing—of when employees become aware of a job’s NCA terms—could have significant policy implications. Government intervention to lower workers’ information costs, and reduce employee/employer information asymmetries, might be very different from—and less costly than—interventions that prohibit NCA usage. They may also have implications for the distribution of policy effects across workers, firms, and (downstream) consumer welfare.

There are other studies suggesting contexts in which NCAs might increase wages or compensation. For example, Lavetti, Simon, and White conducted a survey of primary care physicians in five states.[57] Nearly 2,000 respondents provided input into panel data on both the use of noncompetes and various labor market outcomes of interest, such as earnings, incentive-based payments, and patient characteristics. Those survey data were analyzed with and without the findings from a 2011 survey by Bishara on the relative strength of enforceability across the states.[58] The results suggest that—at least for physicians—greater enforceability is associated with higher, not lower, compensation:

Using three years of longitudinal earnings data per physician, we estimate that [NCAs] increase the annual rate of earnings growth by an average of 8 percentage points in each of the first 4 years of a job, with a cumulative effect of 35 percentage points after 10 years on the job.[59]

Analyzing wage growth in terms of enforceability amplifies the difference: cumulative earnings gain over the first ten years is estimated to be 70% among those with NCAs but only 35% for those without them, on average; “comparable estimates are 89% and 36% respectively in the model using variation in state enforceability.”[60]

They also find a higher incidence of patient referrals associated with NCAs,[61] which may imply allocative and search efficiencies, and potentially patient benefits, in addition to whatever benefits accrue to the physicians. As the authors note, physicians present an interesting and distinctive occupational case study, in part because the practitioner-patient relationship may be a distinctive and durable form of human capital[62] (or, in the alternative, of good will), and in part because legal restrictions—notably anti-kickback laws—restrict both explicit and implicit payments or revenue-sharing for referrals.[63] Those distinctions may suggest other occupations worth scrutiny; they also suggest limits to the generalizability of the physician organization findings.

A 2019 study by Kini, Williams, and Yin examines the impact of NCAs on Chief Executive Officer (CEO) compensation. CEOs are distinctive in several ways. For one, due to SEC filings, CEOs are an exception to our typical inability to know which workers are signing NCAs. Second, CEOs are likely to be relatively well informed about the terms of their employment and better equipped to bargain over terms such as NCAs and non-disclosure terms, as well as compensation.[64] The study exploits staggered, state-level changes in NCA enforceability to estimate the relationship between NCA usage on both the CEO compensation and the monitoring of CEO performance.[65] Results suggest that increases in NCA usage and enforceability are both associated with higher total CEO compensation: among other things, the annual total compensation for CEOs with NCAs is 18.4% higher than it is for CEOs without NCAs.[66] Also,

As stricter enforcement enhances the likelihood that a CEO with an NCA [NCA] will be fired for poor performance and limits the CEO’s outside options, the CEO will demand an increase in total compensation for bearing increased job risk. The board agrees to the higher compensation but increases alignment of interest and risk-taking incentives to reduce the possibility of the CEO taking actions that can harm long-term shareholder value but reduce the CEO’s short-term job risk.[67]

Other studies also suggest potential efficiencies associated with NCAs, if not higher wages. Garmaise (2011), for example, studied the effects of NCA enforceability on both executive compensation and firm investment by analyzing both time series and cross-sectional variation in enforceability across the states.[68] He found that greater enforceability reduces both compensation growth and total compensation.[69] In addition, he found greater enforceability to be associated with a shift in compensation towards salary, and increased salary growth, relative to other forms of compensation.[70] These compensation effects represent benefits to the firms: that is, greater enforceability was found to be associated with lower turnover and greater Board of Directors participation, not just lower total compensation growth.[71]

Gurun, Stoffman, and Yonker find decidedly mixed effects. The authors exploit a commercial policy change, rather than a statutory one, to study the impact of NCAs on financial advisors and their industry.[72] Specifically, they use firms’ adoption of the “Protocol for Broker Recruiting” (“Protocol”) as an event. The Protocol permitted a financial adviser to take client lists and contact information, from a firm participating in the Protocol, to a new place of employment without fear of legal action.[73] In effect, the Protocol reduced both NCA enforceability and enforcement for numerous firms and advisors, even in states with permissive enforcement regimes. Unlike other enforceability studies, this was based on firm-specific data that reflect actual changes in both NCA usage and NCA enforcement.[74]

Adviser turnover was observed to increase, initially and temporarily, after firms join the protocol.[75] Because advisors could decamp for new firms without fear of suit, firms became less willing to fire advisors for misconduct, and broker misconduct increased.[76] In addition, by the second year after adopting the Protocol, “client fees increased by about 13% from pre-adoption levels. After three years, fees remain about 18% higher than pre-adoption fee levels.”[77] As the authors note, “[t]hese ?ndings, along with those on higher misconduct rates, call into question whether unlocking clients makes them better off.”[78]

A 2015 study of hair salons by Johnson and Lipsitz did not examine wage per se, but a wage-related aspect of NCAs,[79] surveying NCA use among hair salons by e-mail.[80] Specifically, the study examined the conjecture that NCAs may be used to transfer utility from employees to employers when the market-clearing wage is constrained.[81] Findings supported the hypothesis that the minimum wage will have a negative effect on employment when NCAs are unenforceable, but not when they are.[82] There was also evidence for the proposition that NCAs were surplus maximizing for some salons, but not others; that is, NCAs may be employed by salons that are wage constrained and lack access to credit, to the detriment of the joint surplus (salon plus employee).[83] Like many of the studies discussed in the NPRM, Johnson and Lipsitz depend centrally on survey evidence, and the cross-sectional convenience sample of 218 salon owners[84] is a conspicuous limitation. Still, the study suggests important questions about the total impact of NCAs on labor markets and, for low-income employees, about the potential interaction of NCAs with minimum wage policies on employment. Additional research into these issues with better data could be important, to the extent one is concerned about the total impact of NCAs on labor markets and, especially, on workers.

In the same paper, Johnson and Lipsitz investigate the impact of NCAs on on-the-job training.[85] They find that salons using NCAs are 14% more likely than the mean to provide training to newly hired workers.[86] Starr also observes a training effect—one similar in magnitude—across categories of workers.[87] His results suggest that, if a state were to adopt a policy change in which it moves from non-enforceability of NCAs to average enforceability, the likelihood of worker training would increase 14.7%.[88] Moreover, Starr’s results demonstrate that “the positive correlation between noncompete enforceability and training… is driven almost entirely by firm-sponsored training. The relationship between noncompete enforceability and self-sponsored training is practically zero.”[89] While Starr, Prescott, and Bishara did not find a relationship between training and the timing of employees learning about NCA terms, these and other studies suggest that NCAs can, in fact, ameliorate hold-up problems associated with investments in employee training, as well as potential tradeoffs in labor markets, such as tradeoffs between wages and firms’ investments in employee training.[90]

None of our discussion is meant to suggest that the various cited papers based on natural experiments[91] are without value. Indeed, in a body of literature based on so few natural experiments (relative to, e.g., the literature regarding the effects of minimum wages), we do not simply dismiss all the studies that lack a clear causal design. Our purpose, rather, is threefold. First, we mean to point out that the empirical basis for regulatory intervention is limited, especially when one considers federal regulations that would sweep as broadly as the Proposed Rule. Second, as we discuss below, the emerging empirical picture is more complex—and the results more mixed—than the Commission seems to recognize. Third, as discussed by Bishara and Starr, McAdams, and numerous participants in the FTC’s 2020 workshop, there remain significant data and methodological limitations across the existing body of literature. Collectively, these undercut both the generality of the Commission’s purported findings about the effects of NCAs (or NCA enforceability), and the confidence that the Commission and other policy makers ought to attribute to such findings.

2.        The downstream effects of NCAs on competition and consumers is theoretically ambiguous and empirically unestablished

Setting aside the study of the Broker Protocol,[92] the NPRM notes precisely one study on the downstream price effects of either NCA usage or enforceability, stating that a 2021 paper by Hausman and Lavetti on the effects of physician NCAs, is “the only study of how non-compete clauses affect prices.”[93] That suggests a gaping hole in the literature. Antitrust has not yet abandoned (and should not abandon) its concern with consumer welfare and downstream prices.[94]

At the outset, we might wonder how well a study of physician NCAs and health care services prices will generalize across occupations, products, and services. We might also wonder about endogeneity and identification issues, given data limitations on specialty distribution within firms, myriad state and federal policy changes pertaining to health care reimbursement, background changes in physician practice organization, and a dearth of major state law NCA policy changes in the period in question, 1996-2007.[95] Changes in physician organization have been ongoing for several decades, and include not just a general trend towards consolidation, but increasing vertical integration, as primary and ambulatory care practices are acquired by hospitals, hospital systems, and networks.[96]

The issue of measuring changes in NCA enforceability seems especially salient, given the difficulty of quantifying changes in enforceability associated with legal changes, especially common law ones,[97] soft or subjective elements of the metric used to attempt quantification, and the fact that neither Bishara’s approach to measurement nor its implementation in the study seems ever to have been tested against any objective measures of litigation impact. The nature, sensitivity, and specificity of the metric also seem critical given the study’s findings, which indicate that the sign of the putative effect changes as one shifts one’s focus from establishment-level changes to firm-level changes in provider organization.[98]

The NPRM’s treatment of the Hausman & Lavetti study seems especially puzzling given the Commission’s considerable experience with health care competition matters and, specifically, economic research on health care competition issues conducted by FTC staff in BE. The paper is, in many ways, a careful and thoughtful attempt to investigate the relationship between NCAs and the organization of physician practices. And, indeed, the authors acknowledge various challenges posed by data limitations, among others.[99]

At the same time, the study employs market definitions and analytic methods eschewed in the Bureau’s investigations of health care provider mergers. The NPRM also is unclear on the confidence the Commission attaches to the study’s striking findings:

we find that a 100 point increase in the establishment-based HHI causes a reduction in negotiated prices of about 1.4 percent to 1.9 percent on average. In contrast, the same increase in concentration caused by firm-level consolidation holding fixed establishment concentration causes prices to increase by 1.7 percent to 2.1 percent. OLS specifications imply very small (but statistically significant) positive price effects of 0.02 percent or less, consistent with within-state evidence from Baker et al. (2014).[100]

While we should not dismiss surprising results out of hand, these findings seem more a red flag than a credible interval estimate. 100-point changes in HHI are not at all likely to signal competitively significant events. Small changes in concentration are not necessarily infra-marginal in their price effects, but this is supposed to be a general result across geographic and service markets, and a decade, not surprising observations in specific geographic and service market. As such, it seems highly unlikely, and at odds with both the FTC’s considerable experience with provider mergers and the larger body of health care competition research.[101]

As the Commission is well aware, calculating HHI based on market share is elementary, given a measure of market share: for a given market, one sums the squares of each firm’s percentage market share. That’s it. And, as the Department of Justice Antitrust Division reports:

The agencies generally consider markets in which the HHI is between 1,500 and 2,500 points to be moderately concentrated, and consider markets in which the HHI is in excess of 2,500 points to be highly concentrated.[102]

Consider, for example, a geographic market in which 10 firms (10 group practices) provide general pediatric services. For the sake of simplicity, assume that each firm has an identical 10% market share. In that case, the HHI is 1,000 (that is, 10(102)). Suppose, further, that two of the ten firms merge, such that eight non-merging firms each retains its 10% market share, leaving the merged entity with a 20% share. In that case, the HHI would be 8(102) + 202 = 1,200. That single acquisition would yield a 200-point change in HHI: double the change that is supposed to be robustly associated with significant price increases. The estimate does not seem credible.

The study employs a commercial database that includes the “medical claims for all active employees and their dependents from a sample of large firms,” from 1996-2007.[103] That is a substantial longitudinal (and nationwide) sample, although it is worth noting that this study, like many, lacked access to All-Payer Claims databases,[104] and one might wonder whether the sample from large firms skews the data. Moreover, the data include prices for ambulatory care services only. Hence, the extent to which hospitals as organizations, and even group and individual practices, cross-cut the delivery of ambulatory and hospital-based services may be a confounding factor of interest, as their longitudinal business database permits firm-level observations, but does not identify the specialties of the physicians at each firm—and, as we noted above, there is evidence of ongoing vertical integration in health care provider markets.[105] Identification may be especially important here, as the findings are directionally inverse depending on the choice of firm-level or establishment-level analysis.

Also noted above, the study depends on “a new database quantifying the variation in state-level NCA laws systematically over time, following the measurement system developed by Bishara (2011).”[106] Note that while a number of the “enforceability” studies cited in the NPRM also follow Bishara’s framework, they do not all employ the same scale. Moreover, although the notion of “enforceability”—like the relative stringency of regulations—carries a rough intuitive connotation, there is no objective measure of “enforceability” and, as we discuss below, it is not clear what the study’s ordering system measures, or how well.

Hausman and Lavetti acknowledge that their “modeling approach follows the general structure-conduct-performance (SCP) frame- work for estimating effects of market structure on prices, which has several well-known limitations.”[107] Indeed, while HHIs may still be used for rough and preliminary screening purposes, merger analysis has, by and large, and for decades, left the SCP framework behind, as both theoretical and empirical work has undermined the approach.[108] We would not expect merger screening or analysis to rely upon regressions of HHIs. Does the Commission’s Bureau of Economics contend that they should?

Work from the Bureau of Economics has reinforced the background methodological trend away from the SCP paradigm in provider markets. Both staff and management in the Bureau of Economics have made substantial contributions to the study of competition in health care markets, with a focus on the study of provider consolidation.[109] That research seems to have had a significant impact on the courts’ treatment of provider mergers. Between 1993 and 2000, the federal antitrust agencies (FTC and DOJ) challenged eight hospital mergers, losing all eight challenges.[110] Hospital merger challenges were nearly abandoned, but the losing streak spurred renewed research efforts, both within the Bureau and across the academy. Critically, BE staff undertook a series of merger retrospective studies, ranging from individual case studies to reviews of dozens of consummated provider mergers.[111] These are, in essence, forensic investigations, aiming “to determine ex post how, if at all, a particular merger affected equilibrium behavior in one or more markets.”[112] Such studies complement diverse cross-sectional and theoretical work on hospital mergers, and on provider consolidation more generally.[113] The retrospectives have helped refine merger screening methods employed within the FTC; and they have been widely credited with reversing the way provider mergers are viewed in the courts.[114]

Research on health care competition from BE and elsewhere, coupled with enforcement by the FTC’s Bureau of Competition, represents a signal model of the application of applied industrial organization research to policy development and law enforcement. Notably, this research program militates against SCP assumptions in provider mergers, and against the market definition alternatives employed by Hausman and Lavetti’s study. Results suggest, for example, that various “the new screening tools (in particular, WTP and UPP) are more accurate than traditional concentration measures at flagging potentially anticompetitive hospital mergers for further.”[115] Results also suggest “no statistically significant relationship between post-merger price change and the HHI screens, regardless of the geographic market or share metric employed.”[116] Hausman and Lavetti are aware of the health care competition literature and attempt to address some of its challenges.[117] Still, given BE’s research, and given the unlikely numerical findings, the NPRM’s discussion of potential limitations to this single study of the downstream effects is curiously oblique:

Generally, greater concentration may or may not lead to greater prices in all situations and may arise for reasons which simultaneously cause higher prices (indicating, therefore, a noncausal relationship between concentration and prices). In this case, the authors claim that researching the direct link between changes in law governing non-compete clauses and changes in concentration allows them to identify a causal chain starting with greater enforceability of non-compete clauses, which leads to greater concentration, and higher consumer prices.[118]

Both points seem correct as far as they go, but the NPRM is entirely unclear on the question what they imply for the significance of the study’s findings. The NPRM states that “[t]here is evidence that non-compete clauses increase consumer prices and concentration in the health care sector.”[119] In the NPRM’s introduction, the suggestion is broader: “research has also shown that, by suppressing labor mobility, non- compete clauses have negatively affected competition in product and service markets in several ways.”[120] Perhaps, but the Commission has identified only one study indicating downstream price effects. Does the Commission find the evidence credible? Or generalizable? The NPRM continues to expound on the study’s dubious findings, and on conjectures about the mechanisms at play, at some length.[121] It also extrapolates on the reported findings, suggesting that they are reinforced by “another study, by Michael Lipsitz and Mark Tremblay, [that] shows increased enforceability of non-compete clauses at the state level increases concentration, as measured by employment-based HHI.”[122] Does the Commission deem that finding important?

None of this is to say that NCAs cannot have an anticompetitive effect in health care markets, and it’s certainly not meant to suggest that provider consolidation cannot be anticompetitive. BE research and FTC enforcement have demonstrated that health care provider mergers and acquisitions can be anticompetitive, under certain facts and circumstances. Many hospital markets are highly concentrated—on any measure—and providers of health care services who have market power might employ NCAs to create (or exacerbate) barriers to entry in both those services markets and input markets, such as professional labor markets. Many provider markets are subject to regulatory barriers to entry as well, such as state law Certificate of Need or Certificate of Public Advantage regimes,[123] which might interact with restraints on labor mobility. Rule of reason inquiry into physician NCAs in specific labor (and service) markets might well find harm to competition and consumers. And further economic research, such as that commenced by Hausman and Lavetti, might well foster successful and pro-consumer antitrust enforcement. But there are serious reasons to doubt the specific interval estimates produced by the one price study available, and there remain questions about the importance of context in assessing the effects of NCAs, and of the distribution of average NCA effects (of whatever accuracy), across distinct labor markets.

In any case, the substantial literature on health care competition, and the distinctive characteristics of health care product, service, and labor markets—highly regulated at the state and federal levels, and subject to a complex mix of public and private payment—strongly suggest that one cannot reliably generalize the results of a single study on NCAs and ambulatory care prices across the entire national work force, much less to the downstream price effects of NCAs across industries, products, and services markets.

3.        The weight of the evidence does not support the claim that NCAs decrease innovation

The Commission argues that the “weight of the evidence indicates non-compete clauses decrease innovation.”[124] The “weight of the evidence” is unclear. There are, indeed, some studies suggesting that greater NCA “enforceability” is associated with some innovation-relevant harm. The main paper that seems to fit the Commission’s model of reliable studies on the topic examines seven legal changes from 1992-2008, which were reported to increase or decrease the level of a state’s NCA enforceability.[125] That paper finds, according to the NPRM, “that the value of patents, relative to the assets of the firm, increase by about 31% when non-compete clause enforceability decreases.”[126] But overall, findings are mixed, the literature is hardly settled or comprehensive, and there remains the question of the confidence one should attach to existing studies, separately or in aggregate.

For the papers that the Commission cites, two find that the enforceability of NCA increases in innovation, one finds a decrease, and one is ambiguous.[127] On the one hand, citing reasonable limitations, the Commission suggests that it puts relatively less weight on those studies. On the other hand, the Commission seems sufficiently confident to conclude that “enforceability broadly diminishes the rate of innovation,”[128] based on one paper that looks at value of patents, which is but one of several commonly used, and oft-debated, measures of innovation.[129] Later, the Commission admits it “is unable to extrapolate from the relevant studies to quantify or monetize this benefit.”[130]

As a background matter—and conspicuous in the economic literature on innovation—innovation (and rates of change in innovation) can be hard to quantify, in part because there are diverse indicators of innovation, but no definitive one.[131] Patents have value and some connection with innovation, but patents vary wildly in their value.[132] Value-adjusted patents are better indicators, but patent value, and the time frame in which it’s best evaluated, may be hard to assess, as evidenced by, e.g., FRAND disputes,[133] or the bundles or thickets in which many patents are sold or licensed. Also, patents may be more (or less) relevant given the technology in question, just as trade-secrets and copyrights might have greater or lesser significance depending on the sector and the nature of the tech; for example, trade secrets and copyright might have greater import in areas as diverse as software and biotech. Factors such as venture capital funding, the establishment and growth of startups, etc. also are significant but, again, of varying significance relative to other signals.[134]

Second, the theoretical impact of NCAs on innovation is ambiguous, and empirical findings regarding the complex subject of innovation suggest mixed effects associated with NCAs (or, more commonly, with changes in NCA enforceability).[135] For example, a 2018 study by Starr, et al., examines the impact of greater NCA enforceability on the creation, growth, and survival of spinouts and other new entrants, based on matched employer-employee data on 30 states and 5.5 million new firms. On the one hand, it finds that greater enforceability is associated with fewer within-industry spinouts; on the other hand, the within-industry spinouts created in greater enforceability states “tend to start and stay larger, are founded by higher-earners, and are more likely to survive their initial years.”[136] They find no impact on entry by firms that are not within industry spinouts.[137] They suggest that greater enforceability may screen the formation of within-industry spinouts by dissuading founders with lower human capital.

A recent working paper by Jeffers suggests that certain labor frictions in knowledge-intensive occupations can play an important role in investment decisions.[138] Using matched employee-employer data from LinkedIn, Jeffers finds that increases in NCA enforceability led to 7-11% declines in worker departures for workers in those occupations where the majority of workers have at least a bachelor’s degree. Those declines, in turn, led to increased investment by those firms that rely more on knowledge intensive occupations.[139]

As we discuss below (and as noted in the NPRM), Marx, et al., 2009 exploited a Michigan statutory change—one deemed to increase NCA enforceability—to study worker mobility—specifically, innovator mobility.[140] Their findings suggest the increased enforceability was associated with lower mobility—or job switching rate—of inventors (roughly, employees who are patent holders). However, a 2019 study by Carlino exploited the same legislative event to investigate the effect of NCA enforceability on startups and job creation.[141] Based on a difference-in-differences analysis, he found that an increase in NCA enforceability had a small effect to none-at-all on startups, and a very small, if positive, effect on job creation.[142]

Third, a 2020 paper by Barnett and Sichelman in the University of Chicago Law Review reviews ambiguities and limitations (including plain errors) in the NCA innovation literature in detail.[143] One of its key observations is that almost none of the relevant studies has a causal design; that is, the studies that employ cross-sectional regressions cannot be said to show that changes in NCA enforceability cause the observed effects.[144] We do not recapitulate their article here, but we commend it to the Commission as another important commentary on the available literature. We note, specifically, as we discuss in Section I.B.3, infra, their observation that several of the event studies cited by the NPRM depend on oversimple, and in some regards erroneous, readings of Michigan law.[145]

We do not suggest that any specific mixed or positive findings be considered definitive. Rather, the piecemeal, mixed, and in some regards infirm findings might be considered suggestive as to some of the impact of NCAs on factors associated with innovation, but they cannot be considered adequate grounds for the general conclusion that “non-compete clauses decrease innovation”; certainly, they are inadequate if they are to be considered a significant plank in the justification of a sweeping federal ban on NCA usage.

B.      The Existing Event Studies Depend on Eccentric Events and Their Results Are Not Sufficiently Generalizable

The NPRM notes the importance of event studies —“‘natural experiments’ resulting from changes in state law”—to assess the effect of changes in state law on earnings.[146] According to the NPRM, “[t]he use of a natural experiment allows for the inference of causal effects, since the likelihood that other variables are driving the outcomes is minimal.”[147] That observation should be subject to significant qualification, but we agree that event studies can support causal inferences and that, broadly speaking, they represent an important means of investigating the economic implications of policy changes.

However, as Starr and others have noted, observable variation in NCA law had long been limited, myriad subtle differences across the states notwithstanding. Workshop panelists and others have noted the difficulty of estimating the causal impact of NCA use, due in part to a dearth of exogenous variation.[148] As Bishara and Starr put it in 2016, “very few states have significantly changed their noncompete policies in the last 30 years.”[149] More recent changes in state NCA law may be more significant. In the past several years, nine states and the District of Columbia have adopted income- or wage-based limits on NCA enforcement.[150] These may yield data for informative event studies without the artifice of “enforceability” measurement, but the studies cited in the NPRM predate these statutory changes. Without such changes, the many, highly varied, and mostly subtler legal changes that had been available constrain the likely generalizability of existing NCA event studies. As McAdams observed, “the more credible empirical studies tend to be narrow in scope, focusing on a limited number of specific occupations (e.g., executives) or potentially idiosyncratic policy changes with uncertain and hard-to-quantify generalizability (e.g., banning non-competes for technology workers in Hawaii).”[151]

1.        Hawaii changed more than NCAs but only for a small number of tech workers

The Hawaii technology workers study, Balasubramanian, et al., is in many regards well designed and well executed. It exploits a 2015 statutory event in Hawaii to study the effect of NCA enforceability changes on employee wages and mobility.[152] The authors “find that Hawaii’s 2015 CNC ban increased new-hire monthly earnings by 4.2 percent, while overall (that is, all worker average) monthly earnings rose 0.7 percent.”[153] Supplementing their initial Hawaii analysis with a cross-state analysis, they find that “eight years after starting a job in an average enforceability state, technology workers have about 8 percent fewer jobs and 4.6 percent lower cumulative earnings relative to equivalent workers starting in a nonenforcing state [NCA].”[154] They suggest that their results are consistent with the notion of a significant lock-in effect associated with NCAs. The finding could be especially significant, as the results suggest that the NCA enforcement effects are not confined to low-wage workers.

At the same time, the study seems to illustrate all of McAdams’s general concerns about “the more credible empirical studies.”[155] Regarding workers’ occupations, the authors correctly observe that the relevant statutory change pertained to the tech sector, or “an employee of a technology business.”[156] And the statute established that NCAs for tech workers “shall be void and of no force and effect.”[157] But there are several wrinkles here.

First, the statute did not so neatly address tech workers. Under the Hawaii statute, tech workers do not include employees of “any trade or business that is considered by standard practice as part of the broadcast industry or any telecommunications carrier.”[158] That is, the statute covered tech workers, but not those in the telecom or broadcast industries, or, indeed, any tech workers employed by any firm “other than a trade or business that derives the majority of its gross income from the sale or license of products or services resulting from its software development or information technology development.”[159] Administrative assistants at tech firms were in, but, say, programmers in telecom, government, education, transportation, or health care were out.

Second, as the authors acknowledge, the occupational definition was not the law’s only idiosyncrasy. The legislation did not simply apply to NCAs. Rather, it restricted a “noncompete clause or a nonsolicit clause in any employment contract.”[160] Given sufficient state law variation, that wrinkle could be a feature rather than a bug: post-employment restrictions often are bundled,[161] and one might like to study the effects of changes in the law bearing on the various elements of the bundle, jointly and severally. Given the current levels of variation across state laws, it is a limitation.

In addition, it’s unclear how much of a change the law effected, even for tech workers (as defined), and for NCAs and non-solicit clauses. As the authors acknowledge, the statute was not retroactive;[162] that is, it would apply to new employment agreements, going forward from the effective date, but not to those already in effect. Workers already covered by NCAS were still covered. Moreover, prior to the statutory change, Hawaii NCAs already were subject to “a reasonableness analysis.”[163] NCAs could easily fail Hawaii’s reasonableness test, as Hawaii courts had considered “the benefits to the employer from noncompete or nonsolicit agreements” to duplicate those of trade secret law, and hence “impose undue hardship upon employees of technology business and the Hawaii economy.”[164]

Finally, we might wonder whether the Hawaii tech sector (as defined under Hawaii law) is representative of tech sectors in other states. Hawaii is a very small state, with a total population (not just its workforce) numbering approximately 1.4 million (approximately 1.36 million in 2010, and approximately 1.44 million in 2022).[165] And Hawaii does not appear to have a vibrant tech sector, even relative to its small size. One source suggests that there is not a single tech firm among the 100 largest employers in Hawaii.[166]

In sum, we have a key legislative event that pertains to one industry (not necessarily one occupational category), on an idiosyncratic definition of that industry; the legal change did not apply exclusively to NCAs, it did not apply retroactively to existing NCAs, and it changed the enforceability of NCAs relative to an uncertain, but apparently somewhat stringent, standard of reasonability. It did so in a very small state, where the workforce included some tech workers, but no significant tech industry to speak of. The authors responsibly acknowledge a few of these idiosyncrasies, and their potential to raise “concerns about generalizability.”[167] But that seems to put it mildly. It’s entirely unclear whether observations that turn on Hawaii’s 2015 NCA legislation can be generalized at all, whether to a potential ban on NCA usage, to changes in NCA enforceability that apply beyond the tech industry, or to potential changes in enforceability pertaining to either the tech industry or tech workers anywhere outside Hawaii. To its credit the FTC, likewise, acknowledges the concern about generalizability.[168] At the same time, the Commission seems comfortable making a “preliminary finding” of estimated wage effects across the nation, occupations, and industries based on a “back-of-the-envelope” extrapolation from unpublished findings regarding an idiosyncratic statutory reform in a state with a very small workforce and – even given the state’s small size – a relatively small tech industry.[169] Picking the mid-point of this back-of-the-envelope range estimate does not make the Commission’s preliminary estimate “conservative,” but highly speculative. Respectfully, this does not seem suitable as an estimated effect for a Proposed Rule that would regulate tens of millions of labor agreements.

2.        Oregon banned NCAs for hourly and low-wage workers during the depths of the Great Recession which muddies general applicability

A 2019 paper by Lipsitz and Starr exploits a 2008 statutory change in Oregon’s NCA law that “banned [NCAs] for hourly and low-wage workers.”[170] The Oregon statutory change, like the Hawaii legislation discussed above, is of interest in part because relatively little of the considerable state-by-state variation in NCA laws has to do with the simple binary question whether, for some tranche of the workforce, NCAs are or are not enforceable in court. And Oregon seems in several regards less of an outlier than Hawaii. First, it’s a substantially larger state;[171] and second, a statutory change focused on hourly and low-wage workers may be more generalizable than one that applies to an eccentric segment of the tech industry in a state lacking a significant tech industry. Looking specifically at hourly workers and comparing observed changes in Oregon against several groups of control states, Lipsitz and Starr find that, “on average, banning… [NCAs] increased the earnings of hourly workers [in Oregon] by 2-3%, with stronger effects for those in jobs most likely to sign… [NCAs], while raising monthly job-to-job mobility by 17%.”

Those are significant effects but, as McAdams notes, the study is subject to potentially confounding factors.

First, Oregon’s 2008 statutory change coincided with the beginning of “the Great Recession”; that is, with the most significant recession since the Great Depression of 1929-39.[172] McAdams also observes that “[r]esearch on regional recessions finds that the timing of recessions (both the onset and recovery) differs across states,” including states in the same census region.[173] Hence, the timing of the statutory change may be regarded as “unfortunate,” from a research perspective. Indeed, it

raises the possibility that the paper’s estimated effects are confounded by macroeconomic factors that—similar to [NCAs]—also influence wage growth and worker mobility, as well as by the differential policy responses by states. Indeed, in Lipsitz and Starr (2019), the mobility of workers in Oregon increased (relative to control states) soon after the ban took force in 2008, but average wages did not increase until a full three years post-ban (in 2011). Actual (or threatened) worker mobility is an important channel through which we expect workers to achieve wage growth in Oregon after its ban on non-competes. The fact that Oregon saw an increase in mobility without an increase in average wages raises the possibility that there are confounding factors at play.[174]

Second, as with Hawaii, we might question the extent to which the 2008 statute changed the state’s law regarding NCAs and low-wage workers. Lipsitz and Starr state that they examine low-wage workers specifically to “focus our empirical analysis on the subset of workers for whom NCAs [NCAs] were enforceable before 2008, but were clearly voidable after 2008.”[175] But Lipsitz and Starr themselves note that the NCA restrictions were not retroactive.”[176] Hence, low-wage workers who did not change jobs were not among the subset of workers against whom NCAs were enforceable before 2008 but not after, at least not until a post-2008 change of jobs. And as with Hawaii, there were certain other wrinkles in the state law. Exceptions to the employees covered by Oregon’s NCA limits included not just professionals—or persons “engaged in administrative, executive or professional work who: (a) Performs predominantly intellectual, managerial or creative tasks; (b) Exercises discretion and independent judgment; and (c) Earns a salary and is paid on a salary basis,”[177] and inter alia, federal employees at any wage level,[178] various agricultural workers, including those paid on a piece-rate,[179] and a person “principally engaged in the range production of livestock and earns a salary and is paid on a salary basis,”[180] persons “employed in domestic service on a casual basis in or about a family home,”[181] and persons “engaged in the capacity of an outside salesperson or taxicab operator.”[182]

In addition, the authors, citing a 2008 law review article by Rassas, note that, pre-2008 Oregon NCAs were subject to a reasonability test, involving “criteria meant to ensure that legitimate business interests were being protected without unduly harming workers.”[183] That, of course, raises the question of the extent to which Oregon courts, prior to 2008, found NCAs for low-wage workers to serve legitimate business interest without harm to those workers.

The law review article they cite provides no objective measure, but it plainly suggests that Oregon courts, and indeed Oregon statutory law, were skeptical of NCAs prior to 2008. As Rassas observed,

The former Oregon statute attempted to balance competing interests of the employee and employer by mostly “codify[ing] the basic common law rules” of reasonableness. Oregon courts imposed additional requirements for enforcement, tipping the balance in favor of the employee’s interest in mobility….

Oregon statutory law mandated that non-competes in any industry were void unless “entered into upon the: (a) [i]nitial employment of the employee with the employer; or (b) [s]ubsequent bona fide advancement of the employee with the employer.”[184]

Statutory limits on NCAs in Oregon had been in place, undergoing piecemeal changes, since 1977.[185] Reviewing the case law, Rassas emphasizes that “Oregon courts did not take these requirements lightly.”[186] They imposed, among other things, both geographic and temporal limits on NCAs, which they deemed “‘covenant[s] in restraint of trade,’ the enforcement of which generally runs counter to public policy.”[187] In addition, decisions by federal courts in the Ninth Circuit reinforced the substance of Oregon’s statutory restrictions on NCAs. For example, in Nike, Inc. v. McCarthy, the Ninth Circuit found it…

apparent that the legislature intended to permit employers to require existing employees to agree to a noncompete agreement, so long as the employee’s job content and responsibilities materially increased and the employee’s status within the company likewise improved.   Otherwise, the employer would merely be imposing a new condition for the “same job.”  Id. Thus, an advancement would ordinarily include such elements as new, more responsible duties, different reporting relationships, a change in title and higher pay.[188]

And in Ikon Office Solutions, Inc. v. Am. Office Prods., Inc., a federal district court held that, “[u]nder Oregon law, the right not to be subjected to a non-competition agreement, except as authorized… is an ‘important employment-related statutory right.’”[189]

We do not argue that Oregon’s 2008 legislation was inframarginal in its effects, or that it did not increase the cost of enforcement of NCAs for at least some employers of low-income workers.[190] Rather, given the statutory idiosyncrasies, and the complex pre-2008 restrictions, the magnitude of the change (on any clear measure) is uncertain. Indeed, it is not at all apparent that it represented a major change for hourly and low-income workers. For those reasons, and the confounding timing of the statutory change at issue and the Great Recession, it is not at all clear how the magnitude of Oregon’s 2008 change in enforceability compares—or should be compared—with the disparate legal changes observed in control states.

We can ask a further question. When measuring or ordering the relative enforceability of state NCA laws, how should we assess, e.g., restrictions pertaining to a specific occupation (such as, e.g., tech industry employees in Hawaii) relative to those pertaining to, e.g., a certain income level, given that the specifics of the statutes vary? We might consider the percentage of the state’s workforce fitting a categorical restriction under state law, the percentage actually or likely covered by NCAs, or various other measures, and we might consider the domain of the restriction somehow normalized according to, e.g., the stringency of limitation. There is no objectively correct way to do this, but one or another means might be more or less useful for economic or antitrust analysis; and, in any case, we might want to know how it is being done within any given study and across the “enforceability” literature.[191]

3.        Michigan’s statutory changes were not a clear switch from unenforceable to enforceable and back again

Several papers exploit 2005 statutory changes in Michigan—the Michigan Antitrust Reform Act (“MARA”)[192]—with or without a subsequent amendment in 2007, to investigate the impact of NCA enforceability on worker mobility, especially as it relates to innovation. MARA—perhaps inadvertently, increased the enforceability of NCAs. Marx, et al. 2009[193] found that the increased NCA enforceability permitted by MARA reduced the mobility—or job switching rate—of inventors; that is, roughly, employees who were patent holders. In a follow-up study, Marx, et al. 2015[194] found what might be viewed as a “brain drain”:

from Michigan to non-enforcing states following the… policy reversal: during a symmetric window from 1975-1996 surrounding [the change], the rate of emigration to non-enforcing states grew in Michigan (0.24%-0.32%) while dropping in states that did not enforce non-competes. The relative risk of post[change] emigration was 1.35 in Michigan, twice as high as in states that continued not to enforce non-competes.[195]

Barnett and Sichelman demonstrate in detail that these studies evidence significant problems in both data and analysis.[196] A central concern has to do with the legal analysis underlying the assessments of changes in enforceability. Marx, et al. (2009)[197] and Marx, et al. (2015)[198] both suppose that NCAs were generally unenforceable, prior to MARA’s enactment in 2005, under a statute providing that “[a]ll agreements and contracts by which any person…agrees not to engage in any avocation or employment…are hereby declared to be against public policy and illegal and void.”[199] They also argue—not without evidence—that MARA’s repeal of Public Act 329 was inadvertent. They also note a 2007 statutory amendment to the pertinent provision of MARA, which represented its retrenchment (not rescission).[200] Specifically, the 2007 amendment added a “reasonableness doctrine” that “did not reinstate the previous ban.”[201]

Neither event was quite what the research supposed. As Barnett and Sichelman explain, the assumption that NCAs were unenforceable in Michigan prior to 1985, but generally enforceable from 1985-87,[202] seems to misread the law. Prior to 2005, Michigan courts might uphold NCA terms or approve changes of venue due to, e.g., choice-of-law provisions in the NCAs or the larger employment agreements within which they were situated. Perhaps more important, the authors of the Michigan studies

appear to overlook that MARA included a “savings clause” that provided that the statute repealed by MARA would “remain in force for the purpose” of enforcing any liability under the repealed act. Consistent with the saving clause, Michigan courts declined to enforce NCAs that were entered into prior to MARA.[203]

That savings clause has implications for both the 2005 and 2007 events. The 2005 adoption of MARA had no bearing on NCAs entered-into prior to the law’s enactment and, hence, no bearing on employees actually or putatively subject to NCAs before 2005. Multi-state firms with strong incentives to employ NCA terms would have had a natural incentive to use choice-of-law provisions to impose or maintain those terms in Michigan pre-2005. The studies assume that California’s relatively recent decisions disfavoring the application of such choice-of-law clauses to NCAs in California represent the general case, but, as Barnett and Sichelman demonstrate, it does not.[204] Moreover, operative NCAs would have included not just employees subject to extra-Michigan NCAs, but some employees whose employment agreement documents pre-2005 included NCA terms, even if, for those employees, the terms were, at least arguably, unenforceable under Michigan’s prior law. As the NPRM notes, and as several authorities have observed, employment agreements commonly contain NCA terms, even in states where such terms are unenforceable; and NCAs are common in engineering and other technical occupations. There are also questions when, to what extent, and on what terms, the 2005 policy change fostered the negotiation (or imposition) of NCAs on tech professionals whose employment remained unchanged from 2005-07.

The savings clause also has implications for Michigan’s 2007 policy reform, beyond whatever retrenchment was accomplished by the savings clause post-2007. Because the 2005 policy change was smaller than the studies suppose, the effect of its 2007 retrenchment was also smaller than the studies suppose. As noted by Barnett and Sichelman, errors and ambiguities in assessing the magnitude of legal changes are especially salient for the Marx et al. [2015] study (as well as a 2009 study by Marx and others), given the relatively small decrease, in absolute terms, in labor mobility observed in Michigan. The 2009 Marx et al. study considers 98,468 inventors and 27,478 inventor moves within Michigan over the period 1963-2006. In absolute terms, labor mobility increased post-MARA over the full time period from 7.18% to 8.98%, although other “non-enforcing” states saw a larger increase, from 7.95% to 10.80%.[205]

While the Marx, et al. studies never report these differences in absolute numbers, they are easy to calculate. Specifically, the difference of in-state mobility in Michigan versus non-enforcing states in absolute terms was roughly 1%, equating to an absolute difference of about 100-200 moves per year purportedly lost within Michigan due to the enforcement of noncompetes. For inventors moving out of Michigan, the numbers are much lower.[206]

In brief, errors and uncertainty in assessing legal changes in Michigan and in control states takes on an outsize import given the small number of job changes potentially attributable to the Michigan statutory change.[207]

One might suppose that such misinterpretations of the law represent “mere” coding errors, and that such errors are occasional (and sometimes minor or debatable), adding some degree of random error, and hence noise, to signals of the economic impact of policy changes, while leaving findings directionally—and approximately—intact. But the Michigan case should remind us that, with small numbers of observations and/or small effects, recoding might well render previously observed effects statistically insignificant or nil. As we have seen above, such coding issues seem significant across key event studies in the literature, rather than outlier events. And as we discuss in Section III.D below, such issues point to fundamental questions about the meaning and reliability of the “enforceability” metric on which so many studies—and the Commission’s conclusions—rely.

4.        California is not a clean event study due to California’s unique attributes

Policy discussions of NCAs often look to Gilson’s 1999 paper,[208] and a few follow-on cross-sectional studies, suggesting that California’s hostility to the enforcement of NCAs helps explain the rise of Silicon Valley and what’s taken to be the fall to tech innovation in Massachusetts.[209] Barnett and Sichelman dissect these arguments with some care, and we commend their discussion to the Commission, even as aspects of Gilson’s comparison now seem dated.[210] Gilson’s account is interesting, but in scientific terms, the Silicon Valley/Rt. 128 comparison seems more of a “just-so story,” than an empirical vindication of any specific theory about NCAs. At best, it is an existence proof for the claim that relatively stringent limits on the private enforcement of NCAs can, under some facts and circumstances, co-exist with vibrant tech innovation. But that proposition is not much at issue.

California would present an especially difficult case for an event study, not least because of timing questions. California’s NCA policy is anchored by a provision in the state’s Business and Professions Code from 1941, [211] and that provision has both statutory and case law roots dating to the 19th Century.[212] Data problems for a credible event study abound, and not just because key events in the state’s growth—as a center of innovation and otherwise—are hard to tie to any specific legal events regarding NCAs.

C.      Because There is No Objective Measure of “Enforceability,” Many of the Causal Studies Contain a Fatal Methodological Weakness

As we noted above, the Commission is well-positioned to contribute to the further development of economic research regarding NCAs and, specifically, to the further application of Industrial Organization economics to research on NCAs and labor market competition. A critical synthesis of the relevant hearings and FTC workshops could contribute to policy debates involving NCAs in Congress and in the states.[213] Additional studies, and the development of better data sources—perhaps in cooperation with the Department of Labor—are well within the staff’s competence, and these too could better inform policymaking and state and federal law enforcement. Moreover, the Commission is well-positioned to help develop the antitrust case law where NCAs and related vertical restrictions on labor agreements demonstrably harm competition and consumers. These tasks are potentially important; they are tractable, given the Commission’s resources, including its human capital; and they fit well within the Commission’s jurisdiction. They should precede, not follow, a proposed federal NCA regulation.

In the NPRM’s account of the empirical evidence, the Commission notes that:

The belief that studies of non-compete clause use do not reflect causal estimates is shared by the authors of at least one of the studies of non-compete clause use. As noted in Starr et al., ‘‘Our analysis of the relationships between noncompete use and labor market outcomes… is best taken as descriptive and should not be interpreted causally.’’ As a result, the Commission gives these studies minimal weight.[214]

We agree that it is important to distinguish between correlation and causation. That is not to say that none of the non-causal studies is suggestive, but we note that the studies to which the Commission ascribes “minimal weight” constitute a significant portion of the available literature. Central to much of the literature—including most of the papers the Commission seems to consider causal—examine the putative effects of NCA “enforceability,” or of changes in levels of enforceability, under state law.

On the surface, there is some intuitive appeal to this approach for several reasons. For one, there is survey evidence on the incidence of NCA usage within and across states, but little evidence on the individuals bound by (or perceived to be bound by) NCAs, so it is difficult to study the impact of NCA usage directly. Second, one might suppose that evidence on the effect of various policies (and policy changes) bears directly on the question what legal policy, if any, to impose. Third, at a high level of abstraction, we might have an intuitive sense that some regulations are more stringent than others, and that some jurisdictions are more (or less) plaintiff friendly, whether with regard to NCAs specifically or across most civil suits. For example, it seems plain enough that the decisions of California courts, applying California Business and Professions Code Section 16600, recognize a stringent restriction on a plaintiff firm’s ability to enforce the terms of an NCA against an employee.[215] Hence, we can think of California as a “low enforceability” state.

However, there is no objective measure of “enforceability,” and, hence, no established metric with which to detect or approximate such a measure. And if we seek to unpack the notion of enforceability, as prologue to identifying or formulating a useful metric, it seems clear that any number of factors or end points might be relevant to our high-level intuition. For example, we might be interested in the cost (average, median, or modal) of litigating an NCA dispute to its conclusion; we might be interested in the ratio of plaintiff to defendant success in litigating such cases to their conclusion; we might be interested in the frequency with which NCA claims are filed and, if filed, settled or, in the alternative, survive, e.g., motions to dismiss for failure to state a claim or motions for summary judgement. We might also be interested in the way the law—and these various factors—affect not just the incidence of NCAs in a state, but the distribution and terms of those NCAs (or the terms of those NCAs employers seek to enforce). These are all related factors, but they are not equivalent, and, a priori, there is no obvious set of them, or weighted sum of them, that is best for all (or any specific) policy purposes.

Given the centrality of “enforceability” to the Commission’s empirical brief for regulation, the underlying enforceability metrics and measurements deserve serious scrutiny. Before turning to the specifics of the scoring tools employed in the various NCA studies, we note that the cited studies of enforceability do not use the same metric, even if many of them share some basic assumptions or sources.

The Commission observes that the various studies are based on Malsberger’s treatise, Non-compete Clauses: A State by State Survey, with some augmented by the 50-state survey conducted by Russell Beck.[216] The Commission also suggests that, while the “studies have defined enforceability of non-compete clauses in slightly different ways, each uses enforceability as a proxy for the chance that a given noncompete clause will be enforced.”[217] It is not at all clear that the claim is correct. That is, at least most of the studies appear to lack any express claim about that proxy, and it is not at clear that anyone has ever investigated empirically the link between such measures and such a likelihood. Perhaps it is simply the ratio of suits (perhaps successful) to employees (putatively?) bound by NCAs, or perhaps the likelihood that an NCA will be enforced, conditional on, perhaps an (arguably) covered employee’s departure to subsequent employment, or perhaps the employee’s departure to a competing employer…. What is more, the differences in measurement approaches are not obviously trivial. Some of the key studies take pains to critique the way other (apparently key) studies seek to implement their assessments of enforceability.[218] Differences in the approach employed may be especially important when considering relatively small effects, relatively few observations, or analyses based on correlations that barely meet significance thresholds.

Second, the various approaches to measuring enforceability are all soft measures; that is, they depend on subjective judgments, and, indeed, on series of subjective judgments. Most of the relevant studies are based, to some extent, on a periodic 50-state review of NCA law by Malsberger and others, as well as a set of accompanying questions suggested to guide state-by-state assessments of NCA laws, as published.[219] For example, Bishara’s 2011 study examines state statutory and, chiefly, decisional law regarding NCAs and, based on twelve criteria of enforceability identified by Malsberger, applies “seven questions because they directly address the legal issues relevant to measuring a given jurisdiction’s intensity of noncompete enforcement.”[220]

  1. Is there a state statute of general application that governs the enforceability of covenants not to compete?”
  2. What is an employer’s protectable interest and how is that defined?”
  3. What must plaintiff be able to show to prove the existence of an enforceable covenant not to compete?”
  4. [numbered 3a by Bishara, but ranked separately] Does the signing of a covenant not to compete at the inception of the employment relationship provide sufficient consideration to support the covenant?”
  5. [labeled 3b and 3c, and scored jointly] Will a change in the terms and conditions of employment provide sufficient consideration to support a covenant not to compete entered into after the employment relationship has begun?
  6. Will continued employment provide sufficient consideration to support a covenant not to compete entered into after the employment relationship has begun? If the restrictions in the covenant not to compete are unenforceable because they are overbroad, are the courts permitted to modify the covenant to make the restrictions narrower and to make the covenant enforceable? If so, under what circumstances will the courts allow reduction and what form of reduction will the courts permit?”
  7. [labeled 8] If the employer terminates the employment relationship, is the covenant enforceable?”[221]

The seven questions were applied to statutory and decisional provisions for each state, with each state receiving a score of zero, five, or ten in response to each question, and then an aggregate score that was a weighted sum of the individual response scores. For example, in applying Question 1,

a score of 10 was awarded to a state that has a statute that favors strong enforcement, a 5 was awarded to a state that either did not have a statute or had a statute that was neutral in its approach to enforcement and a 0 was given to a state that has a statute that disfavors enforcement. This question was given an overall weight of ten.[222]

By way of contrast, for question 3,

a score of 10 was awarded to a state that places a weak burden of proof on the plaintiff employer, a 5 was awarded to a state that has a balanced approach to the burden placed on the employer and a 0 was awarded to a state that places a strong burden of proof on the employer. This question was given an overall weight of 5.[223]

Bishara suggests that:

Ultimately, this research will present a subtle yet authoritative view of the development of noncompete enforcement and provide evidence of trends in enforcement, as well as give guidance for state policymakers, businesses, and employees when evaluating the pros and cons of negotiating and attempting to enforce a noncompete agreement.[224]

We assume that Malsberger’s survey was based on a well-informed review of relevant legal materials. At the same time, the review was not comprehensive, and its identification and characterization of relevant holdings and statutory provisions are matters of subjective—if informed—legal judgment. That is, they are not objective measures. Similarly, the twelve factors of import are matters of subjective—if informed—legal judgment.

Building on that review, Bishara applies his own rubric, which includes the scoring scheme (0, 5, or 10), scoring (or coding) of provisions under that scheme, and weighting of the seven scores to enable a weighted sum for each state, and an ordering of the states according to those sums. Given each sum, the ordering is objective, but the rubric is not: the choice of scoring scheme and—critically—the scoring and weighting of provisions and holdings under that scheme are all matters of subjective judgement or intuition.

The problem is more than simply that any index is imperfect; the limited inputs and rubric makes it difficult for other scholars to investigate and compare different legal changes. It is worth comparing Bishara’s rubric and index to an index like the World Economic Forum’s Global Competitiveness Report, for example.[225] The Global Competitiveness Report comprises 12 “pillars,” each of which aggregates multiple categories and explicit data points.[226] Any change in the index can be explicitly traced to a change in one of the data points or survey questions. Researchers can adjust weightings (but not the specified data points) as they see fit to test the robustness of the index. From there, researchers using the Global Competitiveness Report can debate whether the index is picking up appropriate policy and legal changes, so that causal estimates are properly identified using changes in the index. We discuss identification further in Section I.D, infra.

While Malsberger’s identification of pertinent legal decisions and enforceability criteria reflect considered and informed legal judgment, they track a relatively limited amount of the variation observed in state law. Moreover, they are, as we have said, matters of subjective judgment; and we can find no evidence that the criteria (or questions) were ever tested against any specific outcomes. As noted above, it does not appear to be the case that anyone has investigated, empirically, the contention that the enforceability criteria serve as an effective (or accurate) proxy for the likelihood of litigation to enforce an NCA. And to unpack the “theory” of enforceability further, we might consider the varied litigation criteria we listed at the top of this section: there seems never to have been any investigation of the empirical relationship between, e.g., the presence or absence of a state law generally (on some level of generality) and, e.g., the incidence, duration, or cost of NCA enforcement cases litigated to their conclusion. We don’t know specifically for what “enforceability” is a proxy, and we don’t know how well it serves as a proxy measure for reasonable candidates.

The same can be said of Bishara’s rubric and its implementation. Either or both might reflect considered legal judgment. They nonetheless represent subjective assessments; and again, we are unaware of any attempt at empirical assessment of the relationship between any of the individual scores, or the weighted sums of those scores, on any of the enforcement measures we listed at the top of this section. That is not to say that none of the scoring and ranking criteria signals anything of interest. It is, however, intended to underscore that the enforcement measures, (1) constitute a family of related schema, rather than any objective metrics, (2) it’s not clear what indicia of relevance that the scores are supposed to function as proxies for, (3) all entail several stages of subjective judgment, and (4) that many other approaches may be available, and perhaps preferrable, for some area of inquiry or other. We suspect that the combined econometric and legal expertise of the FTC’s staff could improve upon these metrics, if tasked to do so.

The various measurement schema employed in the enforceability studies also recall our discussion, above, of the importance of coding to the reported results. The Hawaii, Oregon, and Michigan studies all exploited legal events that were in some regards idiosyncratic and in others simply misread. Hawaii presented an idiosyncratic legal change (bearing on both NCAs and non-solicit terms, for employees of certain tech firms but not others, with no application to existing NCAs), in an idiosyncratic context (a very small state lacking a significant tech sector). The Michigan studies, as discussed, seemed to depend upon readings and coding of a legislative event in Michigan that overstate the regards in which the event effected a change in the law, and perhaps in NCA. Oregon, too, involved a legal change that, while apparently non-trivial, may have affected less legal change than it seemed at first glance; and in any case, the Oregon event coincided with the onset of the Great Recession, which might well have been a confounding factor in assessing observed effects in Oregon against those in control states.

We might also wonder about the enforceability scale employed in the Hausman and Lavetti study of physician organization and health care services prices. That study exploits not a distinct legislative event, but “rich variation in the relevant legal environments” across the states, employing an enforceability rating scheme akin to the one in Bishara.[227] The single example of such changes discussed expressly in the article concerns a judicial decision in Louisiana:

For example, in Shreveport Bossier v. Bond (2001) a Louisiana construction company attempted to enforce an NCA against a carpenter. The state Supreme Court ruled that the NCA could only prevent the carpenter from establishing a new business, but not from joining a pre-existing firm. This decision abruptly changed the law in the state, allowing all workers, including employed physicians, who had previously signed NCAs to escape the restrictions and move to other firms.”[228]

This seems to imply that NCAs were generally (or at least typically) enforceable against employees moving to another firm as employees before the decision, but not after. In one regard, that would be simply erroneous. The decision in question, SWAT 24 Shreveport Bossier v. Bond,[229] did not change the law of the entire state of Louisiana; it resolved a circuit split.[230]

The extent to which the Louisiana Supreme Court decision changed the law, as read by courts in any of the state’s circuits, is unclear. The decision was rendered against a backdrop of Louisiana’s longstanding “public policy disfavoring noncompetition agreements between employers and employees.”[231] Specifically, “[p]rior to the enactment of the first statutory prohibition of noncompetition agreements in 1934, Louisiana courts consistently held these agreements to be unenforceable.”[232] Subsequent statutory amendments continued to restrict NCAs, but provided for certain exceptions under which NCAs would be enforceable. At issue in SWAT 24 had been an exception established under a 1989 statutory amendment that did not expressly address NCAs and could be read narrowly or broadly. Under a narrow reading, a Louisiana court would not uphold an NCA if an employee left to work as an employee of another firm, but it might uphold the NCA if the employee left “to pursue his own competing business.”[233] Under a broad reading, an NCA might be found valid even if the employee left to work, as an employee, of another firm. Louisiana’s second circuit court of appeals had read the exception narrowly in the matter on appeal, and had done so in prior decisions,[234] but the state’s fourth circuit read it more broadly in 1998, as did the third circuit in 1999.[235] The state Supreme Court sustained the narrow reading.

The SWAT 24 decision describes other potentially relevant aspects of Louisiana law, but, plainly, firms suing to enforce NCAs in Louisiana were subject to significant statutory and decisional constraints prior to the circuit split. And while the third and fourth circuit decisions repudiated in SWAT 24 did provide employers considerable latitude, this much seems clear, and as close to an objective reading of the law as one can get: the SWAT 24 decision did not change authoritative reading of the law by courts in Louisiana’s second circuit.

It is possible that this represents an isolated coding error in the assessments of enforceability employed by Hausman and Lavetti. But it is an error in the sole legal example they discuss. In conjunction with the more central errors underlying, e.g., the Michigan and Hawaii event studies, it highlights a more general issue about the measurement of complex changes in statutory and decisional law, as well as their coding.

First, the legal changes being coded do not occur in a vacuum; judicial decisions as well as statutory reforms are set within a larger legal context that tends to comprise preceding statutory law (where relevant provisions may or may not be confined to a specific chapter or section of state law) and a body of jurisprudence that may include unpublished decisions as well as published ones. And given that most of the legal changes that have been in evidence are relatively subtle ones, we should worry not just about random errors in coding—about noise—but about systematic errors.

Assessing a legal change by any measure may require familiarity with the body of law in that state. Experienced attorneys in the field—especially those experienced in the state in question—might well be accurate judges of the directional impact of a pertinent new statutory provision or authoritative decision on, say, the plaintiff’s burden in seeking to enforce the terms of an NCA, likely dependent on specifying the sort of burden at issue and the terms and employment context of the NCA in question. We might assume that all of the systems would consider California’s broad statutory limit on NCA enforcement to be a strong one, and that (nearly) all coders would code it as such. There remains the question how much the 1941 enactment of the specific provision of the California Business and Professions Code we see today changed the law in California, and in what respects, given antecedent California statutes and common law restrictions on NCAs dating to the 19th Century.[236]

Quantifying the change is another matter, and for most of the statutory changes that might be observed, one that depends more heavily on identifying both the specifics of the statutory provision and some specific effect, or endpoint—some specific dependent variable—on which the change is supposed to bear, as well as the terms and employment context of the NCAs at issue. Practiced attorneys may or may not have reliable intuitions about how to score such changes. The cruder the scale, the better their chances may be, but the cruder scales may not be much help in scoring or ordering the myriad policy variations one observes in NCA law. What’s more, even with relatively crude scales, we have no evidence of the degree to which they may be reliable in one or another regard. Intuitive estimates of the relative effects of diverse changes across numerous states might be arbitrary or otherwise unreliable. And again, they have never been tested against any objective standard. The further we move from the best case—a licensed practicing attorney experienced in the employment law of a given state—the less confidence we might have in the ability of those reading, interpreting, and coding the law to estimate the magnitude of any specific change on any specific variable of interest. As we have seen, there seem to be plain and substantial errors at the level of reading and interpreting statutory and judicial reforms underlying key studies in the literature.

As we observed already, the endpoint (or dependent variable) of interest does not appear to have been specified in any of the “enforceability” studies. And the Commission’s suggestion that, while the “studies have defined enforceability of non-compete clauses in slightly different ways, each uses enforceability as a proxy for the chance that a given noncompete clause will be enforced,”[237] is itself unclear: the likelihood of what specifically, given what (if any) attributes of an NCA and employment context? In brief, we are left with a search for one or several unspecified dependent variables, without any theory of legal change to identify the quantity of interest, much less to guide how we operationalize its measurement.

None of this proves that the various implementations of the Malsberger-based enforceability rating schema do not signal anything, but we have seen that it’s not clear what they signal, or that they all signal the same thing. Again, there is no objective metric of “enforceability.” At best, we have a family of related subjective approaches to quantifying some related aspects of policy reform. At worst—and arguably—we have the results of running various labor indicators through a black box.

Under the best-case scenario, we have a developing body of economic research, some of it suggestive of reasonable concerns we might have, on average, about some of the effects of NCAs. That is not a solid ground on which to rest a sweeping federal regulation. It is, rather, an invitation for the Commission to continue to gather information on, and experience with, the competitive effects of various NCAs. And it is an invitation to the Commission to commit resources to the further development of this body of research, including improved data sources, as well as refined methods and additional findings. For example, as Starr and others have suggested, we have both an over-reliance on survey data on NCA usage and a dearth of data on who is subject (actually or on paper) to an NCA. The Commission might, for example, help refine available survey instruments—perhaps in cooperation with the Department of Labor—and it might employ its Section 6(b) authority to gather direct evidence of NCA usage, and of what terms are employed in what contexts. Moreover, more recent state-level statutory reform—especially wage-based restrictions on NCA enforcement, as in Virginia, Illinois, Massachusetts—may yield data for a series of event studies that do not require the artifice of the enforceability measure.

We might add a final question: whatever it is that the enforceability studies do or do not signal, and however well, how do the various scoring schema, and the empirical results obtained employing them, array the universe of available policy options? We don’t have any results suggesting regulatory alternatives, as it does not appear that any of the states have approached NCAs via regulation.[238] Not incidentally, we have no documented evidence of the effects of implementing a ban on use or maintenance of NCAs (as in the Proposed Rule), as opposed to limits on the abilities of plaintiff firms to enforce them, in civil court, against former employees. Beyond that, there remains the more complicated question how the various systems illustrate the differential effects of the myriad policy options a legislature, court, or regulator might consider, from diverse presumptions against (or for) plaintiff firms seeking to enforce NCAs, to “red-pencil” or “blue pencil” latitude for judges, to restrictions on one or another tranche of the income distribution, or one or another set of occupations are, or should be, rated and ordered.

The Commission has asked for input on various policy alternatives to the sweeping regulatory ban proposed in the NPRM. But it is not at all clear that the empirical evidence allows anyone to sort the optional wheat from the potential chaff. Unfortunately for the Commission, this is also an invitation for courts to strike down any Proposed Rule as insufficiently supported.

D.     The Predicted Effects of the Proposed Rule Are Flawed Because Observed State-Level Changes May Not Apply Linearly to the Proposed Rule’s National-Level Policy Change

It is not enough to simply lump some studies under the heading of “natural experiment.” A natural experiment is just a name given to a situation outside of researchers’ control that the user of the term believes allows them to identify a causal estimate. Any causal estimate is identified only with respect to a model. Relative to the model of supply and demand, regressing quantity purchased on price has an identification problem.[239] We cannot say that changes in price cause changes in quantity. But relative to a model of standard consumer theory, there is no identification problem. Variation in price does cause a change in the quantity purchased. Much of the debate surrounding modern empirical economics papers is the extent to which people accept a proposed model or identifying assumptions.

Moreover, the question of causation is not a simple yes or no. For example, the Commission quotes Starr, et al., noting that certain results are “‘best taken as descriptive and should not be interpreted causally.’”[240] But in a working paper version of the paper, the authors include an appendix on “Potential Instruments for Noncompetes,” which considers policy changes gathered as an instrumental variable for NCAs.[241] An instrumental variable potentially generates a causal estimate under an appropriate model. However, the authors conclude that the regressions “yield implausible estimates.”[242] In Section I.B, supra, we gave reasons why the identifying assumptions around event studies (such as parallel trends between different states around the Great Recession) may not hold. As the Commission sometimes observes, we should vary the weight we attribute to certain results according to the degree to which we believe the identifying assumptions.[243] Instead of lumping all under “natural experiments,” it is important to clearly delineate papers on changes of an enforceability index from papers that study an explicit policy change (Hawaii, Oregon, or Michigan). The latter comprise a much smaller body literature. Both sorts of studies may provide valuable insights, but they rely on fundamentally different identifying assumptions—as do the papers with regressions that the Commission considers not to be natural experiments, which would have a causal interpretation if the model were the regression used in each paper.

Once we are confident that what we are picking up is likely to be a true causal effect, then we can ask about external validity and how that estimate can inform a policy change beyond the scope considered in the data, such as the FTC’s Proposed Rule. While the Commission uses a “conservative” estimate of the effect on wages, it does not actually provide a robust defense of this estimate; rather, only a “back-of-the-envelope” extrapolation from a single unpublished study.[244] It is difficult to imagine that such a casual approach will satisfy courts assessing whether there is sufficient empirical support for the specific Proposed Rule.

The NPRM assumes a linear relationship between the enforceability index and the log of wages, and that the linear relationship would hold in the context of a national policy change. There are reasons to place little weight on both steps of the extrapolation since it is so far out of sample.

First, the changes picked up in the index used by Johnson, et al. may bear little resemblance to even a state level version of the Proposed Rule.[245] The impact of policy changes at the state level may be linear, supralinear, or sublinear. In other words, there could be linear returns to decreasing enforceability (as assumed), increasing returns, decreasing returns, or even, eventually, negative returns. Taking “the most conservative estimate” does nothing to mitigate this uncertainty, since the estimate technique assumes linearity.[246] Indeed, although a conservative estimate is likely better than the alternative, for any given state (or industry, wage level, type of employee, or any number of other variables) it may still be wildly inaccurate—even directionally so.

An alternative would be to use an estimate from a suitable event study, which would allow an explicit comparison between the event study’s policy change and the Proposed Rule. If the event study resembled the Proposed Rule, we need to worry about whether the relationship is linear or not, as we have the estimated effect for the relevant treatment, at least at the state level. Unfortunately, no such event study exists because no state has implemented such a stringent policy against NCAs as the Proposed Rule would be.

Second, simply extrapolating from a state policy change—even a comparable one—to a national policy change is not straightforward. It is not obvious that so-called “general equilibrium effects” operate the same at the state level as at the local or national level. For a simple example, state level estimates of the tax elasticity of capital gains are different from what we should predict from a national capital gains tax change.[247] In the context of NCAs, we know that firms set uniform policies across states. This is why workers sign NCAs in states where they are illegal; everyone in the company signs one. Similarly, any estimate of a state policy change will not pick up firm responses that occur only when the policy applies sufficiently broadly. The NPRM suggests that businesses have substitutes for NCAs, such as NDAs.[248] We explain why NDAs do not perfectly replicate NCAs in Section II, infra. Nevertheless, they may be partial substitutes. In that case, we should expect that businesses will substitute more to NDAs for a national policy change to NCAs than they do for a state policy change. In that case, simply extrapolating from the state estimate will overestimate any effects—good or bad—of the Proposed Rule.

II.      The Proposed Rule Fails to Account for NCAs’ Procompetitive Benefits and Wrongly Assumes Equivalent Benefit from Alternatives to NCAs

As the Commission observes, courts have long recognized that NCAs may “increase employers’ incentive to make productive investments, including in worker training, client attraction, or in creating or sharing trade secrets with workers.”[249] The Commission concedes that “there is evidence non-compete clauses increase worker training and capital investment (e.g., investment in physical assets, such as machines),”[250] and cites three studies indicating such effects.[251] Nevertheless, it concludes that these well-established, and empirically supported business justifications “do not alter” its conclusion that all NCAs, save those that come within its narrow exception for the sale of a business, merit condemnation as unfair methods of competition.

The Commission states two reasons for refusing to account for procompetitive uses of NCAs:

First, employers have alternatives to non-compete clauses that reasonably achieve the same purposes while burdening competition to a less significant degree. Second, the asserted benefits from these commonly cited justifications do not outweigh the considerable harm from non-compete clauses.[252]

Neither of these reasons justifies the Commission’s sweeping NCA ban. Indeed, the NPRM provides no account of the degree to which, and cost at which, such alternatives function as either alternatives or complements to NCAs.

The Commission identifies four “alternatives to non-compete clauses for protecting valuable investments”: trade secret lawsuits, non-disclosure agreements (NDAs), fixed-duration employment contracts, and enhanced wages and benefits.[253] None of those alternatives is as effective as an NCA in preserving incentives to make output-enhancing investments that could be taken to the investing employer’s rivals, however.

Moreover, any effort by employers to try to make these alternatives as effective as an NCA would also run afoul of the Commission’s’ Proposed Rule. The Commission in its Proposed Rule makes de facto alternatives to non-compete agreements illegal:

[T]he following types of contractual terms, among others, may be de facto noncompete clauses:

(i) A non-disclosure agreement between an employer and a worker that is written so broadly that it effectively precludes the worker from working in the same field after the conclusion of the worker’s employment with the employer.

(ii) A contractual term between an employer and a worker that requires the worker to pay the employer or a third-party entity for training costs if the worker’s employment terminates within a specified time period, where the required payment is not reasonably related to the costs the employer incurred for training the worker.[254]

While these inclusions may be necessary to achieve the Commission’s intended effect – to completely prohibit NCAs, including anything that functions like an NCA – they also undermine the Commission’s claim that “alternatives” remain available. And it is precisely to the extent that alternatives actually function as alternatives that they would run afoul of the Proposed Rule.

Meanwhile, to the extent these alleged alternatives might be found permissible under the “de facto” clause of the Proposed Rule only because they remain less effective, this highlights how the Proposed Rule deviates from accepted competition principles: Prevailing antitrust doctrine does not credit less restrictive alternatives that are less effective than the restraints they would replace.[255]

A.      Trade Secret Law Protects Different Intangible Assets than NCAs and Is More Difficult to Enforce

The Commission maintains that trade secret law provides a substitute means by which an employer may protect valuable information from being transferred to a rival.[256] It states (misleadingly, as explained below) that the Uniform Trade Secrets Act provides legal protection for “information that (1) derives independent economic value from not being generally known to other persons who can obtain economic value from its disclosure or use, and (2) is the subject of reasonable efforts to maintain its secrecy.”[257] The ability to sue workers who misappropriate such information, the Commission says, provides an adequate incentive for employers to produce and engage in the intra-firm sharing of competitively valuable information, negating a key business justification for NCAs. But trade secret law is less effective than NCAs at protecting employer interests for at least five reasons.

First, trade secret law provides little to no protection against the appropriation of skills training. Training an employee how to perform the tasks necessary to be a productive worker for an employer is not typically or chiefly a secret to rivals engaged in the same basic business. The benefit those rivals get from hiring the training employer’s workers is not secret information but the return on the training employer’s sunk training costs. If a firm cannot prevent the loss of such benefits before recouping its investment, it will be less likely to incur such costs in the first place. Trade secret law cannot address that problem.

Second, trade secret law fails to protect valuable competitive information besides that implicated in employee training. The NPRM misstates the definition of a trade secret in a manner that obscures key limits on the law’s protections. The first element of a trade secret defined by the Uniform Trade Secrets Act is not as stated in the NPRM, but is instead “information… that (1) derives independent economic value, actual or potential, from not being generally known to, and not being readily ascertainable by proper means by, other persons who can obtain economic value from its disclosure or use….”[258] The Commission’s characterization obscures the requirement that a protectable trade secret must “not be readily ascertainable by proper means” by rivals. This requirement has often prevented competitively valuable information like customer lists and information about customers’ interests and preferences from qualifying for trade secret protection.[259] An employer who cannot protect such information is less likely to compile it, or to share it with workers who may leave for a rival.

Third, reliance on trade secret law to protect competitively valuable information tends to limit efficient sharing of such information within the firm. Many businesses operate most effectively when numerous employees are aware of competitively valuable information, such as customer preferences, buying patterns, etc. But the second element of a protectable trade secret is that the information be “the subject of efforts that are reasonable under the circumstances to maintain its secrecy.”[260] The more freely a piece of information is shared within a firm, the less likely it is to merit trade secret protection.[261] NCAs, by contrast, do not discourage the intra-firm sharing of competitively sensitive information.

Fourth, trade secret law has much higher enforcement costs than NCAs. An employer who believes that its former employee has shared competitively valuable information with a rival must first prove that the information qualifies as a trade secret—i.e., that it (1) has independent economic value, actual or potential; (2) is not generally known to other persons who would benefit from it; (3) is not readily ascertainable by proper means; and (4) is the subject of reasonable efforts to maintain secrecy.[262] It must then show that the departing worker shared the information with a rival. In many cases, the rival could have acquired the competitively valuable information from numerous sources, and it will be difficult for the employer to prove misappropriation. As a practical matter, then, trade secret violations will be relatively difficult to detect, and relatively costly to prosecute—again, diminishing their effectiveness as a substitute for NCAs. NCAs, then, are a complement to trade secret protection, not fully covering the same scope, but enabling firms to ensure some degree of trade-secret protection at comparatively lower cost.[263]

Fifth, even if an employer succeeds in establishing liability under trade secret law, its remedy will often be inadequate, or even worthless. Once a trade secrete has been appropriated, the cat is out of the bag.[264] An aggrieved employer can seek damages, but those can be extremely difficult to prove with adequate certainty (further adding to enforcement costs above). The appropriating employee will often be judgment-proof, and third-party beneficiaries of the trade secrets will be likely unreachable. Rational employers will often forego trade secret actions even in the rare cases in which they could establish trade secret status, misappropriation, and the degree of their damages without undue cost.

B.      Non-Disclosure Agreements Are Substantially Less Effective than NCAs at Encouraging Worker Training and the Sharing of Valuable Information with Workers

The Commission observes that:

Employers that seek to protect valuable investments also have the ability to enter into NDAs [i.e., Non-Disclosure Agreements] with their workers. NDAs, which are also commonly known as confidentiality agreements, are contracts in which a party agrees not to disclose information the contract designates as confidential. NDAs may also prohibit workers from using information that it designated as confidential. If a worker violates an NDA, the worker may be liable for breach of contract.[265]

According to the Commission, the availability of NDAs obviates the need for NCAs. But NDAs—like trade secret actions, and for many of the same reasons—are substantially less effective than NCAs at encouraging worker training, client attraction, and the creation and intra-firm sharing of competitively valuable information.

Many employer investments, such as skills training, can be transferred to hiring rivals without any disclosure whatsoever. And courts do not enforce NDAs to preclude a trained employee’s subsequent use of skills funded by a prior employer. Favorable customer relationships created by employer investments also cannot be protected by NDAs. The departing employee who attracts her prior employer’s customers engendered by the initial employer’s efforts to foster favorable employee/customer interactions does not breach a non-disclosure commitment.

There are also significant practical impediments to using NDAs to protect employer investments in worker training, customer attraction and loyalty, and competitively valuable information. Simply drafting an NDA that could substitute for an NCA poses a challenge because the employer must anticipate and specify ex ante all the categories of information a departing employer might misappropriate and the ways it might do so. To establish liability, the employer must prove that the employee disclosed or illicitly used the information; again, the mere fact that another learned the information does not establish that that the employee disclosed it. If the employer surmounts this hurdle, it must establish its damages with reasonable certainty[266]—challenging when, as is typical, damages comprise lost profits.[267] And, again, departing employees may often be judgment-proof.

Instead of being substitutes, NDAs may be a complement to NCAs, as NCAs may decrease the enforcement costs of NDAs. In that case, we would expect to see NCAs bundled with other trade secret agreements,[268] which we do, especially among higher earning individuals, which we do, as the NPRM points out.[269]

C.      Fixed Duration Employment Contracts Are Subject to Remedial Limitations that Render Them Ineffective Substitutes for NCAs

The Commission states that:

[I]f an employer wants to prevent a worker from leaving right after receiving valuable training, the employer can sign the worker to an employment contract with a fixed duration. An employer can establish a term of employment long enough for the employer to recoup its training investment without restricting a worker’s ability to compete with the employer after the worker’s employment ends.[270]

The problem with relying on fixed duration employment contracts is the law of contract remedies. Given the repugnancy of involuntary servitudes and the practical difficulty an administering court would face in ensuring that any compelled service is of adequate quality, contract law does not permit specific performance as a remedy for breach of personal service contracts.[271] Hence, no court would order an employee subject to a fixed duration employment contract to abide by her commitment. The remedial options would be either a negative injunction barring the employee from engaging in competing employment[272]—effectively a judicially imposed NCA—or money damages. But here too, lost profits would be difficult to ascertain and, often, impossible to collect.

D.     The Claim that Higher Wages and Enhanced Benefits Can Substitute for NDAs Reflects a Misunderstanding of the Hold-Up Problem

Finally, the Commission maintains that NCAs are unnecessary because employers could prevent their workers from leaving for a rival by providing them with greater benefits:

Employers that wish to retain their workers can also pay the worker more, offer them better hours or working conditions, or otherwise improve the conditions of their employment. These are all viable alternatives for protecting training investments, and other investments an employer may make, that do not restrict a worker’s ability to work for a competitor of the employer or a rival’s ability to compete against the worker’s employer to attract the worker.[273]

These observations by the Commission betray a fundamental misunderstanding of the hold-up problem that justifies particular NCAs.

Employers often must undertake costly investments to enable their employees to generate as much value as possible. For example, they may provide them with costly training, share competitively valuable information with them, and grant them opportunities to build personal relationships with firm clients. Such investments are made at risk: if the employee leaves before the investing employer has received an adequate return on its investment, the cost of the investment is lost, and perhaps transferred to a rival. A higher wage may be justified for a subsequent employer, as the employee comes with the added value provided by the former employer (e.g., training, knowledge of competitively valuable information, relationships with potential customers). Employees in which employers have invested are thus well-positioned to press their employers for greater compensation. The risk of such hold-up prompts a tendency of employers to underinvest in training and information sharing. NCAs ameliorate such risks and, hence, their tendency to prompt such underinvestment.

To provide greater compensation before firm investments in employees have generated adequate returns is to compensate an employee for the firm’s investment. In effect, it endorses, rather than ameliorates, the risk of hold-up. Thus, simple deal-sweetening is not, as the Commission asserts, a “viable alternative[] for protecting training investments[] and other investments an employer may make.”[274]

III.    The Commission’s Relevant Experience, Expertise, and Capacity to Enforce Proposed Rule Is Limited

The NPRM states that the

rulemaking represents the culmination of several years of activity by the Commission related to non-compete clauses and their effects on competition. This activity has included extensive public outreach and fact-gathering related to non-compete clauses, other restrictive employment covenants that may harm competition, and competition in labor markets generally.[275]

Specifically, the NPRM cites to the record of several hearings and workshops: two hearings sessions among the Commission’s Hearings on Competition and Consumer Protection in the 21st Century “FTC 21st C. Hearings”),[276] the FTC 2020 NCA Workshop,[277] and a 2021 workshop, jointly sponsored by the FTC and the Antitrust Division, regarding the broader topic of labor market competition. (“FTC/DOJ 2021 Labor Competition Workshop”).[278] The Commission also cites three competition matters involving NCAs that were resolved by consent orders on the eve of the Commission’s announcement of the NPRM.[279] In addition, the Commission notes a 2019 petition for a rulemaking from the Open Markets institute and various co-signatories.[280]

While we do not doubt that the staff conducted appropriate investigations in each of the three matters – or in a fourth settled in March – we note that the consents were achieved without either trial or adjudication by the Commission, and without any finding or stipulation of any antitrust violations.[281] Moreover, as noted by Commissioner Wilson in her dissenting statement to two of the three orders, the Consent Orders are exceedingly brief, providing little guidance as to how the conduct at issue violated—in the Commission’s view—either the FTC Act or the Sherman Act.[282] A fourth matter has been settled since, but that provides little further guidance. And, like two of the three initial matters, it involves facts and circumstances specific to the glass container industry.[283]

In Prudential Security, et al., security guards allegedly were subject to NCAs that barred the guards from undertaking related employment with any of Prudential’s competitors, and from starting a competing business. Those prohibitions were alleged to apply for two years following conclusion of the guards’ work for the Respondent, anywhere within a 100-mile radius of their main place of work for the Respondent. The NCAs also were alleged to impose liquidated damages of $100,000 per guard, per violation.[284] Such terms seem extreme, given the occupation: they might well be untethered from, e.g., any of the firm’s interests in protecting proprietary information or the firm’s employee-specific investments;[285] and they might well be inefficient, “unreasonable” (as found by a Michigan state court, applying Michigan law),[286] or otherwise objectionable.

Based on the available documents, however, it remains unclear whether competition in specific Michigan labor markets was harmed by the conduct at issue. The Commission’s complaint alleged harm to “competitive conditions,” and to individual security guards in some relevant labor market or markets. [287] The supporting documents also allege that the NCAs were “coercive and exploitative.”[288] The Commission opines that such conduct, causing “harm to competitive conditions,” constitutes a violation of Section 5’s prohibition of unfair methods of competition. It cites, as authority, its recent Policy Statement Regarding Unfair Methods of Competition Under Section 5 of the FTC Act.[289] Indeed, such language is to be found in the policy statement, and terms such as “exploitative” and “coercive” occur in dicta in certain Supreme Court decisions. But, as we have noted elsewhere, while those terms are evocative in colloquial usage, they have no established meanings in antitrust jurisprudence; and their meaning, application, and connection to antitrust jurisprudence is not explained in the Commission’s policy statement.[290] Critically, novel applications of the Commissions 2022 policy statement have not yet been vindicated in the courts.

It is well established that NCAs can vary along multiple dimensions: duration, geographic scope, occupational scope, application to certain types of firms, and stipulated damages, among others.[291] Even supposing, arguendo, that the conduct at issue in Prudential, et al.—like the conduct at issue in Ardagh Group, et al., and O-I Glass—violated Section 5’s prohibition of “unfair methods of competition,” the consent orders seem to turn on specific facts and circumstances, such as the duration of the restrictions and the outsize liquidated damages provisions in Prudential, et al. It remains entirely unclear how well information uncovered in the staff’s investigations might inform competition analyses of NCAs with different terms, or in other labor markets. As Howard Shelanski—a former Director of the FTC’s Bureau of Economics and a former Administrator of the White House Office of Information and Regulatory Affairs during the Obama Administration—noted at the FTC’s 2020 workshop, enforcement is a “slow” way to gather the information requisite to the issuance of useful guidance or regulations, in part due to selection bias.[292] That is not to gainsay the importance of developing the case law. Rather, it is to underscore the need to develop a body of case law that reflects the diversity of NCAs, the contexts in which they are employed, and their effects. Three or four settled fact-specific investigations, in toto, seem a very slender reed on which to hang a major federal regulation on labor agreements generally. That seems all the more significant, given the history of NCA litigation, both in competition matters and at common law, where the specific terms and conditions, and the context in which they are employed, have tended to determine whether or not specific NCAs were found enforceable or lawful.[293]

With respect to the hearings and workshops mentioned in the NPRM,[294] while not focused exclusively on NCAs, these were significant information-gathering efforts on competition issues in labor markets by FTC staff. The Commission’s call for comments, issued in conjunction with the 2020 workshop, solicited responses to various questions, both descriptive and normative, including several on the adequacy of existing NCA laws and regulations; the FTC also asked for input on possible legal reforms. They also addressed the Commission’s practical ability and legal authority to advance policy reforms by regulation. The NPRM notes that 328 comments were submitted to the record of the 2020 workshop, and that 27 comments were submitted to the record of the FTC/DOJ 2021 Labor Competition Workshop.[295] In addition, 280 comments were submitted in response to a 2021 call for “public comments on contract terms that may harm competition, including ‘non-compete clauses that prevent workers from seeking employment with other firms.’”[296] All of these, according to the NPRM, informed the rulemaking process:

As it has developed this Proposed Rule, the Commission has closely considered the views expressed at these forums and the public comments it has received through these engagement efforts. The comments have informed the Commission’s understanding of the evidence regarding the effects of non- compete clauses; the law currently governing non-compete clauses; and the options for how the Commission may seek to restrict the unfair use of non- compete clauses through rulemaking, among other topics.[297]

That may be true, but, as we noted in the Introduction and Executive Summary of these comments, the Commission has never issued a report summarizing or synthesizing the information gleaned from these various endeavors. What is more, references to the evidence gathered through those substantial investigations seem extremely limited and highly selective. The diversity of views and evidence presented at the hearings and workshops, and in submissions to the records of those events, is not in evidence in the NPRM or, specifically, the Proposed Rule. Some well-documented complications get slight treatment in the NPRM, while others are simply absent from the discussion. While workshop presentations and submissions cannot settle the complex questions presented by NCAs and the Proposed Rule, they do offer substantial input on matters ranging from stakeholder views to legal challenges, to the developing state of the empirical evidence. A proper analysis of the record, rather than summary reference to it, is wanted.

Even a brief survey of the record indicates the complexity of stakeholder viewpoints, policy issues, and evidence. At the FTC’s 2020 workshop, panelists and commenters expressed diverse views on the most basic questions whether the FTC could or should regulate NCAs. Several workshop participants and numerous comments endorsed some measure of federal intervention to restrict the use or enforcement of NCAs, at least in some contexts.[298] Some of those comments endorsed federal intervention as a complement to state NCA law. For example, comments submitted jointly by twenty state Attorneys General advocated for “federal rulemaking that is consistent with [the states] ability to pursue enforcement and legislative priorities to the benefit of workers and consumers,” while also noting advantages to “the type of experimentation and variation that our system of government is designed to promote,”[299] with the states serving as Brandeisian “laboratories of democracy.”[300] At the same time, they recommended that federal rules should not preempt state law.[301]

Other comments took a dimmer view of federal intervention, while also lauding state law variation: “[s]tate laws are sufficient to address any harms that may be associated with noncompete agreements. Federal intervention (whether at the statutory or regulatory level) is not necessary.”[302] And the Global Antitrust Institute commented that they were “concerned… that many proposals to address… [concerns about NCAs] through ex ante antitrust regulatory interventions, such as an FTC rule, are ill-suited and will likely do more harm than good.”[303]

Comments from the Antitrust Law Section of the ABA questioned the need for federal intervention.[304] They also questioned whether the available evidence provided an adequate foundation for policy reform:

The Section does not have the impression that the research and analysis of non-compete clauses are far enough along such that lawmakers and policymakers—whether at the federal, state, or local level—have a clear sense of the nature and extent of the harms, an ability to evaluate the adequacy of existing legislative and regulatory regimes to address those harms, and a blueprint for additional legislation or regulation should current regimes be deemed inadequate.[305]

Howard Shelanski stated plainly that an outright ban would be “deeply problematic.”[306]

Commissioner Noah Phillips’s noted the Commission’s extremely limited experience with competition rulemaking in general: “The FTC has issued a competition rule just once in its history, in the 1960s.”[307] And several workshop participants also focused on the demands of rulemaking for the FTC or, more specifically, the demands that would attend any rulemaking likely to survive court challenges. Commissioner Phillips argued that the broad language of Section 5 might raise Constitutional concerns, including those associated with the Nondelegation Doctrine; and, further, that “[n]ondelegation concerns may also be exacerbated by other factors here, including the lack of clarity in the rulemaking authority, the traditional commitment of the issue to the states, the fact that neither the FTC nor any court has found non-competes to violate the FTC Act’s prohibition against unfair methods of competition.”[308]

Such concerns should be all-the-more salient, given the Supreme Court’s recent articulation of the Major Questions Doctrine in West Virginia v. EPA.[309] Reviewing certain power plant emissions standards adopted by the Environmental Protection Agency, the Court observed that…

our precedent teaches that there are “extraordinary cases” that call for a different approach—cases in which the “history and the breadth of the authority that [the agency] has asserted,” and the “economic and political significance” of that assertion, provide a “reason to hesitate before concluding that Congress” meant to confer such authority.[310]

Separation of powers principles, in addition to readings of legislative intent, required “something more than a plausible textual basis for the agency action”; for that reason, given the scope and economic impact of the EPA’s regulation, the Court held that EPA had exceeded its statutory authority.

In the NPRM, the Commission contemplates a regulation that would, by the Commission’s own estimation, alter the terms of employment for approximately 30 million American workers, with an economic impact of “$250 to $296 billion per year,” on wages alone. That is, the Commission asserts that its Proposed Rule would be one of significant economic impact, just as public controversy over NCAs and the proposed preemption of state law suggest significant political impact. Independent of proposed rules and advanced notices of proposed rulemaking, the Commission reports that it has some 18 guides and regulations under review.[311] Most of these were adopted under express statutory authority considerably narrower than the charge of Section 5. The FTC Act comprises no such express grant of authority with respect to NCAs or other terms of labor agreements.

We cannot be certain how the courts might evaluate an FTC NCA regulation, but it’s clear enough that the federal courts have increasing concerns about agency deference. With that in mind, adoption of NCA regulations as proposed would seem to pose a substantial risk to the Commission; that is, to both the substance of such regulations and the Commission’s regulatory authority in competition matters, at least. A thoroughgoing analysis of the scope of the Proposed Rule and the present (and developing) state of agency deference and statutory interpretation in Supreme Court jurisprudence seems sorely needed. Former FTC Chairman William Kovacic took a less settled position on FTC authority, while also observing that the contemporary judiciary is skeptical of agencies’ initiatives to extend their own reach.[312] While Kovacic did not suggest that the courts find all regulatory innovation anathema, he did emphasize the importance of building a comprehensive foundation for any forays into competition rulemaking likely to survive judicial scrutiny.[313] Howard Shelanski similarly advocated for the further development of the empirical evidence before entertaining rulemaking;[314] he also suggested that there was much work—such as the issuance of guidance and the development of research—that the Commission might undertake on NCAs besides, or prior to, regulation.[315]

As discussed in more detail in Section I of these comments, supra, various panelists at the FTC 2020 NCA Workshop—including leading contributors to the empirical literature cited in the NPRM—noted significant limitations to the state of the literature, as did comments submitted to the record. For example, Evan Starr noted the difficulty of estimating the causal effects of using NCAs,[316] and the need for more research on those causal effects.[317] Challenges to such research include, inter alia, the fact that NCAs are commonly bundled with other restrictions, such as non-solicitation and non-disclosure terms (and attendant selection issues),[318] an over-reliance on survey data,[319] a dearth of longitudinal data,[320] a relative dearth of exogenous variation,[321] and a dearth of findings regarding total welfare implications of NCAs and, specifically, of research on the downstream effects of NCAs on product and service markets, and thereby on consumers.[322] Panelists also suggested mixed results, rather than uniform findings on, e.g., pay, and on potential tradeoffs in labor markets, such as tradeoffs between wages and firms’ investments in employee training.[323]

Panelists at the FTC 2020 NCA Workshop also noted mixed results, rather than uniform findings on, e.g., pay, as well as potential tradeoffs in labor markets, such as tradeoffs between wages and firms’ investments in employee training.[324] And several panelists noted both observed and potential benefits to NCAs.

For example, Ryan Williams presented research on the effects of NCAs for CEOs suggesting that NCAs provide compensation benefits for the CEOs themselves, and that firms are more likely to fire a CEO for poor performance when there is a NCA in force, which potentially benefits both shareholders and employees.[325] Overall, he said that the findings imply a positive story for CEO NCAs. Similarly, Kurt Lavetti reviewed research suggesting both physician benefits and efficiencies associated physician NCAs.[326] And Commissioner Noah Phillips noted the potential of NCAs to ameliorate hold-up problems in labor agreements by, for example, encouraging investment in worker training and the sharing of proprietary information with employees.[327] In sum, as Kurt Lavetti concluded, “we’re still far from reaching a scientific standard of concluding that non-compete agreements are bad for overall welfare.”[328]

One more issue seems notable. The NPRM omits any reference to a 2019 literature review conducted by staff in the FTC’s Bureau of Economics.[329] That literature review was much discussed in comments submitted to the 2020 workshop and in the workshop itself.[330] Not incidentally, the named staff author of the review, John McAdams, moderated a session at the 2020 workshop. Yet the McAdams paper is not even mentioned in the NPRM. McAdams observes that economic research regarding NCAs “has made important strides.”[331] At the same time, however, he observes mixed results, and he describes numerous data and methodological limitations running throughout the body of literature. Overall, he finds that the “more credible empirical studies tend to be narrow in scope, focusing on a limited number of specific occupations… or potentially idiosyncratic policy changes with uncertain and hard-to-quantify generalizability.”[332] Of direct relevance to the Proposed Rule, “[t]there is little evidence on the likely effects of broad prohibitions of non-compete agreements.”[333]

That, too, is part of the Commission’s expertise and experience regarding NCAs. But it is a part that suggests caution, and grounds for research development, rather than a rush to adopt a sweeping uniform regulation like the Proposed Rule. The NPRM’s review of the literature is substantial but skewed; and, as we discuss below, the NPRM fails to adequately address many of the well-known limitations to available studies. The Proposed Rule carries real risk to the Commission’s authority, as well as its resources. As Kovacic said at the 2020 workshop, “the bolder the measure, the stronger the evidentiary armor is going to have to be and the more thoughtful the analyses,” if an intervention, and the Commission’s authority, are to be sustained.[334] As a general matter of policy, we cannot recommend adoption of so sweeping a rule as the one that the Commission has proposed. That fundamental policy issue aside, it should be conspicuous that a more fulsome development of the record, and a more critical review of the literature, is needed before the FTC proposes any regulation of NCAs.

The imposition of a sweeping federal regulation and the preemption of state law suppose general and durable market failure causing substantial consumer harm. Observation of certain market imperfections, or frictions, falls well short of that mark.[335] Recent empirical findings suggesting potential harms and benefits associated with NCAs in different, and specific, contexts also fall short. The case for regulation also supposes that regulatory intervention can be effective and efficient, yet there is no model in state law for the ban proposed by the FTC, and the NPRM provides no analysis of the likely effects of the difference between the Commission’s proposal and state law alternatives. No state has adopted the general prohibition on NCA usage that the FTC has proposed. No state chiefly restricts NCAs via a regulatory ban; and no state has adopted the seemingly arbitrary 25% share restriction that the Commission has proposed for permitting certain NCAs in conjunction with the sale of a business.[336] And while the NPRM includes a casual attempt at a cost-benefit analysis, it lacks even a cursory analysis of the resources that would be required for effective implementation and enforcement of the Proposed Rule. These would not be trivial. As noted in the NPRM, there is survey evidence suggesting that NCAs now apply to roughly one fifth of all employed persons in the U.S. labor force; that is, nearly 30 million workers.[337] Regulations are not self-enforcing: and while regulation may be, in certain regards, more streamlined than case-by-case law enforcement, it still requires investigation of alleged infractions, administration, and, in addition to regulatory challenge mechanisms, compliance monitoring, guidance to industry and workers, periodic rule review, and the resources to defend at least some challenges to agency determinations of violations, and assessments of penalties, in federal court. Detection alone may often be a challenge to the extent that many “workers are totally uninformed about the law.”[338]

Effective enforcement need not entail detecting, much less penalizing, every violation, but it does require sufficient enforcement activity to establish a credible threat that violations will be penalized, and that enforcement is not selective. Yet the NPRM contains no assessment of the resources required for adequate enforcement of the Proposed Rule or any alternative NCA regulation. Enforcement staff in the Commission’s Bureau of Competition (“BC”) have substantial antitrust expertise in mergers and diverse conduct matters, but little experience in labor matters and none in the enforcement of competition regulations. Moreover, the Commission has recently reported that BC staff are barely able to meet the Commission’s already established and important workload.[339] Adding an obligation to monitor restrictions in labor agreements across all industries and occupations in the U.S. would be both futile and an unnecessary drain on the staff’s ability to scrutinize mergers and conduct under settled antitrust law.

Enforcement burdens would be greater still, given the Commission’s proposal “that whether a contractual term is a non-compete clause for purposes of the Rule would depend on a functional test,”[340] rather than a nominal one. Currently, NCAs may be confined to distinct and readily parsed provisions among terms of employment, or they may be drafted in more complex terms, and perhaps distributed across multiple provisions or documents. A general bar on NCA use, subject to substantial regulatory penalties, would encourage firms that value NCAs to seek marginally permissible alternatives and various workarounds; these might tax staff resources further still, from detection and investigation through challenges in either administrative process or federal court.

The Commission’s experience with enforcing its own Contact Lens Rule (CLR),[341] which implements a specific statutory charge in the Fairness to Contact Lens Consumers Act (FCLA),[342] may be instructive. The CLR was adopted following “decades of regulatory and research experience regarding the optical goods industry.”[343] That experience included adoption and enforcement of the Eyeglass Rule[344] (adopted in its initial form in 1978), and two substantial studies of competition and consumer protection issues regarding regulation and retail sales of contact lenses specifically, with the latter report conducted pursuant to an express statutory charge in the FCLA.[345]

The key provision in both the FCLA and the CLR was a simple “prescription release” requirement: “[w]hen a prescriber completes a contact lens fitting, the prescriber . . . shall provide to the patient a copy of the contact lens prescription.”[346] Periodic rule review led the Commission to solicit comments on the CLR in September 2015;[347] review of those comments, and other input, led to an NPRM proposing amendments to the CLR in 2016,[348] a supplemental NPRM in 2019,[349] and publication of amendments to the CLR in 2020.[350] As the Commission explained in amending the rule, there was a “need to improve compliance with the Rule’s automatic prescription-release requirement, as well as a need to create a mechanism for monitoring and enforcing the Rule.”[351] In plain language, the Commission found that its own rule was difficult to enforce, and that non-compliance was widespread. To quantify the enforcement challenge might be difficult, but one number seems salient: we are aware of precisely zero matters in which the Commission enforced the CLR’s prescription release requirement between its initial 2004 effective date and its 2020 amendment.

We do not mean to gainsay the challenge of enforcing the CLR. To the contrary, we believe that the Commission’s experience with the CLR illustrates the challenges of drafting, and enforcing, effective regulations, even when an agency has decades of experience with the issues those regulations are meant to address.

IV.    The Commission’s Legal Authority to Issue the Proposed Rule is Contentious—and Dubious

The NPRM implicates a range of questions regarding the Commission’s legal authority. These questions relate both to the scope of the Commission’s substantive legal authority to regulate NCAs and to its authority to undertake such regulation through the adoption of a substantive rule, rather than through adjudication.[352] These issues are made all the more complicated given the infrequency with which the Commission has attempted to undertake competition rulemaking in implementation of its Unfair Methods of Competition (“UMC”) authority under Section 5 of the FTC Act—arguably, just once in over 100 years of FTC UMC authority. Meanwhile, there is very little judicial authority discussing the Commission’s competition rulemaking authority, and none of it is recent.[353] And recent judicial trends exacerbate the issue, as the courts have been increasingly skeptical of claims of regulatory authority such as the Commission makes in the NPRM.

This is a contentious area of law and policy. Several of the key issues are discussed below; other comments submitted to this proceeding develop these arguments in more detail.[354] Our primary purpose here is to emphasize that the Commission’s Proposed Rule, if adopted, would regulate into market uncertainty and legal controversy.

The Commission is the nation’s chief inter-sectoral regulator of domestic trade and commercial activity. With its statutory mandate to prevent unfair methods of competition comes a corollary mission to promote a robust and competitive marketplace. Uncertainty is anathema to such a marketplace. The Commission’s Proposed Rule would upset dozens of state laws. Not incidentally, NCAs already are a topic of extensive legislative discussion at the federal level.[355] If adopted, these rules will be subject to years of litigation. One of the few things that can be said with certainty is that media and other coverage would lead to substantial confusion and disruption for employees and employers alike.

In the comments below, we discuss the following issues: whether the Commission has statutory authority to adopt substantive Unfair Methods of Competition rules (under current D.C. Circuit precedent, yes; but that precedent is unlikely to withstand judicial review today); whether the Proposed Rule presents major questions for the purposes of the Major Questions Doctrine (it does); whether it would withstand judicial scrutiny under the Major Questions Doctrine (it likely would not); and whether the Proposed Rule, if adopted, would be based upon an unconstitutional delegation of authority to the Commission (they likely would be).

A.      The Commission’s Claimed Authority to Adopt Competition Rules Is Unlikely to Withstand Judicial Scrutiny

The Commission’s claim of general competition rulemaking authority under Section 6(g) of the FTC Act rests on an ambiguous statutory clause and a 1973 opinion of the D.C. Court of Appeals in National Petroleum Refiners Association v. FTC.[356] That opinion has not been affirmatively repudiated by the Court of Appeals or reversed by the Supreme Court, but there has been little occasion to revisit it: The Commission has not proposed or enforced competition rules since the 1970s. As the Commission is well aware, the National Petroleum Refiners Court considered an octane labeling rule that operates chiefly as a consumer protection regulation, although one deemed at the time to have both competition and consumer protection elements. And the case was decided before Congress enacted the Magnuson-Moss Warranty Act,[357] which amended the FTC Act to include substantial procedural constraints on the consumer protection rulemaking that had constituted nearly the whole of the FTC’s regulatory activity.[358] What is more, the 1973 opinion reflects a degree of agency deference that is increasingly out of favor with the federal courts.[359]

Other comments will argue the best reading of Section 6(g) in more detail than we undertake here. Our purpose here is more limited. We remind the Commission that its reading of its own authority is contentious.[360] Administrative law scholars have argued that a far more limited reading of 6(g) is likely to prevail in the courts.[361] And, in any case, the Commission must recognize that the promulgation of a broad regulatory prohibition of NCAs under the Commission’s UMC authority, first, is nearly certain to be challenged in the courts and, second, risks both the substantive provisions of such a rule and a Supreme Court repudiation of the Commission’s authority to issue substantive or “legislative” competition rules more generally.

Section 6(g) states: that “the Commission shall also have power… from time to time to classify corporations and… to make rules and regulations for the purpose of carrying out the provisions of this Act.”[362] For the proponents of a broad rulemaking power, this is taken to be a catch-all provision providing a general power to issue “rules and regulations,” subject only to the relatively light-touch procedural requirements for “informal rulemaking” in Section 553 of the Administrative Procedure Act.[363]

As prominent commentators have noted, there is no “plain meaning” of “rules” within the meaning of section 6(g).[364] In National Petroleum Refiners, the D.C. Circuit opted to “favor an interpretation which would render the statutory design effective in terms of the policies behind its enactment and to avoid an interpretation which would make such policies more difficult of fulfillment, particularly where, as here, that interpretation is consistent with the plain meaning of the statute.”[365]

A contemporary court, reading the same statutory language, would not likely agree that the meaning of “rules” in section 6(g) is “plain,” based on suppositions about the general policy behind the initial enactment of the FTC Act.

Most importantly, we note that the remainder of Section 6 empowers the Commission to investigate and report on the business practices of corporations. More recent amendments have to do with investigating, reporting, consulting, and advising by the Commission.[366] No part of Section 6 expressly authorizes the Commission to undertake any enforcement action or impose any penalties, and the authority it does explicitly grant is limited to information gathering and analysis by the Commission.

Recent judicial trends are far less deferential to administrative agencies, and far more likely to curtail agency discretion in the face of statutory ambiguity. For example, in AMG Capital Management, the Supreme Court narrowly interpreted the Commission’s power to obtain equitable remedies, repudiating established Commission practice.[367] And, as explained below, in cases like West Virginia v. EPA, the Supreme Court has demonstrated concern with the general breadth of the administrative state and, specifically, has rejected the proposition that courts defer to agency interpretations of vague grants of statutory authority where such interpretations are of major economic and political import.[368]

B.      The Proposed Rule Presents Major Questions that Can Be Addressed Only by Congress

Adoption of a broad NCA rule such as proposed in the NPRM will likely also face scrutiny under the major questions doctrine. The Supreme Court’s recent opinion in West Virginia v. EPA[369] has brough substantial attention to the Major Questions Doctrine. While the contours of this doctrine are still being defined by the courts, it stands roughly for the proposition that Congress must “speak clearly if it wishes to assign to an agency decisions of vast economic and political significance.”[370] The Proposed Rule is broad—on the Commission’s own account it would affect around 30 million employees and hundreds of billions of dollars in commerce annually. It would also insert the Commission into an area that is already heavily regulated by the states and the federal government: Numerous federal statutes and rules regulate employer/employee relations, and a vast—and active—body of state statutory and judge-made law addresses NCAs specifically. If adopted, the Proposed Rule would clearly be one of vast economic and political significance. Indeed, one could well call the Proposed Rule the very model of a modern major question.

If deemed to be a major question, it beggars belief to think that the courts would find that Congress, through the FTC Act’s capacious but general language, has spoken clearly enough to grant the Commission the authority to regulate labor in this way. Both the substance of the Proposed Rule and the mechanism of issuing such rules are likely to be found infirm. The Commission’s relevant authority is to proscribe unfair methods of competition. The scope of that authority has long been understood as largely coextensive with, but slightly broader than, the scope of the antitrust laws. Historically, the Commission’s UMC authority has been exercised through case-by-case enforcement actions.

The Commission’s Proposed Rule would go far beyond the established scope of the FTC’s UMC authority, and it would abandon case-by-case enforcement entirely. Indeed, there is no question that traditional indicia of anticompetitive conduct are of no relevance to the Proposed Rule. For example, the proposed prohibition is not limited to firms with market power. What’s more, there is no legitimate argument that NCAs are categorically anticompetitive (or otherwise unfair methods of competition). Many rules are somewhat overinclusive—that goes hand in hand with the legislative prerogative—but the Commission’s claim to rulemaking authority is strained at best, and its substantive legal authority is limited on its face to enforcing the prohibition of unfair methods of competition, not to regulating competition to ensure that broad categories of commercial practices are on net competitive.

Moreover, there is no paucity of legislative interest or ability to regulate in this area. Both Congress and the states are very active in the areas that the Commission’s rules would regulate.[371]

Other comments in this proceeding take up the arguments that the Proposed Rule presents major questions and would likely be rejected under the Major Questions Doctrine. We add to those comments, both to join in those concerns and to add a broader institutional perspective. The Commission’s recent moves towards aggressive use of its Unfair Methods of Competition authority run in the opposite direction of contemporary administrative law. The Commission’s recent policy statement on its use of its UMC authority, for instance, cites to myriad cases that are four or more decades old, antedating the modern era of antitrust law, and often rely on dicta in doing so.[372] At the same time, in cases like West Virginia v. EPA, the Supreme Court has shown concern with the general breadth of the administrative state, and in cases like AMG Capital,[373] the Court has taken action to limit, or has shown concern about, the scope of the Commission’s authority specifically. And just last week, in Axon Enterprise,[374] the Supreme Court held that defendants in FTC and SEC administrative proceedings need not exhaust agency process on the merits before raising constitutional challenges to the agencies’ actions in federal district court.

The cost of risky and resource-draining litigation cannot be gainsaid. Importantly, this observation is endogenous to the question of the Commission’s authority: The Commission is the nation’s chief inter-sectoral commercial regulator. The Proposed Rule promises to be exceptionally disruptive to the entire American economy—a destabilizing force that runs counter to the Commission’s purpose and that should, in any case, be an important consideration, even when the Commission exercises a clear statutory mandate. But here, the only certainty is uncertainty. The Commission is considering regulations that would subject vast swaths of the United States’ economy, employees, and employers to confusion and uncertainty. The Commission ought to be more circumspect about the potential to disrupt the process it is charged to protect.

None of this is to reject the Commission’s authority to challenge a specific firm’s specific use of NCAs under Section 5 of the FTC Act. Through case-by-case basis adjudication, the Commission might determine that a specific course of conduct, in a specific factual setting, makes out either a UMC or UDAP claim; that is wholly consistent with the purposes and language of the FTC Act.

C.      A Grant of Substantive Statutory Authority Sufficient to Support the Proposed Rule Would Amount to an Impermissible Delegation of Authority

While the Court’s application of the Major Question Doctrine in West Virginia v. EPA is grounded in several established strands of constitutional jurisprudence, the precise meaning of the doctrine remains uncertain. One line of inquiry suggests the doctrine is a new instantiation of the Non-Delegation Doctrine. There is some sense to that. The Major Questions Doctrine requires that Congress must clearly—and with some specificity—indicate an agency’s authority to engage in significant rulemaking. The non-delegation doctrine, meanwhile, requires that Congress provide an intelligible principle that limits the scope of congressional authority delegated to an agency.[375] But the Non-Delegation Doctrine stands on its own: it could be the case that major questions present delegation issues, but there remain potential non-delegation issues separate from major questions. And the Court has also noted, e.g., separation of powers concerns at play in major questions.

The Non-Delegation Doctrine was famously articulated in Schechter Poultry, a 1935 Supreme Court opinion striking down the National Industrial Recovery Act (NIRA).[376] This is a seminal case in the administrative-law canon: decided on the same day as Humphrey’s Executor, it dealt with the permissibility of Congressional delegations of authority to federal agencies. The central issue is that the United States Constitution vests “all legislative powers” in Congress.[377] Federal agencies are empowered to act on Congress’s behalf, which seemingly could violate the Constitution’s legislative vesting clause, which would render all agencies unconstitutional.

To resolve this issue, the Court found that Congress can empower agencies to exercise specific powers on Congress’s behalf, but that there must be limits to these delegations of authority. The constitutional limit is that “Congress is not permitted to abdicate or to transfer to others the essential legislative functions with which it is thus vested.”[378]

Schechter Poultry is all-the-more relevant to the Proposed Rule because it discusses that doctrine in direct comparison to the Commission’s statutory authority.[379] Both NIRA, which required the National Recovery Agency (NRA) to enforce codes of “fair competition,” and the Federal Trade Commission Act, which prohibits “unfair methods of competition,” have similar and similarly broad grants of statutory authority. In striking down NIRA, the Court explained its flaws in direct comparison to the FTC’s statutory authority to deem certain methods of competition unfair. It explained that…

“unfair methods of competition” are thus to be determined in particular instances, upon evidence, in the light of particular competitive conditions and of what is found to be a specific and substantial public interest.… To make this possible, Congress set up a special procedure. A Commission, a quasi-judicial body, was created. Provision was made [for] formal complaint, for notice and hearing, for appropriate findings of fact supported by adequate evidence, and for judicial review to give assurance that the action of the Commission is taken within its statutory authority.[380]

While the Court does not expressly say that it is the case-by-case, adjudicatory nature of the Commission’s UMC authority that renders the FTC Act a constitutional delegation of authority, the Court did point to the lack of these specific quasi-judicial procedures in holding NIRA’s delegation of authority to the NRA to be unconstitutional.

In other words, if the FTC were successfully to assert that the FTC Act authorizes it to enact broad competition rulemakings like the Proposed Rule, that holding may contain the seeds of its own demise, if the Court determines that such a broad grant of authority without the constraints of adjudicatory process or special Mag-Moss-like procedural rules is contrary to the Non-Delegation Doctrine.

V.      Conclusion

As we said in the introduction to these comments, we cannot recommend that the Commission adopt the Proposed Rule. It is not supported by the evidence, empirical or otherwise; neither is it supported the Commission’s experience, authority, or resources.

Our comments have, like the Commission’s own NPRM, reviewed the empirical literature regarding NCAs in some detail. In doing so, we can conclude only that the Commission’s conclusions about “the weight of the evidence” are untenable.

First, as made amply clear at the FTC 2020 NCA Workshop, evidence about the effects of NCAs themselves is both limited and mixed. And like the more substantial body of evidence on the putative effects of NCA “enforceability,” it is hardly comprehensive. Moreover, as made clear in the literature, and at the FTC’s various workshops and hearings regarding NCAs and other labor competition issues, significant data and methodological limitations are observed throughout the relevant empirical literature. These are endemic and far from trivial. While the NPRM’s review of the literature responsibly notes many of these limitations in discussing individual studies, the Commission seems wholly to ignore such limitations in making its general observations about the available empirical findings.

Second, most of the studies that have employed causal designs depend heavily on a dubious set of “enforceability” metrics. These lack any clearly specified subject; they are variable in their implementation; they depend upon several layers of subjective assessments; and they are highly coding dependent. Each implementation might best be considered a “black box.” There is no such thing as an objective measure of enforceability.

Finally, most of the studies cited by the Commission have limited relevance to antitrust enforcement. The Commission seems to be in no position to offer even a partial equilibrium analysis of NCA effects. To ignore the question of downstream effects on consumers (and the paucity of evidence in this area) would be irresponsible. What is more, findings on, e.g., average wage effects observed in a particular state tell us little about the question of substitution effects, or about the basic question of the extent to which such average effects—even if taken at face value—may be driven by specific local labor markets in which specific employers exploit a significant degree of monopsony power. As Howard Shelanski observed, “[i]t’s very possible that a small employer that ties up six employees in a non-compete has zero effect on the market.”[381] At the same time, imposition of an NCA without notice could be a material omission, and potentially actionable under the Commission’s UDAP authority.[382]

None of this is to say that the literature is without merit, or that none of the cited studies are suggestive of legitimate policy concerns. It is to say that the existing body of literature is developing and substantially incomplete. Available findings are mixed, there are far too many unanswered questions, and most empirical observations are far too uncertain in their findings and in their generalizability to ground a sweeping federal rule.

But beyond the Proposed Rule’s evidentiary infirmities lie still more problems. The Commission plainly lacks both the experience required to ground such a rule and the resources that would be necessary to enforce it. Moreover, as the Commission is aware, adoption of the Proposed Rule would be nearly certain to prompt legal challenges to both the substance of the NCA regulations and, more broadly, to the Commission’s authority to issue substantive or “legislative” competition rules under Section 6(g) of the FTC Act. While the Commission may be persuaded it has been granted such authority, and might cite a fifty-year-old D.C. Circuit case in support of that proposition, the Commission cannot gainsay changes in judicial construction that have occurred since National Petroleum Refiners. More specifically, the Commission cannot ignore the Supreme Court’s more recent holdings on non-delegation and major questions that are wholly at odds with the sort of agency deference that obtained in 1973. That is, the Commission cannot ignore either the litigation burden or the risk to its own authority—nor the legal and economic uncertainty—that the adoption of the Proposed Rule would entail.

All is not lost. As we have also discussed, the Commission is in a position to develop better data sources, and the staff are capable of making substantial contributions to the literature. These could include, among other things, development of directly observed data on NCA terms and usage that would reduce, if not obviate, an excessive reliance on survey data. In addition, recent developments in state law—specifically, income-based restrictions on NCA enforcement—should enable data collection and event studies that do not depend upon soft and untested enforceability metrics.

The Commission, Congress, and state policy makers could all benefit from a more extensive development of the Commission’s experience with NCAs. We note that the Commission has not issued any report of the findings of its 21st C. Competition Hearings, and that it has not issued any report on its 2020 workshop. We recommend that the Commission undertake a careful review of the records of pertinent FTC hearings and workshops, and that it issue a substantial report of its findings as prologue to any consideration of federal NCA regulations. Importantly, such reports could inform policy reforms that do not rest on antitrust. Recent state-level statutes, such as income-based limits on NCA enforcement, are not merely opportunities for event studies. Rather, they highlight the various policy concerns that might motivate state or federal reforms in labor policy, whether in conjunction with, or apart from, any observations of conduct that exploits market power in violation of the FTC Act or the federal antitrust laws, to the detriment of competition and consumers.

Competition policy can make an important contribution to such potential policy reforms, without necessarily coopting them. For example, it may be that NCAs for low-income workers serve no procompetitive goal, even if there are many labor markets in which NCAs do not harm competition but prove otherwise politically unobjectionable. That might provide a foundation for further state or federal policy reform, wholly apart from the question whether there are UMC violations that could support FTC competition rulemaking.

Finally, the Commission has asked about alternatives to the Proposed Rule, and “whether the rule should apply uniformly to all workers or whether there should be exemptions or different standards for different categories of workers.”[383] We believe that the existing literature simply does not permit the making of viable inferences regarding the different effects of potential alternative policies, and thus that the issuance of the Proposed Rule or any alternative NCA rule by the Commission would be premature. Further research could confirm the Starr, Prescott, and Bishara finding that the timing of an NCA disclosure bears critically on the wage impact of an NCA, for example.[384] If so, that might ground a general finding that the failure to disclose NCA terms at some point before the commencement of employment is a material omission, perhaps with sufficient frequency and effect to support a Mag-Moss UDAP rulemaking. At present, that too would be premature, however.

Indeed, if the existing evidence is to be taken at face value, arguably the wage-effect evidence, especially that adduced by Starr, Prescott, & Bishara, counsels against a broad prohibition on NCAs:

To be sure, regulatory regimes must sometimes rely on clear rules that ban (or allow) particular conduct, and such rules will be overinclusive or underinclusive. As then-Judge Breyer once explained, the cost of assessing the exact impact of each type of conduct would be prohibitive. The benefits of additional investigation do not always warrant the costs. However, if the anticipated impact on wages should drive the treatment of employee noncompete agreements, the cost of discriminating between contracts likely to reduce such wages and those likely to increase them is extremely low. Agencies and courts need simply ask whether the employer disclosed the agreement before acceptance. If the answer is “yes,” any presumption that such an agreement will reduce wages must evaporate. If anything, the presumption should shift in favor of a conclusion that the agreement will produce net benefits.[385]

That is not to say that no enforcement is ever warranted. The Commission has brought and settled four Section 5 cases in which they alleged that specific NCAs, under specific facts and circumstances, violated the prohibition of unfair methods of competition. Although the antitrust analysis in the public documents is not entirely clear, we do not maintain that there have been no NCAs that constitute UMC violations; and there might well be uninvestigated matters in which the Commission might demonstrate actual or likely harm to competition and consumers. In the alternative, as noted above, an FTC investigation might find a UDAP violation under some specific set of facts and circumstances.

But the Proposed Rule at issue here is not tied to credible evidence and is not nearly so restrained. The extensive concerns discussed in this comment militate against the Commission’s adoption of the Proposed Rule and, indeed, based on the available record, against any general competition rulemaking restricting the use of NCAs by the Commission.

[1] Non-Compete Clause Rule, 88 Fed, Reg, 3482 (RIN 3084, proposed Jan. 19, 2023) (to be codified at 16 C.F.R. Part 910) [hereinafter NPRM].

[2] See infra., Section II. See also, e.g., Daron Acemoglu & Robert Shimer, Holdups and Efficiency with Search Frictions, 40 Int. Econ. Rev. 827 (1999). The potential benefits of NCAs, and the importance of context in evaluating them, were discussed at the FTC’s 2020 workshop on NCAs. FTC, Non-competes in the Workplace: Examining Antitrust and Consumer Protection Issues (Jan. 9, 2020) [hereinafter FTC 2020 NCA Workshop; references to the workshop transcript will be cited by speaker and transcript page number (“Tr.”)]. A web page for the workshop, with links to the agenda, speaker biographies, public comments, and a transcript of the proceedings, is at FTC 2020 NCA Workshop, Kurt Lavetti, Tr. at 144 (“context matters. So although non-compete agreements can reduce earnings on average, in some contexts there’s evidence they might systematically increase earnings.”); id., the Hon. Noah Phillips, Tr. at 218 (“non-competes can serve good purposes, incentivizing investment in workers and protecting trade secrets, worthy goals in our increasingly knowledge-based economy”); id., Ryan Williams, Tr. at 175-6 (can “say some good things about non-compete contracts”); id., Ryan Nunn, Tr. at 126 (questioning utility of NCAs in various contexts, but noting NCAs can address a hold-up problem in training, and that “[f]irm-sponsored training is more common in states that more stringently enforce their non-compete agreements.”). See also, e.g., Norman D. Bishara & Evan Starr, The Incomplete Noncompete Picture, 20 Lewis & Clark L. Rev. 497, 505 (2016) (“Despite the potential cost of noncompetes for individuals and regions, the use and enforcement of noncompetes may also provide both private and social benefits.”).

[3] See, e.g., Aandrei Iancu & David Kappos, Banning Non-compete Agreements Hurts US Companies and Workers, The Hill (Mar. 23, 2023) (discussing importance of NCAs in protecting trade secrets); FTC 2020 NCA Workshop, Ryan Williams, Tr. at 178; id., Orly Lobel, Tr. at 12; id., Ryan Nunn, Tr. at 122-5, 134.

[4] See, e.g., John McAdams, Non-Compete Agreements: A Review of the Literature, Working Paper (2019), available at

[5] For an early case, see Mitchel v. Reynolds, 24 E.R. 347 (1711) (upholding a noncompete contract between a bakery and a baker, upon finding the contract’s terms, including a geographic restriction to the same parish as the bakery, reasonable).

[6] See infra., Section I.

[7] Id.

[8] See infra note 150, and accompanying text.

[9] See Bur. Labor Stats., Local Area Unemployment Statistics Geographic Concepts (Mar. 20, 2020),; see also Ioana Marinescu & Roland Rathelot, Mismatch Unemployment and the Geography of Job Search, 10 Am. Econ. J. Macro. 42, 42 (2018) (“[J]ob seekers are 35 percent less likely to apply to a job 10 miles (mi.) away from their zip code of residence.”).

[10] Id.

[11] See, e.g., Stephen G. Bronars, FTC Evidence that Noncompetes Reduce Wages is Inconclusive, Edgeworth Insights (Mar. 7, 2023),

[12] See 16 U.S.C. 46(f) (“To make public from time to time such portions of the information obtained by it hereunder as are in the public interest; and to make annual and special reports to the Congress and to submit therewith recommendations for additional legislation; and to provide for the publication of its reports and decisions in such form and manner as may be best adapted for public information and use.” Id.)

[13] See infra., Section III. See also Press Release, FTC Cracks Down on Companies That Impose Harmful Noncompete Restrictions on Thousands of Workers (Jan. 4, 2023), Since publishing the NPRM, the Commission has announced a fourth NCA settlement, also in the glass container industry. In the Matter of Anchor Glass Container, Corp., FTC Matter No. 211 0182 (Mar. 15, 2023) (decision and order).

[14] In the Matter of Prudential Security, et al., FTC File No. 2210026 (Jan. 4, 2023) at 2, (“This Consent Agreement is for settlement purposes only and does not constitute an admission by Proposed Respondents that the law has been violated as alleged in the Draft Complaint, or that the facts as alleged in the Draft Complaint, other than jurisdictional facts are true.”); In the Matter of Ardagh Group, et al., FTC File No. 2110182 (Jan. 4, 2023),; In the Matter of O-I Glass, Inc., FTC File No. 2110182 (Jan. 4, 2023),

[15] Dissenting Statement of Commissioner Christine S. Wilson, id. at 2 (“[E]ach Complaint runs three pages, with a large percentage of the text devoted to boilerplate language. Given how brief they are, it is not surprising that the complaints are woefully devoid of details that would support the Commission’s allegations.”).

[16] Compare Mitchel v. Reynolds, supra note 5 (upholding specific NCA restrictions as reasonable) with John Dyer’s Case, Year-Book Mich. 2 Hen. V, fo. 5, pl. 26 (1414) (rejecting NCA terms in an indenture contract as void under the common law).

[17] See Nat’l Soc’y of Prof’l Engrs. v. United States, 435 U.S. 679, 689 (1978) (“The Rule of Reason suggested by Mitchel v. Reynolds has been regarded as a standard for testing the enforceability of covenants in restraint of trade which are ancillary to a legitimate transaction, such as an employment contract or the sale of a going business.” (citing Mitchel v. Reynolds, 1 P. Wms. 181, 24 Eng. Rep. 347 (1711))); see also United States v. Addyston Pipe & Steel Co., 85 F. 271, 279 (6th Cir. 1898), aff’d as modified, 175 U.S. 211 (1899) (“It was of importance that business men and professional men should have every motive to employ the ablest assistants, and to instruct them thoroughly; but they would naturally be reluctant to do so unless such assistants were able to bind themselves not to set up a rival business in the vicinity after learning the details and secrets of the business of their employers.”).

[18] For a recent overview of state NCA laws, see, e.g., Russell Beck, Employee Noncompetes: A State-by-State Survey (last updated Feb. 11, 2023), available at

[19] The only express exception in the Proposed Rule regards NCAs executed in conjunction with the sale of a business, where the NCA applies to a seller who “is a substantial owner of, or substantial member or substantial partner in, the business at the time the person enters into the non-compete clause. Proposed § 910.1(e) would define substantial owner, substantial member, or substantial partner as an owner, member, or partner holding at least a 25% ownership interest in a business entity.” NPRM at 3515. While an exception providing for NCAs in conjunction with the sale of a business is common in states with some general hostility to NCAs, as under Cal. Bus. & Prof. Code § 16601, the identification of a 25% ownership requirement appears arbitrary and excessive. For example, California law permits certain NCAs for, inter alia, “[a]ny person who sells the goodwill of a business, or any owner of a business entity selling or otherwise disposing of all of his or her ownership interesting in the business entity.” Cal. Bus. & Prof. Code § 16601. We have not found any authority restricting such ownership to anything like a 25% share. That proposed restriction may prove far too narrow, not just when natural persons owning a startup or small business number more than four, but when, e.g., venture capital investment reduces the founders’ shares of a startup.

[20] FTC 2020 NCA Workshop, supra note 2.

[21] See, e.g., FTC 2020 NCA Workshop, supra note 2, Eric Posner, Tr. at 72-73 ((“I took upon myself the dreary task of trying to read every antitrust case ever decided involving non-competes, but it turned out not to be that dreary because there are only a handful of such cases — a few dozen or maybe more. Virtually none of them successful, basically they all fail. The plaintiffs always lose in these cases.”); id., Randy Stutz, Tr. at 60-68 (discussing difficulties of making out an antitrust case against an NCA under the rule of reason); Cf. Business Electronics Corp. v. Sharp Electronics Corp. 485 U.S. 717, 729 n. 3 (1988) (Justice Scalia citing the English common law case of Mitchel v. Reynolds in support of the proposition that “[t]he classic ‘ancillary’ restraint is an agreement by the seller of a business not to compete within the market.”) The NPRM notes that the Commission has identified 17 antitrust matters brought by private parties or state or federal antitrust authorities, under either the Sherman Act or state antitrust law, NPRM at 3496, suggesting that two of the matters the plaintiffs “were successful to some degree.” Id. In a 2015 matter, the “degree” of success reported was a federal district court’s denial of a motion to dismiss. Id. In the other matter—American Tobacco—the Supreme Court, in 1911, held that certain covenants not to compete were among a number of practices that, collectively violated the Sherman Act, although the Court expressly did not consider the various practices “legality, isolatedly viewed.” U.S. v. Am. Tobacco Co., 221 U.S. 106, 183 (1911). The other 15 matters did not reflect some degree of success. NPRM at 3496.

[22] NPRM at 3497-8.

[23] 15. U.S.C. § 46 (especially subsections (a), (b), and (f)).

[24] See infra., Section III.

[25] U.S. Bur. Labor Stats., Monthly Labor Review (Jun. 2022) (reporting 149,785 total employed),

[26] FTC 2020 NCA Workshop, supra note 2, Evan Starr, Tr. at 171.

[27] For a general discussion, see, e.g., A. Mitchell Polinsky & Steven Shavell, The Theory of Public Enforcement of the Law, in Handbook of Law and Economics, vol. 1, c. 6 (A. Mitchell Polinsky & Steven Shavell, eds., 2007); Steven Shavell, The Optimal Structure of Law Enforcement, 36 J. Law & Econ. 255 (1993).

[28] See, e.g., Oversight and Enforcement of the Antitrust Laws, Before the S. Comm. on the Judiciary, Subcomm. on Antitrust, Competition Policy, and Consumer Rights, 117th Cong. (2022) (Prepared Statement of the Fed. Trade Comm.) (“While we constantly strive to enforce the law to the best of our capabilities, there is no doubt that—despite the increased appropriations Congress has provided in recent years—we continue to lack sufficient funding.”),

[29] See infra., Section IV.

[30] The only express exception in the Proposed Rule regards NCAs executed in conjunction with the sale of a business, where the NCA applies to a seller who “is a substantial owner of, or substantial member or substantial partner in, the business at the time the person enters into the non-compete clause. Proposed § 910.1(e) would define substantial owner, substantial member, or substantial partner as an owner, member, or partner holding at least a 25% ownership interest in a business entity.” NPRM at 3515. While an exception providing for NCAs in conjunction with the sale of a business is common in states with some general hostility to NCAs, as under Cal. Bus. & Prof. Code § 16601, the identification of a 25% ownership requirement appears arbitrary and excessive. For example, California law permits certain NCAs for, inter alia, “[a]ny person who sells the goodwill of a business, or any owner of a business entity selling or otherwise disposing of all of his or her ownership interesting in the business entity.” Cal. Bus. & Prof. Code § 16601. We have not found any authority restricting such ownership to anything like a 25% share. That proposed restriction may prove far too narrow, not just when natural persons owning a startup or small business number more than four, but when, e.g., venture capital investment reduces the founders’ shares of a startup.

[31] For an example of a current legislative proposal, see, e.g., S.379—Freedom to Compete Act of 2023, 118th Cong. (2023-24) (which would amend the Fair Labor Standards Act to prevent the use of NCAs in employment contracts for certain non-exempt employees).

[32] See infra Section I.D.

[33] FTC 2020 NCA Workshop, supra note 2, Kurt Lavetti, Tr. at 139.

[34] See Local Area Unemployment Statistics Geographic Concepts, BLS, (Mar. 20, 2020); see also Marinescu & Rathelot, supra note 9 (“more than 80% of [all] job applications occur where the applicant and prospective employer are within the same ‘commuting zone.’”); Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law ¶ 552 (5th ed. 2022) (explaining that “commuting costs” limit a supplier’s ability to operate in a distant geographic market).

[35] See, e.g., Stephen G. Bronars, supra note 11.

[36] Bishara & Starr, supra note 2, at 537.

[37] Id. at 538.

[38] Id.

[39] Id. at 539.

[40] Id.

[41] Bishara & Starr, supra note 2, at 525.

[42] See NPRM at 3491.

[43] See id. at 3492. For example, the Commission admits the paper is not causal: “The study by Samila and Sorensen examines the enforceability of noncompete clauses across all states but does not consider changes in enforceability: they are therefore unable to rule out that their results could be due to underlying differences in the states rather than non-compete clause enforceability.”

[44] The NPRM discusses at least 10 of Professor Starr’s articles (and co-authored articles) repeatedly, and at length, with more than 40 citations.

[45] See McAdams, supra note 4. We also note that the named staff author of the review, John McAdams, moderated a session at the FTC 2020 NCA Workshop.

[46] See, e.g., FTC 2020 NCA Workshop, Kurt Lavetti, Tr. at 140 (“There’s also a new working paper by John [McAdams] that provides a great overview of this literature.”)

[47] McAdams, supra note 4, at 4.

[48] Id.

[49] Id.

[50] See, e.g., FTC 2020 NCA Workshop, Evan Starr, Tr. at 158 (noting “much harder to estimate causal effects using noncompete agreements); Tr. at 159 (lack of studies isolating random variation in use of noncompetes); Id., Ryan Williams, Tr. at 192 (regarding identification issues); Id., Ryan Nunn, Tr. at 192.

[51] The NPRM’s misapplication of a model in the Lavetti, Simon, and White paper is one example of a strained attempt to discount—and indeed invert—research findings. NPRM at 3501, 3524. The presentation of an alternative model—one that leads to merely “suggestive” observations, to make an ad hoc adjustment to account for an unobserved base rate of “enforceability” is simply conjecture. As we explain below, the “enforceability” assessment itself is deeply problematic. More than that, the NPRM seems to be suggesting a weak rewrite of the paper at issue, without any replication of the original work, all in the service of a finding that no existing study demonstrates or suggests. That is not credible evidence that anyone has demonstrated a negative impact of NCAs or NCA enforceability on physician wages.

[52] For a discussion of some of the difficulties raised by studies’ use of “enforceability” assessments, see Jonathan Barnett & Ted Sichelman, The Case for Noncompetes, 87 U. Chi. L. Rev. 953 (2020).

[53] Matthew S. Johnson, Kurt Lavetti, & Michael Lipsitz, The Labor Market Effects of Legal Restrictions on Worker Mobility, Working Paper (2020) at 2,

[54] Evan P. Starr, James J. Prescott, & Norman D. Bishara, Noncompete Agreements in the U.S. Labor Force, 64 J.L. & Econ. 53 (2021). Note that, whereas the prior Starr study considered the impact of NCA enforceability, this finding by Starr, Prescott, and Bishara has to do with signing an NCA.

[55] Id. at 75. In a footnote, the authors explain that this is among the observations that may be driven by unobservables. Id. at n. 34-35. We are not suggesting that the finding is definitive. Indeed, we spend a large part of these comments on data and methodological questions arising across the body of empirical literature. For those reasons, we suggest that this is an area that merits additional research.

[56] Alan J. Meese, Don’t Abolish Employee Noncompete Agreements, 57 Wake Forest L. Rev. 631, 702 (2022). On the inducement of additional training and/or production of information, see infra notes 85-90 and accompanying text, and Section II.

[57] Kurt Lavetti, et al., The Impacts of Restricting Mobility of Skilled Service Workers: Evidence from Physicians, 55 J. Human Resources 1025 (2020). We note that while many of these workers may be employees, others may be partners, other types of co-owners of a practice, or independent contractors.

[58] NPRM at 3495 (citing Norman D. Bishara, Fifty Ways to Leave Your Employer: Relative Enforcement of Non-Compete Clauses, Trends, and Implications for Employee Mobility Policy, 13 U. Pa. J. Bus. L. 751, 778–79 (2011)).

[59] Lavetti, et al., supra note 57.

[60] Id.

[61] Id. at 1055-7

[62] Id. at 1049.

[63] Id. at 1031.

[64] Omesh Kini, et al., CEO Non-Compete Agreements, Job Risk, and Compensation, 34 Rev. Fin. Stud. 4701 (2021).

[65] Id. Data regarding CEO contracts were compiled by hand based on SEC filings. The authors were able to identify 7,661 unique CEOs from ExecuComp, but found employment contracts for only 3,192; that is “only 41.67% of all CEOs in the ExecuComp database have employment contracts during our sample period.” Still, the study incorporates data on nearly half of all CEOs of publicly traded firms.

[66] Id. at 25-6.

[67] Id. at 26.

[68] Mark J. Garmaise, Ties that Truly Bind: Noncompetition Agreements, Executive Compensation, and Firm Investment, 27 J. L. Econ. & Org. 376 (2011). Data regarding executive compensation and board participation were taken from Standard and Poor’s ExecuComp database, which includes such data on the five most highly paid executives for 2,610 large publicly traded U.S. firms; on R&D investment, capital expenditures, and acquisitions were obtained at the firm level from Compustat. Id. at 388.

[69] Id. at 21 (“For a given executive, a shift to a tougher enforcement regime reduces compensation growth by 8.2%, which is 25% of the mean growth rate.”). Garmaise defines total compensation as the sum of salary, bonus, “other annual,” total value of restricted stock granted, total value of stock options granted, long-term incentive payouts, and “all other total” as defined and reflected in the ExecuComp data.

[70] Id. at 22.

[71] Id. at 25. Garmaise did not, however, find significant impact on firm value or profitability. Id. at 27-8.

[72] Umit G. Gurun, Noah Stoffman, & Scott E. Yonker, Unlocking Clients: Non-compete Agreements in the Financial Advisory Industry, 141 J. Fin. Econ. 1218 (2021).

[73] See id. at 1219. Eventually, over 1,500 firms adopted the Protocol. Prior to implementation of the Protocol, NCAs and NCA-related litigation had both been common in the industry.

[74] Id. at 1219-20.

[75] See Gurun, et al., supra note 72, at 1228.

[76] Id. at 1220. From a sample of advisors at 100 large firms, it was observed that misconduct tends to increase the likelihood of being fired by 23%, absent the Protocol, “but that this discipline is effectively undone when firms join the protocol.” See also id. at 1232 (“Once adviser fixed effects are included in the model, the coefficient estimates on “Firm in protocol” become both statistically and economically significant. The estimate in column 4, which is calculated using the sample of advisers working for employers with at least 100 advisers, indicates that the probability that an adviser engages in misconduct increases by 20 bps once his employer joins the protocol. Compared to an unconditional probability of misconduct of 47 bps, this is an increase in likelihood of over 40%.”).

[77] Id. at 1220.

[78] Id.

[79] See Matthew S. Johnson & Michael Lipsitz, Why are Low-Wage Workers Signing Noncompete Agreements?, J. Human Resources 0619-10274R2 (May 12, 2020).

[80] Id.

[81] Id. at 2, 17.

[82] Id. at 30. (“employment elasticity of the minimum wage in the lowest NCA enforcement states is much more negative (-0.38) than the average effect (p = .024). On the other hand, the point estimate on the interaction term … implies that the employment elasticity of the minimum wage is significantly closer to zero when NCAs are available.”)

[83] Id. at 28; see also, p. 43, Table 6.

[84] Id. at 16.

[85] Johnson & Lipsitz, supra note 79.

[86] Id. at 26.

[87] Evan Starr, Consider This: Wages, Training, and the Enforceability of Covenants Not to Compete, 72 Indus. & Labor Rel. Rev. 783 (2019).

[88] See id. at 785, 796-7. Note that Starr also observes lower wages associated with increased NCT enforceability. See id.

[89] See id. at 797.

[90] See FTC 2020 NCA Workshop, Evan Starr, Tr. at 162, 166, 174 (regarding, e.g., evidence of training incentives and wage/training tradeoffs); id., Kurt Lavetti, Tr. at 144-6 (regarding physician compensation and potential referral/patient sharing); id., Ryan Williams, Tr. at 187, et seq. (NCAs and risk management for CEOs). Cf. id., Howard Shelanski, Tr. at 263 (noting “ambiguity” in the research).

[91] The main natural experiment papers cited on wages are: Johnson, et al., supra note 53, Michael Lipsitz & Evan Starr, Low-Wage Workers and the Enforceability of Noncompete Agreements, 68 Mgmt. Sci. 143 (2021), and Natarajan Balasubramanian, Jin Woo Chang, Mariko Sakakibara, Jagadeesh Sivadasan, & Evan Starr, Locked In? The Enforceability of Non-Compete Clauses and the Careers of High-Tech Workers, 57 J. Human Resources S349 (2022).

[92] See supra notes 72-77 and accompanying text.

[93] NPRM at 3490 (citing Naomi Hausman & Kurt Lavetti, Physician Practice Organization and Negotiated Prices: Evidence from State Law Changes, 13 Am. Econ. J. Applied Econ. 258 (2021)).

[94] Or with other cognizable downstream effects, such as the impact of qualitative aspects of goods or services, output, etc.

[95] See, e.g., Hausman & Lavetti, supra note 93, at 259, 269, 271, fig. 1 & table 1.

[96] See, e.g., Martin Gaynor & Deborah Haas-Wilson, Change, Consolidation, and Competition in Health Care Markets, 13 J. Econ. Persp. 141 (1999); Deborah Haas-Wilson, Managed Care and Monopoly Power: the Antitrust Challenge (2003); Martin Gaynor, et al., The Industrial Organization of Health-care Markets, 53 J. Econ. Lit. 235 (2015); Dep’t Justice, Fed. Trade Comm’n. Improving Health Care: a Dose of Competition [Internet]. Washington (DC): FTC; 2004 Jul [cited 2017 Jul 31]. Available from:; Brent D. Fulton, Health Care Market Concentration Trends in The United States: Evidence and Policy Responses, 35 Health Affs. 1520 (2017).

[97] See Barnett & Sichelman, supra note 52.

[98] See Hausman & Lavetti, supra note 93, at 260 (“100 point increase in the establishment-based HHI causes a reduction in negotiated prices of about 1.4 percent to 1.9 percent on average. In contrast, the same increase in concentration caused by firm-level consolidation holding fixed establishment concentration causes prices to increase by 1.7 percent to 2.1 percent.”).

[99] See id. at 277-8.

[100] Id. at 260.

[101] See generally, e.g., Christopher Garmon, The Accuracy of Hospital Screening Methods, 48 RAND J. Econ. 1068 (2017) (reviewing post-merger price changes for 28 hospital mergers, initially published as BE Working Paper).

[102] See Herfindahl-Hirschman Index, Department of Justice (Jul. 31, 2018),; see also Competitive Effects, Federal Trade Commission (last visited Apr. 12, 2023),

[103] Hausman & Lavetti, supra note 93, at 269.

[104] See All-Player Claims Databases, Agency for Healthcare Research and Quality (February 2018),

[105] See supra note 95.

[106] Hausman & Lavetti, supra note 93, at 270.

[107] Hausman & Lavetti, id., at 276.

[108] See, e.g., Richard Schmalensee, Inter-Industry Studies of Structure and Performance, in 2 Handbook of Industrial Organization 951–1009 (Richard Schmalensee & Robert Willig eds., 1989); William N. Evans, Luke M. Froeb & Gregory J. Werden, Endogeneity in the Concentration–Price Relationship: Causes, Consequences, and Cures, 41 J. Indus. Econ. 431 (1993); Steven Berry, Market Structure and Competition, Redux, FTC Micro Conference (Nov. 2017), _steven_berry_keynote.pdf. See also Nathan Miller, et al., On the Misuse of Regressions of Price on the HHI in Merger Review, 10 J. Antitrust Enforcement 248 (2022).

[109] See, e.g., Thomas Koch & Shawn W. Ulrick, Price Effects of a Merger: Evidence from a Physicians Market, 59 Econ. Inquiry 790 (2021); Keith Brand & Ted Rosenbaum, A Review of the Economic Literature on Cross-Market Healthcare Mergers, 82 Antitrust L.J. 533 (2019); Thomas Koch, et al., Physician Market Structure, Patient Outcomes, and Spending: An Examination of Medicare Beneficiaries, 53 Health Servs. Res. 3549 (2018); Julie A. Carlson, et al., Economics at the FTC: Physician Acquisitions, Standard Essential Patents, and Accuracy of Credit Reporting, 43 Rev. Indus. Org. 303 (2013); Devesh Raval, et al., Using Disaster Induced Closures to Evaluate Discrete Choice Models of Hospital Demand, 53 RAND J. Econ. 561 (2022). See also, e.g., Martin Gaynor & Robert J. Town, The Impact of Hospital Consolidation—Update, Robert Wood Johnson Foundation, The Synthesis Project (2012) (Gaynor is a former Director of the FTC’s Bureau of Economics); Martin Gaynor & William B. Vogt, Competition Among Hospitals, 34 RAND J. Econ. 764 (2003); Leemore S. Dafny, et al., Regulating Hospital Prices Based on Market Concentration Is Likely to Leave High-Price Hospitals Unaffected, 40 Health Aff. 1386 (September 2021) (Dafny was Deputy Director for Health Care Antitrust in the FTC’s Bureau of Economics from 2012-13); Leemore S. Dafny, Hospital Industry Consolidation—Still More to Come?, 370 New Eng. J. Med. 198 (2014).

[110] See, e.g., Christopher Garmon, Hospital Mergers—Retrospective Studies to Improve Prediction, CPI Antitrust Chronicle (July 2017).

[111] See, e.g., Garmon, supra note 101 (reviewing post-merger price changes for 28 hospital mergers, initially published as BE Working Paper); Deborah Haas?Wilson & Christopher Garmon, Hospital Mergers and Competitive Effects: Two Retrospective Analyses, 18 Int. J. Econ. Bus. 17 (2011); Orly Ashenfelter, et al., Retrospective Analysis of Hospital Mergers, 18 Int. J. Econ. Bus. 5 (2011); Patrick S. Romano & David J. Balan, A Retrospective Analysis of the Clinical Quality Effects of the Acquisition of Highland Park Hospital by Evanston Northwestern Healthcare, 18 Int. J. Econ. Bus. 45 (2010); John Simpson, Geographic Markets in Hospital Mergers: A Case Study, 10 Int. J. Econ. Bus. 291 (2003); Michael G Vita & Seth Sacher, The Competitive Effects of Not?For?Profit Hospital Mergers: A Case Study, 49 J. Indus. Econ. 63 (2001).

[112] Joseph Farrell, Paul Pautler, & Michael Vita, Economics at the FTC: Retrospective Merger Analysis with a Focus on Hospitals, 35 Rev. Indus. Org. 369 (2009).

[113] See citations referenced supra, note 108.

[114] See Overview of the Merger Retrospective Program in the Bureau of Economics, Federal Trade Commission (last visited Apr. 12, 2023),

[115] Garmon, Accuracy of Hospital Screening Methods, supra note 101, at 1070.

[116] Id.

[117] See Hausman & Lavetti, supra note 94, at 275-7.

[118] NPRM at 3490.

[119] Id.

[120] NPRM at 3482.

[121] Id.

[122] Id. (citing Michael Lipsitz & Mark Tremblay, Noncompete Agreements and the Welfare of Consumers, Working Paper (2021),

[123] See, e.g., Joint Statement of the Antitrust Division of the U.S. Department of Justice and the Federal Trade Commission on Certificate-of-Need Laws and Alaska Senate Bill 62 (2017),; FTC Policy Perspectives on Certificates of Public Advantage, Staff Policy Paper (2022); Statement of the Federal Trade Commission In The Matter of Phoebe Putney Health Services, Inc., et al., FTC Docket No. 9348 (Sep. 4, 2014), The Phoebe Putney matter illustrates, among other things, how certificate of need programs can impede effective remedies to demonstrably anticompetitive provider mergers. Cf. FTC v. Phoebe Putney Health Sys., Inc., 568 U.S. 216 (2013).

[124] NPRM at 3492.

[125] Zhaozhao He, Motivating Inventors: Non-Competes, Innovation Value and Efficiency, Working Paper (2021),

[126] NPRM at 3492.

[127] See NPMR at 3492-3. The papers are: Gerald A. Carlino, Do Non-Compete Covenants Influence State Startup Activity? Evidence from the Michigan Experiment, Fed. Reserve Bank of Phila. Working Paper 21-26 (2021) (finding a correlation that suggests an increase in patenting with enforceability); Fenglong Xiao, Non-Competes and Innovation: Evidence from Medical Devices, 51 Rsch. Pol’y 1 (2022) (finding enforceability correlates with an increase in the quantity of innovation as measured by the introduction of new medical devices); Sampsa Samila & Olav Sorenson, Noncompete Covenants: Incentives to Innovate or Impediments to Growth, 57 Mgmt. Sci. 425 (2011) (finding a correlation that suggests venture capital induces less patenting when non-competes are enforceable); and Raffaele Conti, Do Non-Competition Agreements Lead Firms to Pursue Riskier R&D Strategies?, 35 Strategic Mgmt. J. 1230 (2014) (finding an ambiguous effect that the Commission summarizes as “riskier research and development strategies lead to more breakthrough innovations, but also lead to more failures, leaving the net impact unclear”).

[128] NPRM at 3493.

[129] He, supra note 125.

[130] NPRM at 3527.

[131] See, e.g., Thomas M. Jorde & David J. Teece, Innovation and Cooperation: Implications for Competition and Antitrust, 4 J. Econ. Persp. 75 (1990) (regarding the organization requirements (and implications) of innovation, modeling complexities, and common market failures in the “market for know-how”).

[132] For a classic review of the literature on the economic significance of patents, and difficulties in determining what aspects of economic activity are, and should be, captured by patent statistics, see Zvi Griliches, Patent Statistics as Economic Indicators: A Survey, 28 J. Econ. Lit. 1661, (1990),

[133] See, e.g., Eli Greenbaum, No Forum to Rule Them All: Comity and Conflict in Transnational FRAND Disputes, 94 Wash. L. Rev. 1085 (2019). Because objective valuation of FRAND terms may often be difficult, authorities tend to focus on the conditions under which (and forums in which) good faith negotiation can occur. Compare U.S. PTO/U.S. DOJ, Draft Policy Statement on Licensing Negotiations and Remedies for Standards-Essential Patents Subject to Voluntary F/RAND Commitments (Dec. 19, 2019), (emphasizing conditions of negotiation) with U.S. PTO/NIST/U.S. DOJ, Withdrawal of 2019 Draft Policy Statement on Licensing Negotiations and Remedies for Standards-Essential Patents Subject to Voluntary F/RAND Commitments (Jun. 8, 2022), (emphasizing case-by-case evaluation of conduct).

[134] Cf. Bronwyn H. Hall & Dietmar Harhoff, Recent Research on the Economics of Patents, 4 Ann. Rev. Econ. 541 (2012) (reviewing the literature and noting the importance of patents in certain sectors, while also concluding that “the sheer size and growth of the recent literature might lead one to assume that patents are an extremely important instrument of economic development and growth, which therefore attract a great deal of interest from researchers and policy makers. But this seems at odds with the weak evidence that patents serve as an incentive for innovation and the fact that relatively few firms find them an important means of securing returns to innovation”).

[135] See, e.g., Evan Starr, Natarajan Balasubramanian & Mariko Sakakibara, Screening Spinouts? How Non-compete Enforceability Affects the Creation, Growth, and Survival of New Firms, 64 Mgmt. Sci. 552 (2018); see also FTC 2020 NCA Workshop, Evan Starr, Tr. at 162-163 (observing wage vs. training tradeoffs); id., Lavetti, Tr. at 144-145 (findings indicating wage gains in certain contexts, but not others).

[136] Starr, et al, supra note 135 at 552.

[137] Id.

[138] Jessica Jeffers, The Impact of Restricting Labor Mobility on Corporate Investment and Entrepreneurship, Working Paper (September 7, 2022), available at

[139] Id. at 1. Jeffers also found decreased entry.

[140] Matt Marx, Deborah Strumsky, and Lee Fleming, Mobility, Skills, and the Michigan Non-compete Experiment, 55 Mgmt Sci 875 (2009) [hereinafter Marx, et al., 2009].

[141] Gerald Carlino, Do Non-Compete Covenants Influence State Startup Activity? Evidence from the Michigan Experiment, Fed. Res. Bank of Philadelphia Working Paper #17-30 (2017) (comparing Michigan with both an all-states control group and with 10 states with statutory limits on NCA enforcement both before and after the Michigan change).

[142] Id. at 16, 20.

[143] See Barnett & Sichelman, supra note 52.

[144] Id. at 1010. (“The simplification of these doctrinal complexities in the Marx et al. study renders that study’s key assumption—namely, that nonenforcing states always apply their own law—flawed, and thus confounds its causal identification strategy.”)

[145] Id. at 1018 (“Marx et al., however, overlook this complexity and erroneously assume that nonenforcing states always apply their own law so as to void a noncompete agreement that falls under the law of another state.”).

[146] NPRM at 3486.

[147] Id.

[148] See, e.g., FTC 2020 NCA Workshop, Evan Starr, Tr. at 173.

[149] Bishara & Starr, supra note 2, at 537.

[150] Since 2019, five states (Maine, Maryland, New Hampshire, Rhode Island, and Virginia have adopted statutes preventing enforcement of NCAs against low-wage workers; and since 2020, four states (Colorado, Illinois, Oregon, and Washington) and the District of Columbia have adopted similar limits pertaining to middle-income (to mid-plus) workers. For a recent overview of state NCA laws, see, e.g., Beck, supra note 18. The NPRM’s examples of event studies mostly concern estimates of relative “enforceability” across many, and often subtle or ambiguous, changes in state laws, instead of studies that focus on unique, major changes in NCA law. It’s not at all clear that these are properly regarded as event studies, but, in any case, as we discuss in detail below, they rest on a soft and problematic metric for legal change.

[151] McAdams, supra note 4, at 4.

[152] See Balasubramanian, et al, supra note 91.

[153] Id. at S351.

[154] Id. at S349.

[155] McAdams, supra note 4, at 4.

[156] Haw. Rev. Stat. § 480-4(d) (2021).

[157] Id.

[158] Id.

[159] Id.

[160] Id.

[161] FTC 2020 NCA Workshop, Orly Lobel, Tr. at 10; id., Evan Starr, Tr. at 172-73 (regarding NCAs, non-disclosure, non-solicitation of clients, non-solicitation of co-workers, IP-assignment terms, “most firms… are using all of these provisions together.”).

[162] Balasubramanian, et al, supra note 91, at S353, n. 9.

[163] Technicolor, Inc v. Traeger, 551 P.2d 163 (1976).

[164] Id.

[165] Quick Facts, Hawaii, US Census Bureau (last visited Apr. 12, 2023),

[166] Chris Kolmar, 100 Largest Employers in Hawaii for 2022, Zippia (June 2021),

[167] Balasubramanian, et al, supra note 91, at S351.

[168] NPRM at 3523 (“Caution is recommended in interpreting this extrapolation, however, since results from one sector within one state may not necessarily inform outcomes that would occur in the rest of the country.”).

[169] Id. (“Extrapolating from the estimates for Hawaii to the average impact on high-tech workers in each state, a prohibition such as the one in this proposed rule would increase earnings of high-tech workers in the average state by 4.8%.”).

[170] See Lipsitz & Starr, supra note 91, at 162.

[171] State Population Totals and Components of Change, US Census Bureau, (estimating Oregon’s population at 4,237,291 as of July 1, 2020).

[172] McAdams, supra note 4 at 17.

[173] Id. at n. 34 (citing James D. Hamilton & Michael T. Owyang, The Propagation of Regional Recessions, 94 Rev. Econ. and Stats. 935 (2012)).

[174] McAdams, supra note 4, at 17-18 (internal citations omitted).

[175] Lipsitz & Starr, supra note 91, at 148.

[176] Id. at 147 (stating that the law brought about “dramatic changes to Oregon’s policy on NCAs, effective January 1, 2008 for new contracts” (pre-existing contracts were governed by the old law)).

[177] Or. Rev. Stat. § 653.020(3).

[178] Id. at 653.020(4).

[179] Id. at 653.020(1)(a)-(d).

[180] Id. 653.020(1)(e).

[181] Id. at 653.020(2).

[182] Id. at 653.020(6).

[183] Id.

[184] Melissa Ilyse Rassas, Explaining the Outlier: Oregon’s New Non-Compete Agreement Law & the Broadcasting Industry, 11 U. PA. J. Bus. L. 452-3 (2009) (internal citations omitted).

[185] Elizabeth H. White & Jonathan G. Rue, Effective Use of Non-solicitation and Confidentiality Agreements in Oregon after S.B. 169, K & L Gates Hub (Apr. 1, 2022),

[186] Rassas, supra note 184, at 453.

[187] Id. (internal citations omitted).

[188] Nike, Inc. v. McCarthy, 379 F.3d 576 (9th Cir. 2004).

[189] Ikon Office Solutions, Inc. v. Am Office Prods., Inc., 178 F. Supp. 2d 1154, 1160 (D. Or. 2001) (quoting Dymock v. Norwest Safety Protective Equip. for Oregon Indus., Inc., 172 Or. App. 399, 405-06 (2001)) (expounding the bounds of Oregon noncompetition law).

[190] See M. Scott McDonald and Jacqueline C. Johnson, Across the Board: Changes Are in the Works for Noncompete Agreements, Littler Mendelson, PC (Aug. 2007), (noting that the 2008 bill took “a hard (and more complex) stance on noncompetes”).

[191] See also notes 167 & 168, supra, and accompanying text.

[192] Michigan Antitrust Reform Act § 18, MCL § 445.788.

[193] Marx, et al., 2009, supra note 140.

[194] Matt Marx, Jasmit Singh & Lee Fleming, Regional Disadvantage? Employee Non-compete Agreements and Brain Drain, 44 Res. Pol’y 394 (2015) [hereinafter Marx, et al., 2015].

[195] Id. at 397.

[196] See Barnett & Sichelman, supra note 52.

[197] Marx, et al., 2009, supra note 140 at 875.

[198] Marx, et al., 2015, supra note 194, at 394.

[199] Mich. Public Act No. 329 of 190 (Michigan Antitrust Reform Act § 18, MCL § 445.788.)

[200] Id. at 445.774a(2).

[201] Marx, et al., 2015, supra note 194, at 396.

[202] Barnett & Sichelman, supra note 52, at 1022. Marx, et al., 2015 supra note 194, at 395, cites as the “governing case,” Application Group Inc. v. Hunter Group Inc., 72 Cal. Rptr. 2d 73 (1st Distr. 1998).

[203] Barnett & Sichelman, supra note 149, at 1022 (citing Compton v. Joseph Lepak, D.D.S., P.C., 397 N.W.2d 311, 316 (Mich. Ct. App. 1986) (“When an agreement or contract is entered into in violation of the statute, repeal of that statute does not make the agreement valid because the Legislature cannot validate a contract which never had a legal existence.”).

[204] See Barnett & Sichelman, supra note 52, at 1018. Bishara cites Edwards v. Arthur Anderson, LLP, 189 P.3d 285 (Cal. 2008) for the proposition that California courts have a strong public policy interest in upholding California NCA law). See Norman Bishara, Fifty Ways to Leave Your Employer, supra note 58, at 757. But see, e.g., In re Howmedica Osteonics Corp, 867 F.3d 390 (3d Cir. 2017) (upholding forum selection clauses specifying New Jersey and Michigan jurisdiction for suit to enforce NCAs against two California residents).

[205] See Barnett & Sichelman, supra note 52, at 1021.

[206] Id.

[207] Needless to say, it is highly unlikely that all tech professional job changes in the pertinent interval were caused by the change in NCA enforceability.

[208] Ronald Gilson, The Legal Infrastructure of High-Technology Industrial Districts: Silicon Valley, Route 128, and Covenants Not to Compete, 74 N.Y.U. L. Rev. 575, 578-9, (1999).

[209] See, e.g., Lina Khan, Noncompetes Depress Wages and Kill Innovation, N.Y. Times (Jan. 9, 2023),; FTC 2020 NCA Workshop, Orly Lobel, Tr. at 15, 22; Starr, Tr. at 168.

[210] See Barnett & Sichelman, supra note 52. See also, Russell Beck, Correlation Does Not Imply Causation: The False Case of Silicon Valley and Boston’s Route, Fair Competition Law (Jul. 9, 2019),

[211] California law provides that “Except as provided in this chapter, every contract by which anyone is restrained from engaging in a lawful profession, trade, or business of any kind is to that extent void.” Cal. Bus. & Prof. Code § 16600. California statutes also provide an exception for NCAs for a person selling ownership interest in a business, the assets of a business, or the goodwill of a business. Cal. Bus. & Prof. Code § 16600.

[212] See, e.g., City Carpet Beating Works v. Jones, 102 Cal. 506 (Cal. 1894) (citing Cal. Civil Code §§ 1673, 1674).

[213] For an example of a current legislative proposal, see, e.g., S.379—Freedom to Compete Act of 2023, 118th Cong. (2023-24) (which would amend the Fair Labor Standards Act to prevent the use of NCAs in employment contracts for certain non-exempt employees).

[214] NPRM at 3487 (citing Starr, Prescott & Bishara, supra note 54).

[215] California law provides that “[e]xcept as provided in this chapter, every contract by which anyone is restrained from engaging in a lawful profession, trade, or business of any kind is to that extent void.” Cal. Bus. & Prof. Code § 16600. It also provides for an exception for NCAs for a person selling ownership interest in a business, the assets of a business, or the goodwill of a business. Cal. Bus. & Prof. Code § 16600.

[216] NPRM at 3486, n. 62 (citing Brian M. Malsberger, et al., Covenants Not to Compete: A State-by-State Survey, Bloomberg BNA (2012) and Beck, supra note 18. An earlier version of the Malsberger survey was P. Jerome Richey, Brian M. Malsberger, Covenants Not to Compete, A State-by-State Survey (1990 & Cum. Supp. 1991).

[217] Id. at 3486.

[218] See generally, e.g., Bishara & Starr, supra note 2.

[219] That is, as published in state codes and, to a lesser extent, in published judicial decisions. See, e.g., Richey & Malsberger, supra note 216.

[220] Norman Bishara, Fifty Ways to Leave Your Employer, supra note 58, at 771.

[221] Id. at 773-7. It’s not clear whether or how answers to individual questions influenced each other. For example, would a given court holding on an employer’s protectable interest be scored differently according to the state’s statute of general application?

[222] Id. at 773.

[223] Id. at 775.

[224] Id.

[225] Klaus Schwab, World Economic Forum Global Competitiveness Report 2019 (2019), available at

[226] Id. at 611-25.

[227] Hausman & Lavetti, supra note 93, at 263.

[228] Id.

[229] SWAT 24 Shreveport Bossier v. Bond, 808 So.2d 294 (La. 2001).

[230] Id. at 296, 307.

[231] Id. at 298 (citations omitted).

[232] Id. at 303.

[233] Id. at 303.

[234] Id. at 299 (citing Summit Inst. For Pulmonary Medicine & Rehabilitation, Inc. v. Prouty, 691 So.2d 1384 (La. App. 2 Cir. 1997).

[235] Id. at 300-1 (citing Scariano Bros., Inc. v. Sullivan, 719 So.2d 131 (La. App. 4 Cir. 1998) and Moreno & Assocs. v. Black, 741 So.2d 91 (La. App. 3 Cir. 1999).

[236] California law provides that “Except as provided in this chapter, every contract by which anyone is restrained from engaging in a lawful profession, trade, or business of any kind is to that extent void.” Cal. Bus. & Prof. Code § 16600. It also provides for an exception for NCAs for a person selling ownership interest in a business, the assets of a business, or the goodwill of a business. Cal. Bus. & Prof. Code § 16600. Section 16600, in its present form, appears to date to 1941, but the provision has both statutory and common law roots extending into the 19th Century. See, e.g., City Carpet Beating Works v. Jones, 102 Cal. 506 (Cal. 1894) (citing Cal. Civil Code §§ 1673, 1674).

[237] NPRM at 3486.

[238] A few states do provide for suits by the employee. 2020 amendments to the Virginia code restrict NCAs for low-wage employees. Va. Code § 40.1-28.7:8. Such employees can bring an action to have a putative NCA declared void, and for other equitable relief including restitution and money damages. While violations of the pertinent provision involve attempts to enforce an NCA against a low-wage worker, not the mere existence of sanctioned terms, such violations are subject to regulatory penalties. But as such, neither the administration of those regulatory penalties nor their effects have been studied.

[239] See, e.g., Scott Cunningham, Causal Inference: The Mixtape 21 (2021).

[240] NPRM at 3487 (quoting Starr, Prescott & Bishara, supra note 54, at 73.

[241] Starr, Prescott & Bishara, supra note 54, at 57.

[242] Id.

[243] NPRM at 3487.

[244] See Johnson, et al., supra note 53.

[245] Id.

[246] NPRM at 3522.

[247] See, e.g., Ole Agersnap & Owen Zidar, The Tax Elasticity of Capital Gains and Revenue-Maximizing Rates, 3 Am. Econ. Rev. Insights 399 (2021).

[248] NPRM at 3505.

[249] Id.

[250] Id.

[251] Id. at 3493 (citing Starr, Consider This, supra note 87 (finding that moving from mean NCA enforceability to no NCA enforceability would decrease the number of workers receiving training by 14.7% in occupations that use NCAs at a relatively high rate); Jeffers, supra note 138 (finding that knowledge-intensive firms invest 32% less in capital equipment following decreases in the enforceability of NCAs); Johnson & Lipsitz, supra note 79 (finding that hair salons that use NCAs train their employees at a higher rate and invest in customer attraction through the use of digital coupons at a higher rate, both by 11 percentage points)).

[252] NPRM at 3505.

[253] Id. at 3505, et seq.

[254] NPRM at 3535.

[255] See O’Bannon v. Nat’l Collegiate Athletic Ass’n, 802 F.3d 1049, 1076 (9th Cir. 2015) (refusing to credit less restrictive alternative that was not “virtually as effective” as challenged restraint); County of Tuolumne v. Sonora Cmty. Hosp., 236 F.3d 1148, 1159 (9th Cir. 2001) (requiring plaintiffs show alternative is “virtually as effective” in serving defendant’s objective, concluding proposed LRAs were less effective, and ruling in favor of defendants at final net-effects step). See also C. Scott Hemphill, Less Restrictive Alternatives in Antitrust Law, 116 Colum. L. Rev. 927, 944-5 (2016) (“Equal effectiveness is an explicit limitation in cases, jury instructions, and commentary.”).

[256] NPRM at 3506. The Commission points to the Uniform Trade Secrets Act, which provides a state law civil cause of action for trade secret misappropriation; the federal Defend Trade Secrets Act, which establishes a similar cause of action under federal law; and the Economic Espionage Act of 1996, which criminalizes theft of a trade secret for either the benefit of a foreign entity or the economic benefit of anyone other than the owner.

[257] Id., citing Uniform Trade Secrets Act § 1(4).

[258] Id. See, e.g., Florida Uniform Trade Secrets Act, FL. Stat. § 688, et seq. (2022) for an example of the USTA, as adopted by the state of Florida.

[259] See, e.g., Hamer Holding Group, Inc. v. Elmore, 202 Ill. App. 3d 994, 560 N.E.2d 907 (1st Dist. 1990) (determining that the plaintiff’s customer list information was not protectable as a trade secret because it could be readily duplicated by anyone with access to the Secretary of State’s information, even though it cost the plaintiff $60,000 to condense).

[260] USTA § 1(4)(ii).

[261] See Michelle L. Evans, Trade Secret Misappropriation of Former Employer’s Customer List §8, 139 Am. Jur. Trials 293 (orig. published 2015) (“Limiting the number of employees within a company who are aware of the trade secret information tends to protect trade secret status.”).

[262] See UTSA § 1(4).

[263] See Iancu & Kappos, supra note 3.

[264] See id. (“If a high-level executive at a company that depends on proprietary technology moves to a Chinese competitor, for example, and shares highly confidential information taken from his last employer, that last employer’s competitive edge might evaporate forever to China’s benefit. By the time the afflicted company sues to enforce trade secret laws, it may be too late; irreparable damage is often done when the information is disclosed to the new employer because that bell can’t be unrung.”). See also Lauren Weber, FTC Plan to Ban Noncompetes Clauses Shifts Companies’ Focus, Wall St. J. (Jan. 17, 2023), (“’Once someone goes to another company, you’re really on the honor system. You have no way to monitor what information is being disclosed or not.’”) (quoting Julie Levinson Werner of Lowenstein Sandler LLP).

[265] NPRM at 3507.

[266] Restatement (Second) of Contracts § 352 (1981) (“Damages are not recoverable for loss beyond an amount that the evidence permits to be established with reasonable certainty.”).

[267] Id. at § 352, cmt. a (“Courts have traditionally required greater certainty in the proof of damages for breach of contract than in the proof of damages for a tort. … [This principle] excludes those elements of loss that cannot be proven with reasonable certainty. The main impact of the requirement of certainty comes in connection with recovery for lost profits.”).

[268] See Natarajan Balasubramanian, Evan Starr, & Shotaro Yamaguchi, Bundling Employment Restrictions and Value Appropriation from Employees Working Paper (Jan. 2023) at 35, See also NPRM at 3485 n. 42 (“This survey also found that non-compete clauses are often used together with other restrictive employment covenants, including non-disclosure, nonrecruitment, and non-solicitation covenants. Id. at 17 (reporting that respondents that had a noncompete clause reported having all three of the other restrictive employment covenants 74.7% of the time).”).

[269] NPRM at 3487 (“Balasubramanian et al. [supra, note 91] find that while non-compete clause use is associated with 2.1–8.2% greater earnings (compared with individuals with no post-contractual restrictions), this positive association is due to noncompete clauses often being bundled with non-disclosure agreements.”).

[270] NPRM at 3507.

[271] Restatement (Second) of Contracts § 367(1) (1981) (“A promise to render personal service will not be specifically enforced.”); id. at § 367, cmt. a (“A court will refuse to grant specific performance of a contract for service that is personal in nature. The refusal is based in part on the undesirability of compelling the continuance of personal association after disputes have arisen and confidence and loyalty are gone and, in some instances, of imposing what might seem like an involuntary servitude.”).

[272] See, e.g., Lumley v. Wagner, 42 Eng. Rep. 687 (1852).

[273] NPRM at 3507.

[274] Id.

[275] NPRM at 3498.

[276] The FTC convened fourteen sets of Hearings on Competition and Consumer Protection in the 21st Century, running from September 2018 through June 2019. Two sets of hearings were of special relevance: On October 16, 2018, a full day of hearings was devoted to issues to do with Antitrust in Labor Markets, including NCAs; and on June 12th, 2019, one of the panels in the hearing comprising a Roundtable with State Attorneys General included discussion of NCAs and other labor restrictions. Information regarding the full set of hearings, including links to agendas and transcripts for individual hearings, can be found at Fed. Trade Comm’n, Hearings on Competition and Consumer Protection in the 21st Century [hereinafter FTC 21st C. Hearings],

[277] FTC 2020 NCA Workshop, supra note 2.

[278] Fed. Trade Comm’n and U.S. Dep’t Justice, Workshop: Making Competition Work: Promoting Competition in Labor Markets (Dec. 6-7, 2021) [hereinafter FTC/DOJ 2021 Labor Competition Workshop],

[279] Press Release, FTC Cracks Down on Companies That Impose Harmful Noncompete Restrictions on Thousands of Workers (Jan. 4, 2023), Since publishing the NPRM, the Commission has announced a fourth NCA settlement, also in the glass container industry. In the Matter of Anchor Glass Container, Corp., FTC File No. 211 0182 (Mar. 15, 2023) (decision and order)

[280] NPRM at 3497. To the best of our knowledge, the Commission has never reported any evaluation of the 2019 petition.

[281] See citations supra at note 14.

[282] See Dissenting Statement of Commissioner Christine S. Wilson, supra note 15.

[283] See supra note 279.

[284] In the Matter of Prudential Security, et al., supra note 14, at 3.

[285] That is our initial reaction to what are, of course, questions of fact, on which the Commission has not reported.

[286] Prudential Security, Inc. v. Pack, No. 18-015809-CB (Mich. Cir. Ct. Dec. 13, 2018).

[287] In the Matter of Prudential Security, supra note 14, at 5.

[288] Id.

[289] NPRM at 3499 (citing FTC, Policy Statement Regarding the Scope of Unfair Methods of Competition Under Section 5 of the Federal Trade Commission Act, Commission File No. P221202 (Nov. 10, 2022)).

[290] See, e.g., Daniel Gilman & Gus Hurwitz, The FTC’s UMC Policy Statement: Untethered from Consumer Welfare and the Rule of Reason, ICLE Issue Brief (Nov. 23, 2022),

[291] For a 50-state review, see Beck, supra note 18. See also FTC 2020 NCA Workshop, Evan Starr, Tr. at 195 (regarding 14 dimensions along which state NCA laws vary, under categorization in Malsberger, et al. treatise).

[292] FTC 2020 NCA Workshop, supra note 2, Howard Shelanski, Tr. at 263.

[293] See text accompanying notes 16-21, supra.

[294] NPRM at 3497-8.

[295] Id. at 3497.

[296] Id.

[297] Id. at 3498.

[298] See generally Comments submitted to FTC to Hold Workshop on Non-compete Clauses Used in Employment Contracts, Docket No. FTC-2019-0093 [hereinafter NCA Workshop Comments], See, e.g., Comment Submitted by Public Citizen—Alex Harman, Comment #0286; Comment Submitted by Open Markets Inst.—Udit Thakur, Comment #0313; Comment Submitted by United States Senate—9 Signatures, Comment #0017. See also, FTC 2020 NCA Workshop supra note 2, Eric Posner, Tr. at 71, 74-77.

[299] NCA Workshop Comments, id., Comment Submitted by Office of the Atty. General for the District of Columbia on Behalf of State Attorneys General (20 signatures), Comment #322.

[300] Id. See also, e.g., FTC 2020 NCA Workshop supra note 2, Orly Lobel, Tr. at 11.

[301] Id.

[302] NCA Workshop Comments, supra note 298, Comment Submitted by Russell Beck (and 21 co-signatories), Comment #0319. See also, e.g., Comment of the Global Antitrust Institute, Antonin Scalia Law School, George Mason University, Comment #0243 (“We are concerned, however, that many proposals to address … [concerns about NCAs] through ex ante antitrust regulatory interventions, such as an FTC rule, are ill-suited and will likely do more harm than good.”); Comment Submitted by the U.S. Chamber of Commerce—Sean Heather and Glenn Spencer, Comment #0303 (“the Chamber sees as unnecessary rulemaking by the FTC under either its unfair methods of competition authority (assuming such authority even exists), or its unfair and deceptive practices authority.”); Comment Submitted by The Center On Executive Compensation (“Center”)—Andrew Maletz, Comment #0264 (association representing chief human resources officers of large firms stating that “We believe an FTC rule regarding non-compete agreements is unnecessary.”)

[303] Id., Comment of the Global Antitrust Institute, Antonin Scalia Law School, George Mason University, Comment #0243.

[304] We note that the Section has submitted comments in response to the NPRM Comment from the American Bar Ass’n, Antitrust Law Section, FTC-2023-0007-9980, March 2, 2023, available at These more recent comments note that the literature remains limited, but that it tends to support the proposition that NCAs for low-wage workers are “generally harmful and not justified” by procompetitive rationales that may apply with other workers. The comment does not appear to advocate for any specific regulation, and it does not appear to address the question whether the FTC, specifically, should adopt a regulation restricting the use of NCAs for low-wage workers.  We agree with the Section’s claim that the literature remains limited, and with their suggestion that standard rationales for NCAs can seem strained in the case of low-wage workers. Restrictions on low-wage workers may generate various policy concerns. However, as should be clear from our comments overall, we do not agree that the literature provides adequate grounds for the adoption of any FTÇ competition regulation under Sections 5 and 6(g) the FTC Act.

[305] NCA Workshop Comments, supra note 298, Comment Submitted by the Antitrust Law Section of the ABA—Brian R. Henry, Comment #0329.

[306] FTC 2020 NCA Workshop, supra note 2, Howard Shelanski, Tr. at 283.

[307] Id, Noah Phillips, Tr. at 220.

[308] Id. at 221.

[309] West Virginia v. Envtl. Prot. Agency, 142 S. Ct. 2587 (2022).

[310] Id. at 17.

[311] See FTC, Rules and Guides Currently Under Review, (last visited Apr. 12, 2023),

[312] See FTC 2020 NCA Workshop, supra note 2, William Kovacic, Tr. at 36. Cf. Aaron Nielson, Tr. at 234-44 (stating that it is an “open question” whether courts would sustain a challenge to the FTC’s authority.)

[313] See id., William Kovacic, Tr. at 37.

[314] See id., Howard Shelanski, Tr. at 264-5.

[315] See id.

[316] See id., Evan Starr, Tr. at 158, 173.

[317] See id. at 173.

[318] See id. at 166.

[319] See id. at 174.

[320] See id., at 173.

[321] See id..

[322] See id., Kurt Lavetti, Tr. at 151-2.

[323] See id., Evan Starr, Tr. at 162, 166, 174; Ryan Williams, Tr. at 179, et seq. (negotiation and compensation for CEOs); id., Kurt Lavetti, Tr. at 144-46 (regarding physician compensation); id., Howard Shelanski, Tr at 263 (noting “ambiguity” in the research) and 284 (describing training investments what would not occur under a ban).

[324] Id., Evan Starr, Tr. at 162, 166, 174; Ryan Williams, Tr. at 179, et seq. (negotiation and compensation for CEOs); Kurt Lavetti, Tr. at 144-6 (regarding physician compensation); Howard Shelanski, Tr. at 263 (noting “ambiguity” in the research).

[325] Id., Ryan Williams, Tr. at 178.

[326] Id., Kurt Lavetti, Tr. at 144-6.

[327] Id., Noah Phillips, Tr. at 218.

[328] Id., Kurt Lavetti, Tr. at 139.

[329] See John McAdams, Non-Compete Agreements: A Review of the Literature, supra note 4.

[330] See, e.g., FTC 2020 NCA Workshop, supra note 2, Kurt Lavetti, Tr. at 140 (referring to John McAdams, the workshop panel moderator: “There’s also a new working paper by John that provides a great overview of this literature.”).

[331] See id., at 4.

[332] See id.

[333] See id.

[334] Id., William Kovacic, Tr. at 37.

[335] Regarding competition in labor markets generally, see, e.g., Diana Furchtgott-Roth, Antitrust and Modern U.S. Labor Markets: An Economics Perspective, Harv. J. L. & Pub. Pol’y Per Curiam 1 (Summer 2022) (“data from the Bureau of Labor Statistics of the U.S. Department of Labor show that exercise of monopsony power is generally not occurring in today’s 21st century economy, nor has it been characteristic of labor markets over the past half century.”); Richard A. Epstein, The Application of Antitrust Law to Labor Markets: Then and Now, 15 N.Y.U. J. Law & Liberty 709 (2021).

[336] The only express exception in the Proposed Rule regards NCAs executed in conjunction with the sale of a business, where the NCA applies to a seller who “is a substantial owner of, or substantial member or substantial partner in, the business at the time the person enters into the non-compete clause. Proposed § 910.1(e) would define substantial owner, substantial member, or substantial partner as an owner, member, or partner holding at least a 25% ownership interest in a business entity.” NPRM at 3515. While an exception providing for NCAs in conjunction with the sale of a business is common in states with some general hostility to NCAs, as under Cal. Bus. & Prof. Code § 16601, the identification of a 25% ownership requirement appears arbitrary and excessive. For example, California law permits certain NCAs for, inter alia, “[a]ny person who sells the goodwill of a business, or any owner of a business entity selling or otherwise disposing of all of his or her ownership interesting in the business entity.” Cal. Bus. & Prof. Code § 16601. We have not found any authority restricting such ownership to anything like a 25% share. That proposed restriction may prove far too narrow, not just when natural persons owning a startup or small business number more than four, but when, e.g., venture capital investment reduces the founders’ shares of a startup.

[337] NPRM at 3485. The latest survey from the BLS survey suggests approximately 18%, but survey findings vary somewhat, and, at least roughly, cluster in the neighborhood of 20%. And BLS estimates a workforce of approximately 150 million employed persons. U.S. Bur. Labor Stats., Monthly Labor Review (Jun. 2022) (reporting 149,785 total employed),

[338] FTC 2020 NCA Workshop, supra note 2, Evan Starr, Tr. at 171.

[339] See, e.g., Oversight and Enforcement of the Antitrust Laws, Prepared Statement of the Federal Trade Commission, supra note 28 (“While we constantly strive to enforce the law to the best of our capabilities, there is no doubt that—despite the increased appropriations Congress has provided in recent years—we continue to lack sufficient funding.”).

[340] NPRM at 3509.

[341] 16 C.F.R. § 315.

[342] 15 U.S.C. §§ 7601-7610.

[343] FTC Staff Comment on Proposed Additional Regulations Issued by the North Carolina State Board of Opticians (Jan. 13, 2011), available at

[344] 15 U.S.C. § 7608.

[345] See FTC, Possible Barriers to E-Commerce: Contact Lenses: A Report From the Staff of the Federal Trade Commission (Mar. 29, 2004),; FTC, The Strength of Competition in the Sale of Rx Contact Lenses: An FTC Study (Feb. 2005),

[346] 16 C.F.R. § 315.3(a)(1).

[347] Contact Lens Rule Request for Comment, 80 FR 53272 (Sept. 3, 2015).

[348] CLR Notice of Proposed Rulemaking, 81 FR 88526 (Dec. 7, 2016).

[349] CLR Supplemental Notice of Proposed Rulemaking, 84 FR 24664 (May 28, 2019)

[350] CLR Final Rule, 85 Fed. Reg. 50668 (2020).

[351] Id. at 50671 (citing the 2016 CLR NPRM, supra note 348).

[352] See generally, e.g., Thomas W. Merrill, Antitrust Rulemaking: The FTC’s Delegation Deficit, Admin. L. Rev. (forthcoming 2023), available at

[353] The most recent case to opine on the Commission’s substantive competition rulemaking authority dates to 1973. See Nat’l Petroleum Refiners Ass’n v. Fed. Trade Comm’n, 482 F.2d 672, 697-8 (D.C. Cir. 1973).

[354] See, e.g., comments submitted to this Docket by TechFreedom, the United States Chamber of Commerce, and the Washington Legal Foundation.

[355] See Workforce Mobility Act of 2023, H.R. 731, 118th Cong. (2023-2024) and Workforce Mobility Act of 2023, S. 220, 118th Cong. (2023-2024).

[356] Nat’l Petroleum Refiners, 482 F.2d at 697-8.

[357] P.L. 93-637 (1975) (codified at 15 U.S.C. § 2301, et seq.).

[358] See, e.g., Maureen K. Ohlhausen & Ben Rossen, Dead End Road: National Petroleum Refiners Association and FTC “Unfair Methods of Competition Rulemaking, Truth on the Market (Jul. 13, 2022),

[359] Id.

[360] See, e.g., Noah Joshua Phillips, Against Antitrust Regulation, Am. Enterprise Inst. Report (Oct. 13, 2022),

[361] See, e.g., Thomas W. Merrill, Antitrust Rulemaking: The FTC’s Delegation Deficit, Colum. Pub. L. Res. Paper, available at; Thomas W. Merrill, Re-Reading Chevron, 70 Duke L.J. 1153 (2021).

[362] 15 U.S.C. § 6(g) (reference to s.57a(a)(2) omitted).

[363] 5 USC §553.

[364] See Merrell, Antitrust Rulemaking, supra note 352, at 28.

[365] National Petroleum Refiners, 482 F.2d, at 689.

[366] See Merrell, Antitrust Rulemaking, supra note 352, at 28.

[367] AMG Capital Management, LLC v. FTC, 141 S. Ct. 1341 (2021). The FTC had argued that monetary damages were impliedly available under the power in section 13(b) of the FTC Act to seek injunctive relief, but the Supreme Court disagreed and restricted the Agency to injunctive relief only, without the implicit grant of damage.

[368] West Virginia v. Envtl. Prot. Agency, 142 S. Ct. 2587 (2021).

[369] Id.

[370] Util. Air Regul. Grp. v. Envtl. Prot. Agency, 573 U.S. 302, 324 (2014).

[371] See, e.g., Workforce Mobility Act of 2023, supra note 355.

[372] See FTC, Policy Statement Regarding the Scope of Unfair Methods of Competition Under Section 5 of the Federal Trade Commission Act (Nov. 10, 2022), available at See also Gilman & Hurwitz, supra note 290.

[373] AMG Capital Mgmt., 141 S. Ct. 1341.

[374] Axon Enterprise, Inc. v. F.T.C., 598 U.S. __ (2023).

[375] Regarding nondelegation generally, see, e.g., Thomas W. Merrill, Rethinking Article I, Section 1: From Nondelegation to Exclusive Delegation, 104 Colum. L. Rev. 2097 (2004). Compare Gundy v. United States, 139 S. Ct. 2116, 2123 (2019) (reaffirming the traditional test permitting the delegation of discretionary authority if constrained by an “intelligible principle”) with id. at 2135-7 (Gorsuch, J. dissenting) (insisting that delegations should be limited to filling the details in statutes with major questions resolved by Congress). See also Noah Joshua Phillips, Against Antitrust Regulation, supra note 360.

[376] A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935).

[377] U.S. Const., Art. I.

[378] Schechter Poultry, 295 U.S. at 529.

[379] This discussion draws from the analysis in William C. MacLeod, Regulating Beyond the Rule of Reason, __ Geo. Mason. L. Rev. __ (forthcoming).

[380] Shechter Poultry, 295 U.S. at 533-4.

[381] FTC 2020 NCA Workshop, supra note 2, Shelanski, Tr. at 293.

[382] Id., Shelanski, Tr. at 191.

[383] NPRM at 3516.

[384] See Starr, Prescott & Bishara, supra note 54.

[385] Meese, supra note 56, at 702-3.

Continue reading
Antitrust & Consumer Protection

Brief of ICLE and Law & Economics Scholars in Deslandes v. McDonald’s

Amicus Brief Interest of Amicus Curiae[1] The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the . . .

Interest of Amicus Curiae[1]

The International Center for Law & Economics (“ICLE”) is a nonprofit, non-partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law and economics methodologies to inform public policy debates and has longstanding expertise in antitrust law.

Amici also include twenty scholars of antitrust, law, and economics at leading universities and research institutions across the United States. Their names, titles, and academic affiliations are listed in Appendix A. All have longstanding expertise in antitrust law and economics.

Amici respectfully submit that their amicus brief will aid the Court in reviewing the order of dismissal by explaining that Plaintiffs did not plead and could not prove any plausible product or geographic market. This is a point that Plaintiffs attempt to elide in their appellate brief that warrants this Court’s attention. The foundation of almost every antitrust claim is a plausible market definition, yet Plaintiffs’ claims in this case are premised on a labor market—limited to one company (McDonald’s), but nationwide in scope—that has no basis in economic reality.

In addition, amici explain why Plaintiffs’ claims are subject to rule of reason scrutiny. This case involves a vertical, intrabrand restraint between McDonald’s and its franchisees, which promoted interbrand competition. It was not a naked restraint on trade, but rather an ancillary restriction that furthered McDonald’s procompetitive goal of creating a strong and stable brand. Treating such a restraint as per se unlawful, as Plaintiffs ask this Court to do, would stifle the type of legitimate cooperation that facilitates output and would ultimately harm consumers.


Until 2017, McDonald’s franchise agreements included a provision that prevented franchisees from hiring workers from other McDonald’s restaurants within the six-month period immediately following the workers’ prior employment. Two employees sued McDonald’s in a putative class action, alleging that this provision—which the parties refer to as “Paragraph 14”—was an unlawful agreement under Section 1 of the Sherman Act, 15 U.S.C. § 1, that harmed competition by artificially suppressing wages. Plaintiffs argued that Paragraph 14 was per se unlawful or failed “quick look” review; on appeal (but not below), Plaintiffs also argue that Paragraph 14 fails scrutiny under the rule of reason. Brief of Appellants at 31-33, Deslandes v. McDonald’s USA LLC, No. 22-2333 & 22-2334 (7th Cir.) (“App. Br.”).

Under the rule of reason, Plaintiffs had the burden to plead and prove the relevant product and geographic markets within which McDonald’s allegedly exerted market power and caused the alleged anticompetitive effects. Plaintiffs’ claim has always been premised—explicitly or implicitly—on a single-brand and nationwide labor market for McDonald’s employees. (§ I.A). Although Plaintiffs shy away from that market on appeal, it is the only one discernible from the record and the only one on which Plaintiffs’ claims could possibly be premised. But that market is fatally flawed along the two axes that typically delineate antitrust markets: (1) geography and (2) products or services.

First, the relevant labor market is local: there is no national market for fast-food restaurant employees, as Plaintiffs suggest. Low-skilled restaurant workers sell their labor in local markets, primarily to avoid long commutes or relocation. The economic (and commonsense) reality is that a person applying for a McDonald’s job in Chicago, Illinois is not also looking for a McDonald’s job in Florida or Montana; nor are local McDonald’s restaurants recruiting employees nationwide. (§ I.B). Second, there is no McDonald’s-specific labor market for restaurant employees. McDonald’s restaurants compete vigorously with other fast-food and quick-service restaurants—and with firms outside the restaurant industry—for labor. Empirical economic evidence refutes Plaintiffs’ arguments to the contrary. (§ I.C).

Further, Plaintiffs cannot avoid the rule of reason (or their burden to plead and prove a relevant market) by invoking per se scrutiny. (§ II). First, Paragraph 14 was not a horizontal restraint; it was a vertical, intrabrand restraint between McDonald’s and its franchisees. (§ II.A). Second, Paragraph 14 was not a naked restraint on trade but was instead “ancillary” to McDonald’s procompetitive endeavor of developing its brand. (§ II.B). Third, Paragraph 14 does not fall in the narrow class of restrictions—like price fixing—universally condemned as anticompetitive. Nor are there sufficient economic studies showing that restraints like Paragraph 14 have a demonstrable and negative impact on trade, such that there would be a basis to depart from the default rule of reason standard. (§ II.C).

I.             There is no Nationwide, Single-Brand Market for McDonald’s Employment

Market definition enables courts to determine whether firms possess market power capable of restricting competition. The market provides a locus for the assessment of that market power and of whether that power has been exploited to harm competition and consumers.

Accordingly, “courts usually cannot properly apply the rule of reason without an accurate definition of the relevant market.” Ohio v. Am. Express Co., 138 S. Ct. 2274, 2285 (2018). The rule of reason requires a court to assess the “actual effect” that a defendant’s conduct has on competition. Id. at 2284 (quoting Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 768 (1984)). And without knowing the relevant market, “there is no way to measure the defendant’s ability to lessen or destroy competition.” Id. at 2285 (quoting Walker Process Equip., Inc. v. Food Mach. & Chem. Corp., 382 U.S. 172, 177 (1965)).[2]

A.             Plaintiffs Must Allege and Prove a Relevant Market

In the district court, Plaintiffs did not clearly define the relevant market in which Paragraph 14 allegedly harmed competition; and that failure defeats their claim under the rule of reason. See Agnew v. Nat’l Collegiate Athletic Ass’n, 683 F.3d 328, 347 (7th Cir. 2012) (affirming dismissal where “[p]laintiffs appear to have made the strategic decision to forgo identifying a specific relevant market,” and rejecting “post hoc arguments attempting to illuminate a buried market allegation”).

Plaintiffs are wrong to assert that they can rely on “direct evidence” of anticompetitive harm to avoid establishing a relevant market. App. Br. at 31. Even if they had such evidence, Plaintiffs would still have the burden to sketch out the “rough contours” of the relevant market and to show that McDonald’s commanded a substantial share of that market. See Republic Tobacco Co. v. N. Atl. Trading Co., 381 F.3d 717, 737 (7th Cir. 2004) (“[I]f a plaintiff can show the rough contours of a relevant market, and show that the defendant commands a substantial share of the market, then direct evidence of anticompetitive effects can establish the defendant’s market power[.]” (emphases added)). But Plaintiffs never inform this Court what product or geographic markets are involved here, even “roughly” speaking. Accordingly, Plaintiffs’ references to their supposed “direct evidence” of anticompetitive harm, without regard to any market boundaries, App. Br. at 31-32, do not suffice.[3]

Rather than define a market, Plaintiffs fault the district court for assuming that their claim “depended upon a single, nationwide geographic market.” App. Br. at 31. But that assumption came not from the district court but from Plaintiffs themselves—as that was the only market potentially discernible in their complaints. See First Am. Compl. ¶¶ 1, 117; Turner Compl. ¶¶ 109-13 (implying a single-brand, nationwide market). While Plaintiffs try to obscure their single-brand, nationwide market for McDonald’s labor on appeal, that is the only one that Plaintiffs rely upon in their opening brief, albeit obliquely. See App. Br. at 31 (arguing that Paragraph 14 “suppressed worker pay nationwide”); id. at 33 (arguing that McDonald’s and its franchisees were the “discrete group of buyers” that were able to “hold down wages”).

Plaintiffs’ proposed market is both implausible and economically unsound. Antitrust markets typically have two dimensions: (1) a geographic market and (2) a product or services market. See Brown Shoe Co. v. United States, 370 U.S. 294, 324 (1962). Plaintiffs’ single-brand, nationwide market fails along both dimensions.

B.        The Relevant Market is Local, not National

First, to identify a relevant geographic market, the court must make a “careful selection of the market area in which the seller operates, and to which the purchaser can practicably turn for supplies.” Republic Tobacco, 381 F.3d at 738 (citing Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 327 (1961)). In labor markets, the “sellers” are workers or job applicants selling their services (like Plaintiffs), and “purchasers” are employers (like McDonald’s) that compete with other firms to hire, employ, and retain the workers.

As the district court correctly found, low-wage restaurant employees sell their labor locally and McDonald’s restaurants compete only with geographically proximate employers to purchase that labor. See D.E. 372 (“Class Cert. Op.”) at 20-21. Fast-food and quick-serve restaurant employees are generally low-skilled and/or entry-level workers who “are looking for a position in the geographic area in which they already live and work, not a position requiring a long commute or a move.” Id. at 21.

While some employees might relocate for other reasons first, and then seek a restaurant job, it is not economically plausible that they would “search long distances for a low-skill job with the idea of then moving closer to the job.” Id. The costs of relocation—in economic terms, the “search” costs and “transition” costs—would far exceed any cost-adjusted increases in pay and benefits.

In practical terms, an hourly McDonald’s worker in Apopka, Florida who does not otherwise desire to move to California would not uproot her life, and leave family and friends, for a McDonald’s job in Los Angeles—even if the Los Angeles franchisee offers to raise her wages a few dollars per hour. The employee has many alternative opportunities that do not require relocation and, in any event, the higher cost of living in Los Angeles would negate the benefits of the wage increase. The total costs of relocation likely outweigh the marginal wage gain.

For similar reasons, an employee is unlikely to commute long distances—for example, from Urbana to Chicago, Illinois—to work at a McDonald’s, even if the McDonald’s in Chicago pays slightly higher wages than the one in Urbana. A marginal wage increase would not offset the time and “commuting costs”—i.e., gas and mileage. See Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles and Their Application ¶ 552 (5th ed. 2022) (explaining that “commuting costs” limit a supplier’s ability to operate in a distant geographic market).

Nor is it plausible that a McDonald’s franchisee in Apopka, Florida would recruit workers nationwide. That franchisee also faces search costs in the labor market. It must advertise its job openings, hire recruiters, and interview applicants, among other things. It would not be worthwhile to incur the substantial costs of a nationwide search for employees, when those employees would likely remain at the job for a relatively short time, and when there are many local workers with similar skills who could fill the role.

To be sure, some highly skilled employees in other industries—for example, corporate executives or professional athletes—undoubtedly search for high-paying or prestigious jobs nationwide. And their potential employers recruit nationwide. For those types of job-seekers, their decisions turn on the scarcity of those jobs, the substantial personal and financial investments (“sunk costs”) they have made to be qualified for such positions (which essentially “lock” them into the nationwide market), and the high salaries or total compensation that make relocation worthwhile. Likewise, for the hiring firm, the search is justified by the small number of qualified candidates, widely distributed across the country, and by the expected benefits. For example, it would be worthwhile for a firm to  search far and wide for a new CEO, knowing that there are only a few people in the country with the skills and leadership ability to lead the company out of financial troubles.

These rarefied exceptions confirm that the vast majority of labor markets are “geographically quite small.” Herbert Hovenkamp, Competition Policy for Labour Markets, U. Pa. Inst. L. & Econ. ¶ 12 (May 17, 2019). Applicants for low paying and fungible jobs have fundamentally different incentives than do corporate executives and professional athletes. The former do not have adequate incentives for a national search, given the substantial costs, plus relocation or long-distance commuting.

These intuitive points are supported by data and established economic methods. Several recent economic studies demonstrate that, “in a wide range of industries[,] geographic markets for employment are rather small,” and that this is “particularly true of low-wage employees.” Areeda & Hovenkamp, supra ¶ 550b (collecting studies). One empirical study shows that “more than 80% of [all] job applications occur where the applicant and prospective employer are within the same ‘commuting zone.’” Ioana Marinescu & Roland Rathelot, Mismatch Unemployment and the Geography of Job Search, 10(3) Am. Econ. J. Macro. 42 (2018).

The U.S. Bureau of Labor Statistics (“BLS”) employs a methodology that confirms the district court’s findings with respect to localized markets. When BLS collects employment and unemployment statistics, it examines the “economically integrated geographic area within which individuals can reside and find employment within a reasonable distance or can readily change employment without changing their place of residence.” BLS, Local Area Unemployment Statistics Geographic Concepts, (Mar. 20, 2020). According to BLS, metropolitan and micropolitan areas are the “major” “labor market areas” (LMAs)—not the country as a whole. Id. The balance of the nation’s LMAs comprise a larger number of smaller geographic areas. Id. BLS identifies thousands of these small LMAs for each decennial census, based on its analysis of highly localized commuting flows. Id; see also Data.Gov, Commuting Zones and Labor Market Areas,  (Nov. 10, 2020).

Plaintiffs’ own expert witnesses acknowledged that a nationwide market was implausible in this case. Dr. Peter Capelli conceded that restaurant workers are employed in local geographic markets, defined by commuting distances. Class Cert. Op at 22-23 (citing Capelli Dep. at 235-36, D.E. 302-1 at 608-09) (“My testimony is that for the restaurant employees in particular, the crew employees, there may be labor markets of different geographic size and that the key issue there might not even be size, it might be commuting distance.”). And Dr. Hal Singer calculated that only 8% of McDonald’s employees commute ten or more miles to work. Id. at 23 (citing Singer Rep. ¶ 64, D.E. 271-5 at 54).

Other evidence in this case also demonstrates that market conditions vary substantially by location. McDonald’s own guidelines on worker pay account for local conditions. D.E. 302-19 at -997; D.E. 302-19 at -432, -444; see also D.E. 380 ¶¶ 40-50. The guidelines do not set forth a single, rigid nationwide formula—which one would expect to see if McDonald’s understood that its franchisees were competing for labor nationwide.

As this Court recently explained, “[t]he antitrust statutes require a ‘pragmatic’ and ‘factual’ approach to defining the geographic market,” and “[t]he market must ‘correspond to the commercial realities of the industry.’” Sharif Pharmacy, Inc. v. Prime Therapeutics, LLC, 950 F.3d 911, 917 (7th Cir. 2020) (quoting Brown Shoe, 370 U.S. at 336). Plaintiffs’ proposed nationwide market is neither pragmatic nor factual. It ignores the obvious commercial reality that local McDonald’s franchisees do not compete nationwide for low-skilled labor, and it ignores the empirical evidence that McDonald’s employees (and similarly situated low-wage employees) do not commute long distances or relocate for these types of jobs.

C.        The Market is not McDonald’s-Specific

The outer boundaries of a product market are defined by the “reasonable interchangeability of use or the cross-elasticity of demand between the product itself and substitutes for it.” Sharif Pharmacy, 950 F.3d at 918 (quoting Brown Shoe, 370 U.S. at 325). Cross-elasticity of demand, here, reflects the degree to which a significant increase or decrease in wages paid by alternative employers changes the number of workers hired or hours worked (quantity demanded) at the employer in question. In the classic company town, the cross-elasticity of demand is zero. When there are substitutes, it is positive, indicating that “consumers would respond to a slight increase in the price of one product by switching to another product.” Todd v. Exxon Corp., 275 F.3d 191, 201-02 (2d Cir. 2001); IIA Areeda & Hovenkamp, supra ¶ 562a.

In a “buyer-side” labor monopsony, such as Plaintiffs have alleged, the market is defined not by the competing sellers (employees), but by the availability of competing buyers (employers). See Todd, 275 F.3d at 201 (citing Roger D. Blair & Jeffrey L. Harrison, Antitrust Policy and Monopsony, 76 Cornell L. Rev. 297, 297-301, 308 (1991)). Thus, the key question is whether employees would see the various employers as reasonable substitutes for one another, such that they would respond to compensation changes by seeking those substitutes. Id. If so, then any of the reasonable substitutes must fall within the market definition for the plaintiff’s market to be plausible. See Rock v. Nat’l Collegiate Athletic Ass’n, 928 F. Supp. 2d 1010, 1021 (S.D. Ind. 2013) (“Plaintiffs’ proposed market is impermissibly narrow because it ignores the existence of [substitutes].”).

In a properly defined labor market, the greater the availability of substitute employers, the less “market power” each employer can have, as employees can go elsewhere when one employer lowers its wages or fails to meet wage increases by others. When a labor market is highly concentrated, by contrast, the employer may be able to exert monopsony buying power. For example, a pure labor monopsony might exist in the classic “company town,” where there is only one large employer—such as a lumber mill or coal mine—that has wide discretion to set wages without employees leaving for other jobs. Those employees are, in effect, “locked in” to selling their labor to the single employer. Cf. Eastman Kodak Co. v. Image Tech. Servs., Inc., 504 U.S. 451, 472-73, 476 (1992) (seller can “lock in” customer to aftermarket for equipment repairs, if customers already purchased equipment in the foremarket and switching costs are high). In the modern economy, however, such examples are rare; and the market for hourly restaurant employees bears no resemblance to a company town.

Nevertheless, Plaintiffs here allege that McDonald’s is a monopsony buyer of labor—not just in a highly concentrated market, but in a single-brand market for McDonald’s labor. App. Br. at 7. In other words, McDonald’s allegedly has monopsony power to set the wages at which it hires and retains employees because employees (or applicants) generally would not see any non-McDonald’s employment opportunities as reasonable substitutes.

Plaintiffs’ theory thus depends upon implausible assumptions about McDonald’s market power and the elasticity of demand: that substantial wage increases by alternative local employers—including, but not limited to, fast food or quick-service restaurants like Wendy’s, Burger King, KFC, and Subway—would have little or no impact on the ability of a local McDonald’s franchise to hire or retain workers at a given antecedent wage. That is not only implausible in this case, but there is no evidence that it is the norm across low-wage labor markets. See, e.g., Jordan D. Matsudaira, Monopsony in the Low-Wage Labor Market? Evidence from Minimum Nurse Staffing Regulations, 96(1) Rev. Econ. & Statistics 92, 102 (2014) (empirical data in low-wage labor markets are “difficult to reconcile with the notion that low-wage labor markets such as those for fast food workers are monopsonist”).

These assumptions underlying Plaintiffs’ alleged market defy basic economic principles and common sense. Courts are highly skeptical of alleged single-brand markets with no substitutes at all, as those markets are almost always artificial and litigation-driven. See, e.g., Sheridan v. Marathon Petroleum Co., 530 F.3d 590, 595 (7th Cir. 2008) (rejecting single-brand market); Todd, 275 F.3d at 200 & n.3 (“Cases in which dismissal on the pleadings is appropriate frequently involve . . . failed attempts to limit a product market to a single brand, franchise, institution, or comparable entity….” (collecting cases)).

McDonald’s does not have monopsony power in any relevant market because, from an employee’s perspective, there are many reasonable substitutes in the geographic areas in which a given McDonald’s franchisee operates. McDonald’s franchisees compete vigorously for labor with other local employers within and without the quick-service industry. As the district court observed, there are multitudes of adequate, substitute employers for low-wage employees—including (1) other quick-serve restaurants, like Burger King, Wendy’s, Arby’s, KFC, Taco Bell, Chick-fil-A, Chipotle, and Jimmy John’s; (2) other restaurants, like Applebee’s; (3) larger retailers, like Walmart, Sam’s Club, and Costco; and (4) a host of other businesses like grocery stores and hotels. Class Cert. Op. at 21-22.

Even narrowing substitutes to just quick-service restaurants, the district court found numerous (even hundreds of) substitute employers within close geographic proximity to each of the named Plaintiffs, and the number of alternative quick-service restaurants dwarfed the number of McDonald’s franchises in the same area. Id. at 6.

Plaintiffs’ claim of monopsony buying power thus depends on the far-fetched premise that McDonald’s can suppress wages, notwithstanding hundreds of non-McDonald’s quick-service restaurants—and numerous other alternative employers—near Plaintiffs and other putative class members. In reality, if McDonald’s lowered its wages (or other employers raised theirs, and McDonald’s did not match), McDonald’s would lose its supply of labor. See Madison 92nd St. Assocs., LLC v. Courtyard Mgmt. Corp., 624 F. App’x 23, 29 (2d Cir. 2015) (in labor markets involving entry-level work, “it is beyond doubt that [employers] would have to increase their wages to retain any employees” if nearby employers “suddenly doubled the wages they paid to their employees”).

To evade this economic and commonsense reality, Plaintiffs and their expert, Dr. Singer, attempted to attribute monopsony power to McDonald’s as a structural feature of the labor economy in general. In other words, in their view, all employers of low-wage workers enjoy monopsony power in their labor markets, and, ipso facto, McDonald’s has market power in a single-brand market nationwide (or in each and every local labor market) because it employs low-wage workers.

In his expert report, for example, Dr. Singer characterizes economic literature as arguing that the ability of firms to suppress wages is “surprisingly common throughout the economy.” D.E. 271-5 (Singer Rep. ¶ 17). Thus, he stated, it “would be consistent with this literature . . . [t]hat McDonald’s-branded restaurant owners also face a low elasticity of labor supply[.]” Id.; see also id. ¶ 39 (“In light of [the economic literature], it is likely that both McDonald’s Franchisees and the McOpCos would continue to exercise some degree of monopsony power over their employees, even in the absence of the No-Hire Agreement.”).

On appeal, Plaintiffs point to Dr. Singer’s opinions about labor monopsonies, in general, as evidence that McDonald’s, specifically, had market power and could suppress wages in a proposed single-brand market for McDonald’s labor. See App. Br. at 30-33 (arguing that Plaintiffs “buttressed their direct proof of detrimental effects with substantial economic scholarship showing that low?wage employers, including those in the fast food sector, possess market power over their employees”); see also id. at 51 (arguing that “an overview of economic research demonstrat[es] that employers exercise significant monopsony power over their employees”).

Whatever the “structural” features of broader markets—comprising many low-wage employers and firms—those features say nothing about one company’s individual market power and ability to harm competition in a properly defined market. In any event, even if such a sweeping claim could suffice to carry one’s burden of proof as to a specific defendant, the claim is still inaccurate and inconsistent with the economic features of the quick-service restaurant industry. That industry is characterized by low barriers to entry for employees, extremely high turnover rates, and substantial wage growth.[4]

First, “[f]ood and beverage serving and related workers typically have no requirements for formal education or work experience to enter the occupation.” BLS, Occupational Outlook Handbook, Food and Beverage Serving and Related Workers, (Sept. 8, 2022). Unlike corporate executives and professional athletes, fast-food restaurant employees do not make large investments to obtain restaurant positions. Thus, they are not “locked in” to any one restaurant or even the restaurant industry, contrary to Plaintiffs’ argument that there are “high switching costs” for restaurant employees. App. Br. at 34.

Indeed, recent data show that the turnover in the quick-service sector is incredibly high, at around 144%—which means that if a restaurant has a total of 30 people on staff at any given time, it faced about 43 departures in the last year alone. See Daily Pay, The Turnover and Retention Rates for QSR Businesses (Nov. 15, 2022). BLS also recently found that the seasonally-adjusted “quit rate” for the accommodation and food services industry was 5.8% as of October 2022—higher than any other industry. See BLS, Economic News Release, Job Openings and Labor Turnover, Table 4, Quits levels and rates by industry and region, seasonally adjusted (Jan. 4, 2023). When restaurant workers quit, moreover, they frequently leave the restaurant industry altogether, creating high numbers of job openings for new entrants into restaurant sector employment. See BLS, Occupational Outlook Handbook, Food and Beverage Serving and Related Workers, (Sept. 8, 2022).

Moreover, fast-food workers in 2021, on average, benefitted from a 10% wage increase from 2020. See Dominick Reuter and Madison Hoff, A 10% pay increase and 8 other stats show how crazy it is to work in fast food right now, Business Insider (Aug. 24, 2021). Such rapid wage growth undermines Plaintiffs’ claim that monopsony power is a structural feature of the restaurant industry. To the contrary, this wage growth suggests that normal market factors of supply and demand are at play. Cf. Richard A. Epstein, Antitrust Overreach in Labor Markets: A Response to Eric Posner, 15 NYU J. L. & Liberty 407, 432 (2022) (“There is no global evidence, given the chronic fluctuations and frequent shortages, to believe that labor markets are rife with hidden pockets of monopsony power that function as economic black holes.”).

Finally, the most robust study of the relationship between wages and hours worked in fast-food labor markets yielded results that are consistent with competitive labor markets, rather than a monopsony model. See David Neumark and William Wascher, Minimum Wages and Employment: A Case Study of the Fast-Food Industry in New Jersey and Pennsylvania: Comment, 90 Am. Econ. Rev. 1362, 1382 (2000) (critiquing the famous Card & Krueger minimum-wage study using direct data, rather than surveys, and looking at hours worked rather than overall employment; and finding that fast-food employers make wage and hour decisions consistent with “the prediction of the textbook [competitive] model”); see also David Neumark and William L. Wascher, Minimum Wages 106 (MIT Press 2007) (reviewing literature on low-wage restaurant workers and concluding that “the low-wage labor market can be reasonably approximated by the neoclassical competitive model”).

Overall, the empirical data show that the labor markets in which restaurants participate are dynamic and competitive, not dominated by monopsonists with the power to suppress wages at will.

II.         Paragraph 14 is Subject to the Rule of Reason

Plaintiffs cannot avoid their burden to plead and prove a market by insisting on per se analysis or even “quick look” review of Paragraph 14. The rule of reason governs Plaintiffs’ antitrust claims with respect to Paragraph 14 for at least three reasons. First, the restriction was a vertical, intrabrand restraint, not a horizontal one. Second, even if it were horizontal, it was still ancillary to the procompetitive franchise agreement. Third, Paragraph 14 does not fall within that narrow class of restrictions—such as price fixing—universally and historically condemned as anticompetitive. See Nat’l Collegiate Athletic Ass’n v. Alston, 141 S. Ct. 2141, 2155-56 (2021) (explaining “spectrum” of antitrust analysis).

As noted above, McDonald’s vigorously competes with numerous firms in both labor markets and the output market. Its competitive efforts have included various intrabrand restraints among its franchisees that foster a strong, consistent brand identity, along with shared marketing and product development. That successful brand identity is what attracts individual franchisees to open and operate McDonald’s restaurants. Because Paragraph 14 was such a vertical restraint, and ancillary to McDonald’s procompetitive objectives, it cannot be per se unlawful. Rather, it is subject to full rule of reason analysis.

A.        Paragraph 14 is a Vertical, Intrabrand Restraint

Paragraph 14 is a vertical, not horizontal, restraint. It was conceived and imposed by the franchisor, McDonald’s—not competing franchisees.

In 1955, McDonald’s included in its franchise agreement the predecessor to the Paragraph 14 restriction as part of an initial bundle of brand standards. That original franchise agreement also included limits on, among other things, product offerings and territorial exclusivity, as many franchise agreements do. D.E. 380-20. The terms of the agreement were consistent across franchisees, and were designed, insisted upon, and monitored by McDonald’s itself. Paragraph 14 was not created as part of an agreement among horizontal competitors, nationally or in any particular geographic labor market. Indeed, the large national (and subsequently international) network of McDonald’s franchises did not yet exist when the key elements of the franchise agreement were established.

Plaintiffs suggest the restraint was per se unlawful because corporate-owned restaurants, McOpCos, were horizontal competitors with independently owned franchisees. See App. Br. at 25, 44. In those local markets comprising both McOpCos and independently owned franchises, the district court found the restraint horizontal, but ancillary and subject to the rule of reason. Op. at 9. But the district court also identified vertical aspects to the terms that, in fact, obtain in most geographic markets. Id. at 4-5.

Paragraph 14 was necessarily a vertical restraint in the twenty states in which there were no McOpCo restaurants at all. Class Cert. Op. 17. In the remaining states, Paragraph 14 still operated as a vertical restraint in the local labor markets that had only independently owned franchisees or McOpCos, but not both. In other words, it was impossible for the restraint to operate horizontally on a national level, because the putative competition between independent franchisees and McOpCos could not have occurred in the many labor markets in which there were no McOpCos. Id. (finding that Plaintiffs have “not [ ] put forth evidence that McOpCos compete with franchisees in every part of the United States”).

Vertical restraints, like Paragraph 14, are generally evaluated under the rule of reason because they often foster interbrand competition. Thus, for decades, the Supreme Court has whittled down the types of vertical restraints that are subject to per se condemnation. In 1977, the Court refused to extend per se illegality to vertical non-price restraints, noting that vertical restrictions tend to promote interbrand competition, “the primary concern of antitrust law.” Cont’l T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36, 49, 52 n.19, 58 (1977). A decade later, the Court observed that “a rule of per se illegality for vertical nonprice restraints was not needed or effective to protect intra brand competition.” Bus. Elecs. Corp. v. Sharp Elects. Corp., 485 U.S. 717, 725 (1988). And, in 2007, the Court repudiated the prohibition of vertical price restraints that it had adopted nearly a century earlier. See Leegin Creative Leather Prods., Inc. v. PSKS, Inc., 551 U.S. 877 (2007) (overruling Dr. Miles Med. Co. v. John D. Park & Sons Co., 220 U.S. 373 (1911), and subjecting vertical price restraints to rule of reason analysis).

In these decisions, the Court repeatedly emphasized that any departure from the rule of reason “must be based on demonstrable economic effect, rather than . . . upon formalistic line drawing.” Bus. Elecs., 485 U.S. at 724; see also Leegin, 551 U.S. at 889 (applying rule of reason in part because “economics literature is replete with procompetitive justifications for a manufacturer’s use of resale price maintenance”).

While Plaintiffs seek a departure from the rule of reason here, economic research confirms that vertical restraints—including franchisor-franchisee restraints—tend to be procompetitive. Reviewing the empirical and theoretical literature on vertical restraints, Lafontaine and Slade observed that:

[T]he empirical evidence concerning the e?ects of vertical restraints on consumer wellbeing is surprisingly consistent. Speci?cally, it appears that when manufacturers choose to impose such restraints, not only do they make themselves better o? but they also typically allow consumers to bene?t from higher quality products and better service provision.

Francine Lafontaine & Margaret Slade, Exclusive Contracts and Vertical Restraints: Empirical Evidence and Public Policy, 10 Handbook of Antitrust Economics 391, 408-09 (Buccirossi ed., 2008); see also Francine Lafontaine & Margaret E. Slade, Transaction Cost Economics and Vertical Market Restrictions—Evidence, 55(3) The Antitrust Bulletin 587 (2010).

The rule of reason is especially appropriate here because Paragraph 14 is not only a vertical restraint, but an intrabrand one as well. And intra-franchise no-hire agreements are fundamentally different from inter-company restraints for two principal reasons.

First, intra-franchise labor restraints do not restrict output or price in the labor market because they do not affect the ability of alternative employers to compete for workers—whether those employers operate other types of quick-service restaurants or any of the myriad establishments that compete for the same pool of lower-skilled workers.

Second, even if McDonald’s did have the ability to confer labor monopsony power on local franchisees (it does not, see infra), it had no economic incentive to do so. Creating local labor monopsonies would suppress wages in those areas. That would reduce the quantity of labor employed and, in turn, suppress output in the downstream product market (i.e., food sales). See, e.g., Herbert J. Hovenkamp, Worker Welfare and Antitrust, __ U. Chi. L. Rev. 1, 10, 13 (2022) (“[T]he demand for labor as an input is closely correlated with the amount of product or service output that the firm is generating.”). But McDonald’s as a franchisor depends on product output for royalties; it has no desire to reduce its royalties by creating dysfunction in the labor market.

Because Paragraph 14 is a vertical and intrabrand restraint, it must be subject to the rule of reason.

B.        Paragraph 14 is Ancillary to the Franchise Agreement

Paragraph 14 is also subject to scrutiny under the rule of reason (rather than per se or quick look analysis) because it was “ancillary” to a procompetitive franchise agreement. In other words, there was a clear procompetitive rationale for the entire bundle of vertically imposed restraints embodied in the McDonald’s franchise agreement: brand quality and consistency. Paragraph 14 promoted that larger endeavor and was not a naked restraint on trade.

For example, Paragraph 14 limited the ability of individual franchisees to free-ride on training investments by McDonald’s and other franchisees. That in turn encouraged investment in employee development and training, and in the McDonald’s system and brand more generally. Limits on intrabrand employee raiding foster brand consistency and stability, which also are advantageous for the growth of a nationwide brand. See, e.g., Gregory J. Werden, The Ancillary Restraints Doctrine after Dagher, 8 Sedona Conf. J. 17, 21 (2007).

Plaintiffs nonetheless argue that Paragraph 14 cannot be regarded as ancillary because it was not “reasonably necessary” to the procompetitive goals of the franchise agreement, purportedly demonstrated by the fact that the cessation of the restraint in 2017 was not fatal to McDonald’s franchise system. App. Br. at 23. That argument claims too much and shows too little.

As a legal matter, ancillarity does not require a showing that restraints are strictly necessary, but only that they “may contribute to the success of a cooperative venture.” Polk Bros., Inc. v. Forest City Enters., Inc., 776 F.2d 185, 189 (7th Cir. 1985). A restraint is ancillary, in other words, if, “at the time it was adopted,” it bears a reasonable relationship to the joint venture’s success. Id. (emphasis added); see also Major League Baseball Props., Inc. v. Salvino, Inc., 542 F.3d 290, 339-40 (2d Cir. 2008) (Sotomayor, J., concurring). By pointing to the fact that McDonald’s was still able to sign franchisees after 2017, Plaintiffs implicitly argue that a restraint can be ancillary only if its removal destroys the entire endeavor. That is not the law. In other words, a company’s decision to remove one part of a bundle of its practices does not mean the part that was removed never contributed to success of the business. Moreover, to require that businesses precisely calibrate the timing of their policy changes and the substance of those changes would chill the ability of businesses to develop and test new policies and business models. Cf. Werden, supra, at 23-24 (comprehensive analysis by DOJ economist rejecting strict-necessity test).

C.        Paragraph 14 Does not Fall in the Narrow Category of Per Se Unlawful Conduct

Paragraph 14 is not one of those few restrictions—such as price fixing and boycotts—that have been universally and historically condemned as anticompetitive; we are aware of no cases holding that this restraint is per se unlawful, and Plaintiffs cite none. See App. Br. at 25 (citing only inapposite decisions and Arrington v. Burger King Worldwide, Inc., 47 F.4th 1247, 1257 (11th Cir. 2022), which held that a no-hire agreement was “concerted activity” under Section 1 of the Sherman Act but instructed the district court “in the first instance” to determine the level of scrutiny). Accordingly, there is a strong presumption that the rule of reason applies. Bus. Elecs., 485 U.S. at 726; see also Alan J. Meese, In Praise of All or Nothing Dichotomous Categories: Why Antitrust Law Should Reject the Quick Look, 104 Geo. L.J. 835, 878-79 (2016) (“Declaring all horizontal restraints inherently suspect would presumptively condemn all manner of cooperation necessary to allocate resources to their highest valued use, relegating economic actors to cooperation achieved through atomistic interaction in the spot market or complete integration.”).

Plaintiffs do not offer any reason to depart from that standard, nor is there one. As the Supreme Court has held, a departure from rule of reason must be “justified by demonstrable economic effect.” Id. Yet there is a dearth of economic studies on the effects of this type of intrabrand no-hire or no-poach agreement on the labor market. Indeed, the Ashenfelter and Krueger study, on which Plaintiffs’ expert relied, states that “systematic evidence on the impact of no-poaching agreements on workers’ pay and within-franchise job mobility is unavailable.” Ashenfelter & Krueger, supra n.4, at 21. Moreover, to evaluate the impact of no-poach agreements on pay and mobility, one would have to control for a number of important variables—e.g., inter-firm variation in the terms of no-poach agreements, inter-firm variation in the bundling of employment restrictions, and interstate variation in the enforceability of employment restrictions. The only study we are aware of that attempts to do so concludes that elimination of no-poach clauses “causes minimal reductions in job concentration and no increase in wages.” Daniel S. Levy, et al., No-Poaching Clauses, Job Concentration and Wages: A Natural Experiment Generated by a State Attorney General, Advanced Analytical Consulting Group, Inc., at 1 (Jan. 23, 2020).

While the Court has said that a departure from the rule of reason should not be based on “formalistic line drawing,” that is precisely what Plaintiffs (and amici Federal Trade Commission and U.S. Dept. of Justice) propose. For example, Plaintiffs argue that per se treatment is appropriate because (1) in a small number of geographic markets, there are both McOpCos and independently owned franchisees,[5] and (2) interbrand no-poach agreements have been found by some courts analogous to market allocation agreements. App. Br. at 25, 44. But two analogies do not an equivalence make. Even in the limited and atypical markets comprising both McOpCos and independently owned McDonald’s franchises, there is still no evidence of monopsony power over the labor market. As the district court noted regarding Plaintiff Deslandes, alternative employers outnumbered McDonald’s franchises by more than a factor of twenty.

Judicial inexperience and limited economic literature, as well as the facts on the ground, all suggest that this was precisely the type of business practice for which more elaborate economic study is needed before subjecting it to per se condemnation.


For the foregoing reasons, this Court should affirm.


Amici Scholars of Law and Economics

Dirk Auer is the Director of Competition Policy at the International Center for Law & Economics and Adjunct Professor at the University of Liege in Belgium.

Jonathan M. Barnett is the Torrey H. Webb Professor of Law at the University of Southern California Gould School of Law.

James C. Cooper is a Professor of Law at George Mason University School of Law.

Richard A. Epstein is the Laurence A. Tisch Professor of Law at New York University School of Law.

Luke M. Froeb is the William C. Oehmig Chair in Free Enterprise and Entrepreneurship at Vanderbilt University Owen School of Management.

Harold Furchtgott-Roth is a former commissioner of the Federal Communications Commission and a senior fellow at the Hudson Institute.

Daniel J. Gilman is a Senior Scholar at the International Center of Law & Economics.

Janice Hauge is a Professor of Economics at the University of North Texas Department of Economics.

Justin (Gus) Hurwitz is a Professor of Law at the University of Nebraska College of Law.

Stan J. Liebowitz is the Ashbel Smith Professor of Economics at the University of Texas at Dallas.

Abbott (Tad) Lipsky, Jr. is an Adjunct Professor at George Mason University School of Law.

Daniel A. Lyons is a Professor & Associate Dean for Academic Affairs at Boston College Law School.

Geoffrey A. Manne is President and Founder of the International Center for Law & Economics and a Distinguished Fellow at the Northwestern University Center on Law, Business & Economics.

Scott E. Masten is a Professor of Business Economics and Public Policy at the University of Michigan Ross School of Business.

Alan Meese is the Ball Professor of Law and Dean’s Faculty Fellow at William & Mary Law School.

Paul H. Rubin is the Samuel Candler Dobbs Professor of Economic Emeritus at the Emory University Department of Economics and Law School.

Vernon L. Smith is the George L. Argyros Endowed Chair in Finance and Economics at the Chapman University Argyros School of Business of Economics. Professor Smith was awarded the Nobel Memorial Prize in Economic Sciences in 2022.

Michael E. Sykuta is an Associate Professor of Economics at the University of Missouri.

Gregory J. Werden is a retired economist at the U.S. Department of Justice, Antitrust Division.

John M. Yun is an Associate Professor of Law at George Mason University School of Law.

[1] All parties have consented to the filing of this brief. Pursuant to Federal Rule of Appellate Procedure 29(a)(4)(E), counsel for ICLE represents that no counsel for any of the parties authored any portion of this brief and that no entity, other than amicus curiae or its counsel, monetarily contributed to the preparation or submission of this brief.

[2] See Gregory J. Werden, Why (Ever) Define Markets? An Answer to Professor Kaplow, 78 Antitrust L.J. 729, 741 (2013) (“Alleging the relevant market in an antitrust case . . . identifies the competitive process alleged to be harmed.”); Jonathan B. Baker, Market Definition: An Analytical Overview, 74 Antitrust L.J. 129, 129 (2007) (“Market definition is often the most critical step in evaluating market power and determining whether business conduct has or likely will have anticompetitive effects.”).

[3] Cf. Toys “R” Us, Inc. v. FTC, 221 F.3d 928, 937 (7th Cir. 2000) (“direct evidence” suffices where the parties agreed on a nationwide market, and the defendant held “20% of the national wholesale market and up to 49% of some local wholesale markets”).

[4] Dr. Singer misconstrues a key study in the limited economic literature regarding franchise hiring terms. That study—Orley Ashenfelter & Alan B. Krueger, Theory and Evidence on Employer Collusion in the Franchise Sector, IZA Discussion Paper, No. 11672 (July 2018) (cited at Singer Rep. ¶¶ 20, 23)—did not employ a causal design, and it did not establish—nor even purport to establish—that employers generally have or exercise antitrust-relevant market power, as Dr. Singer suggests. Singer Rep. ¶ 23. The study also found widespread use of no-poach terms among many brands that did not enjoy significant market, indicating that market power is not a prerequisite for a franchise to impose no-poach agreements.

[5] FTC and DOJ argue that assessing the horizontality of “employee-allocation agreements in the franchise context . . . requires a fact-bound evaluation of whether the agreement limits rivalry between actual or potential competitors.” DOJ & FTC Br. at 26-27 n.8. But, as noted above, there is undisputed evidence that in several states and local markets there was no competition between independently owned franchisees and McOpCos. Class Cert. Op. 17.

Continue reading
Antitrust & Consumer Protection

ICLE Brief for 9th Circuit in Epic Games v Apple

Amicus Brief In this brief for the 9th U.S. Circuit Court of Appeals, ICLE and 26 distinguished scholars of law & economics argue that the district court in a suit brought by Epic Games rightly found that Apple’s procompetitive justifications outweigh any purported anticompetitive effects in the market for mobile-gaming transactions.

United States Court of Appeals
For the
Ninth Circuit

Plaintiff/Counter-Defendant, Appellant/Cross-Appellee,
Defendant/Counter-Claimant, Appellee/Cross-Appellant

Appeal from a Decision of the United States District Court
for the Northern District of California,
No. 4:20-cv-05640-YGR ? Honorable Yvonne Gonzalez Rogers





The International Center for Law & Economics (“ICLE”) is a nonprofit, non- partisan global research and policy center aimed at building the intellectual foundations for sensible, economically grounded policy. ICLE promotes the use of law & economics methodologies to inform public policy debates and has longstanding expertise in the evaluation of antitrust law and policy.

Amici also include 26 scholars of antitrust, law, and economics at leading universities and research institutions around the world. Their names, titles, and academic affiliations are listed in Addendum A. All have longstanding expertise in, and copious research on, antitrust law and economics.

Amici have an interest in ensuring that antitrust promotes the public interest by remaining grounded in sensible legal rules informed by sound economic analysis. Amici believe that Epic’s arguments deviate from that standard and promote the private interests of slighted competitors at the expense of the public welfare.


Epic challenges Apple’s prohibition of third-party app stores and in-app payments (“IAP”) systems from operating on its proprietary, iOS platform as a violation of the antitrust laws. But, as the district court concluded, Epic’s real concern is its own business interests in the face of Apple’s business model—in particular, the commission charged for the use of Apple’s IAP system. See Order at 1-ER22, Epic Games, Inc. v. Apple Inc., No. 4:20-CV-05640 (N.D. Cal. Sept. 10, 2021), ECF No. 812 (1-ER3–183). In essence, Epic is trying to recast its objection to Apple’s 30% commission for use of Apple’s optional IAP system as a harm to consumers and competition more broadly.

Epic takes issue with the district court’s proper consideration of Apple’s procompetitive justifications and its finding that those justifications outweigh any anticompetitive effects of Apple’s business model. But Epic’s case fails at step one of the rule of reason analysis. Indeed, Epic did not demonstrate that Apple’s app distribution and IAP practices caused the significant market-wide effects that the Supreme Court in Ohio v. Am. Express Co. (“Amex”) deemed necessary to show anticompetitive harm in cases involving two-sided transaction markets. 138 S. Ct. 2274, 2285–86 (2018). While the district court found that Epic demonstrated some anticompetitive effects, Epic’s arguments below focused only on the effects that Apple’s conduct had on certain app developers and failed to appropriately examine whether consumers were harmed overall. This is fatal. Without further evidence of the effect of Apple’s app distribution and IAP practices on consumers, no conclusions can be reached about the competitive effects of Apple’s conduct.

Nor can an appropriate examination of anticompetitive effects ignore output. It is critical to consider whether the challenged app distribution and IAP practices reduce output of market-wide app transactions. Yet Epic did not seriously challenge that output increased by every measure, and Epic’s Amici ignore output altogether.

Moreover, the district court examined the one-sided anticompetitive harms presented by Epic, but rightly found that Apple’s procompetitive justifications outweigh any purported anticompetitive effects in the market for mobile gaming transactions. The court recognized that the development and maintenance of a closed iOS system and Apple’s control over IAP confers enormous benefits on users and app developers.

Finally, Epic’s reliance on the theoretical existence of less restrictive alternatives (“LRA”) to Apple’s business model is misplaced. Forcing Apple to adopt the “open” platform that Epic champions would reduce interbrand competition, and improperly permit antitrust plaintiffs to commandeer the judiciary to modify routine business conduct any time a plaintiff’s attorney or district court can imagine a less restrictive version of a challenged practice, irrespective of whether the practice promotes consumer welfare. See NCAA v. Alston, 141 S. Ct. 2141, 2161 (2021) (“[C]ourts should not second-guess ‘degrees of reasonable necessity’ so that ‘the lawfulness of conduct turn[s] upon judgments of degrees of e?ciency.’”). Particularly in the context of two-sided platform businesses, such an approach would sacrifice interbrand, systems-level competition for the sake of a superficial increase in competition among a small subset of platform users.

Read the full brief here.

Continue reading
Antitrust & Consumer Protection

ICLE Brief for D.C. Circuit in State of New York v Facebook

Amicus Brief In this amicus brief for the U.S. Court of Appeals for the D.C. Circuit, ICLE and a dozen scholars of law & economics address the broad consensus disfavoring how New York and other states seek to apply the “unilateral refusal to deal” doctrine in an antitrust case against Facebook.

United States Court of Appeals
for the District of Columbia Circuit


No. 1:20-cv-03589-JEB (Hon. James E. Boasberg)




Amici are leading scholars of economics, telecommunications, and/or antitrust. Their scholarship reflects years of experience and publications in these fields.

Amici’s expertise and academic perspectives will aid the Court in deciding whether to affirm in three respects. First, amici provide an explanation of key economic concepts underpinning how economists understand the welfare effects of a monopolist’s refusal to deal voluntarily with a competitor and why that supports affirmance here. Second, amici offer their perspective on the limited circumstances that might justify penalizing a monopolist’s unilateral refusal to deal—and why this case is not one of them. Third, amici explain why the District Court’s legal framework was correct and why a clear standard is necessary when analyzing alleged refusals to deal.


This brief addresses the broad consensus in the academic literature disfavoring a theory underlying plaintiff’s case—“unilateral refusal to deal” doctrine. The States allege that Facebook restricted access to an input (Facebook’s Platform) in order to prevent third parties from using that access to export Facebook data to competitors or compete directly with Facebook. But a unilateral refusal to deal involves more than an allegation that a monopolist refuses to enter into a business relationship with a rival.

Mainstream economists and competition law scholars are skeptical of imposing liability, even on a monopolist, based solely on its choice of business partners. The freedom of firms to choose their business partners is a fundamental tenet of the free market economy, and the mechanism by which markets produce the greatest welfare gains. Thus, cases compelling business dealings should be confined to particularly delineated circumstances.

In Part I below, amici describe why it is generally inefficient for courts to compel economic actors to deal with one another. Such “solutions” are generally unsound in theory and unworkable in practice, in that they ask judges to operate as regulators over the defendant’s business.

In Part II, amici explain why Aspen Skiing—the Supreme Court’s most prominent precedent permitting liability for a monopolist’s unilateral refusal to deal—went too far and should not be expanded as the States’ and some of their amici propose.

In Part III, amici explain that the District Court correctly held that the conduct at issue here does not constitute a refusal to deal under Aspen Skiing. A unilateral refusal to deal should trigger antitrust liability only where a monopolist turns down more profitable dealings with a competitor in an effort to drive that competitor’s exit or to disable its ability to compete, thereby allowing the monopolist to recoup its losses by increasing prices in the future. But the States’ allegations do not describe that scenario.

In Part IV, amici address that the District Court properly considered and dismissed the States’ “conditional dealing” argument. The States’ allegations are correctly addressed under the rubric of a refusal to deal—not exclusive dealing or otherwise. The States’ desire to mold their allegations into different legal theories highlights why courts should use a strict, clear standard to analyze refusals to deal.

Read the full brief here.

Continue reading
Antitrust & Consumer Protection

Chairman Pai’s Legacy of Transparency

TOTM For many, the chairmanship of Ajit Pai is notable for its many headline-grabbing substantive achievements, including the Restoring Internet Freedom order, 5G deployment, and rural . . .

For many, the chairmanship of Ajit Pai is notable for its many headline-grabbing substantive achievements, including the Restoring Internet Freedom order, 5G deployment, and rural buildout—many of which have been or will be discussed in this symposium. But that conversation is incomplete without also acknowledging Pai’s careful attention to the basic blocking and tackling of running a telecom agency. The last four years at the Federal Communications Commission were marked by small but significant improvements in how the commission functions, and few are more important than the chairman’s commitment to transparency.

Read the full piece here.

Continue reading
Telecommunications & Regulated Utilities

ICLE Ninth Circuit Amicus Brief in FTC v. Qualcomm

Amicus Brief INTRODUCTION The district court’s decision is disconnected from the underlying economics of the case. It improperly applied antitrust doctrine to the facts, and the result . . .


The district court’s decision is disconnected from the underlying economics of the case. It improperly applied antitrust doctrine to the facts, and the result subverts the economic rationale guiding monopolization jurisprudence. The decision—if it stands—will undercut the competitive values antitrust law was designed to protect.

Antitrust law should seek to minimize error and decision costs to maximize consumer welfare and reduce the likelihood of self-defeating antitrust interventions. See Frank H. Easterbrook, The Limits of Antitrust, 63 Tex. L. Rev. 1 (1984). The Supreme Court has thoroughly incorporated the economic logic of this “error cost” framework into its antitrust jurisprudence. See Ohio v. Am. Express Co., 138 S. Ct. 2274, 2287 (2018) (“Any other analysis would lead to ‘mistaken inferences’ of the kind that could ‘chill the very conduct the antitrust laws are designed to protect.’ ”) (quoting Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 226 (1993)); see also Thomas A. Lambert & Alden F. Abbott, Recognizing the Limits of Antitrust: The Roberts Court Versus the Enforcement Agencies, 11 J. Competition L. & Econ. 791 (2015).

In contrast, this case is a prime—and potentially disastrous— example of how the unwarranted reliance on inadequate inferences of anticompetitive effect lead to judicial outcomes utterly at odds with Supreme Court precedent.

The district court’s decision confuses several interrelated theories of harm resting on the central premise that Qualcomm’s business model is purposefully structured to preserve its ability to license its standard essential patents (SEPs) to device makers (OEMs) at “unreasonably high royalty rates,” thus “impos[ing] an artificial surcharge on all sales of its rivals’ modem chips,” which “reduces rivals’ margins, and results in exclusivity.” FTC v. Qualcomm Inc., No. 17-CV-00220-LHK, 2019 WL 2206013, slip op. at 183 (N.D. Cal. May 21, 2019) (hereinafter slip op.).

But, without more, high royalty rates, artificial surcharges, the reduction of rivals’ margins, and even exclusivity do not violate the Sherman Act. Indeed, high prices are as likely the consequence of the lawful exercise of monopoly power or the procompetitive offering of higher quality products, and harm to competitors is a hallmark of vigorous competition. See, e.g., Verizon Commc’ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004) (“The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system.”). Avoiding the wrongful condemnation of such conduct is precisely the point of the Court’s error cost holdings.

The district court commits several key errors inconsistent with both Supreme Court precedent and its underlying economic framework.

First, the court failed to require proof of the anticompetitive harm allegedly caused by Qualcomm’s conduct. Instead, the court infers both its existence and its cause, see slip op. at 42–43, justifying its approach with reference to a single case: United States v. Microsoft, 253 F.3d 34, 79 (D.C. Cir. 2001) (“We may infer causation when exclusionary conduct is aimed at producers of nascent competitive technologies as well as when it is aimed at producers of established substitutes.”).

But the court misreads Microsoft and disregards contrary Supreme Court precedent. Indeed, both the Court and Microsoft made clear that a finding of illegal monopolization may not rest on an inference of anticompetitive harm.

In Brooke Group, the Court took the unusual step of reviewing an appellate decision for the sufficiency of evidence, prodded by the need to protect against the costs of erroneously condemning procompetitive conduct. See 509 U.S. at 230. It held that only evidence defendant’s conduct injured “competition, not competitors” supports a monopolization claim. Id. at 224 (citation omitted). And because harm to competitors doesn’t necessarily mean harm to competition, inferring anticompetitive harm from such evidence would not suffice: “mistaken inferences are especially costly, because they chill the very conduct the antitrust laws are designed to protect.” Id. at 226 (citation omitted).

In subsequent cases, the Court redoubled its commitment to minimizing error costs arising from erroneous inferences of anticompetitive effect. See Trinko, 540 U.S. at 414 (“The cost of false positives counsels against an undue expansion of § 2 liability.”) (citation omitted); Pac. Bell Tel. Co. v. linkLine Commc’ns, Inc., 555 U.S. 438, 451 (2009).

As law and economics scholars, we are concerned that, because the district court’s decision rests on tenuous, unsupported inferences, “[i]f the district court’s holding is not repudiated on appeal, then the obvious consequence will be for companies to be deterred from much innocent and potentially procompetitive business conduct.” Douglas H. Ginsburg, Joshua D. Wright & Lindsey M. Edwards, Section 2 Mangled: FTC v. Qualcomm on the Duty to Deal, Price Squeezes, and Exclusive Dealing 2 (George Mason Univ. Law & Econ. Research Paper Series 19-21, Aug. 19, 2019),

This concern is not just academic. See FTC v. Qualcomm, No. 19- 16122, Order at 6 (9th Cir. Aug. 23, 2019) (recognizing the DOJ and Departments of Energy and Defense all classified this decision as a costly false positive).

Second, the court erred in finding Qualcomm had an antitrust duty to deal with rivals. The evidence adduced could sustain the district court’s ruling through only one theory: an illegal unilateral refusal to deal.2 See Aspen Skiing Co. v. Aspen Highland Skiing Corp., 472 U.S. 585 (1985)). But this narrow exception—“at or near the outer boundary of § 2 liability,” Trinko, 540 U.S. at 409—is subject to strict limitations.

Finding a duty to deal requires that the company gave up a profitable course of dealing with rivals and adopted a less profitable alternative. The evidence before the district court uniformly shows that Qualcomm’s challenged practices were more profitable, and thus insufficient to support an antitrust duty to deal.

Finally, because the court didn’t perform a competitive effects analysis, it failed to demonstrate the “substantial” foreclosure of competition required to sustain a claim of anticompetitive exclusion. To avoid the costs of mistaken condemnation, the Court placed tight guardrails around finding exclusionary conduct anticompetitive, requiring foreclosure of “a substantial share of the relevant market.” See Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320, 328 (1961). Without this finding, which also may not be inferred, a claim of anticompetitive foreclosure is unsupportable.

In sum, the district court’s approach extends antitrust law beyond the clear boundaries imposed by the Supreme Court and risks deterring significant pro-competitive conduct. If upheld, amici anticipate significant harm from the district court’s decision.

Continue reading
Antitrust & Consumer Protection

Amicus Brief, Mozilla v. FCC

Amicus Brief ICLE filed a  brief in support of Petitioners in the D.C. Circuit case, Mozilla v. FCC, a case that challenged the FCC's authority to issue the Restoring Internet Freedom Order ("RIFO").


ICLE filed a  brief in support of Petitioners in the D.C. Circuit case, Mozilla v. FCC, a case that challenged the FCC’s authority to issue the Restoring Internet Freedom Order (“RIFO”). In RIFO, the FCC repealed the Title II classification on ISPs, preempted conflicting state laws, and applied a transparency rule against ISPs, among other provisions. In our brief, we argue that:

Contrary to the claims of Petitioners, the Commission acted well within its authority in adopting the Order. The Commission developed a comprehensive regulatory scheme for ISPs that includes both obligations imposed under the Communications Act, as well as complementary regulation and potential enforcement under antitrust law by the Commission’s sister agencies. As we show below, this competition-oriented, light touch regulatory regime comports with the relevant provisions and stated goals of the Communications Act far better than the ex ante rules adopted in the Title II Order.

In adopting this competition-oriented regulatory regime, the Commission also acted within its authority to preempt contradictory state laws under well- established precedent. The Commission did so while properly allowing for states to continue to regulate under other laws of general applicability that do not conflict with or frustrate the federal policies underlying the Order.

Accordingly, the Order should be upheld and the petitions for review should be denied.

Signatories on the Brief

  • Michelle Connolly
    Professor of Economics
    Duke University
    Former chief economist, FCC
  • Janice A. Hauge
    Professor, Department of Economics
    University of North Texas
  • Justin (Gus) Hurwitz
    Director of Law & Economics Programs
    International Center for Law & Economics
    Associate Professor of Law And Co-Director of Space,
    Cyber, and Telecom Law Program
    Nebraska College of Law
  • Mark A. Jamison
    Director and Gunter Professor, Public Utility Research Center
    University of Florida
  • Stan Liebowitz
    Ashbel Smith Professor of Managerial Economics
    University of Texas at Dallas
  • Daniel A. Lyons
    Associate Professor of Law
    Boston College Law School
  • Geoffrey A. Manne
    President and Founder
    International Center for Law & Economics
  • Michael Sykuta
    Associate Professor, Applied Social Sciences
    University of Missouri – Columbia
Continue reading
Telecommunications & Regulated Utilities