The majority of private sector, nonunion workers and e-commerce transactions are subject to arbitration agreements, which require litigating disputes in private, often confidential, proceedings, rather than in courtrooms.1 Paired with waivers of the right to bring class actions, arbitration agreements often suppress claims altogether.2 A series of Supreme Court decisions upholding arbitration of virtually all consumer and employment law claims pursuant to the Federal Arbitration Act3 (FAA) and fortifying the statute through a sweeping preemption regime have allowed arbitration clauses to flourish.4
This Note proposes that states and municipalities seeking to limit the impacts of forced arbitration turn to an alternative route that circumvents preemption altogether: the market participant doctrine. Under the market participant doctrine, when states act pursuant to proprietary rather than regulatory interests — such as when purchasing goods and services on the market — their actions are not subject to preemption absent a clear statement of congressional intent.5 Limiting the use of forced arbitration against the employees who deliver government services and the people who consume them aligns neatly with states’ proprietary interests in ensuring business partners provide safe, efficient, and lawful service delivery. This Note thus renews and builds on calls for experimentation in state procurement initially proposed by the National Consumer Law Center in 2015.6
Part I provides a brief overview of how the Supreme Court’s broad interpretation of the FAA has frustrated state efforts against forced arbitration.7 Part II traces the origins of the market participant exception, from its birth in the dormant commerce clause cases to its application in the National Labor Relations Act8 (NLRA) context.9 Though the Court has never considered the doctrine’s application to statutes other than the NLRA, circuit courts have applied it to various other federal statutes.
Part III argues that the market participant exception, rooted in general principles of federalism, should also apply to the FAA. Labor scholars might demur, however, that this exception is of limited utility in light of its narrowness in the NLRA context.10 This Note further argues that, because the FAA does not create a highly centralized regulatory regime like the NLRA does, market action should be defined more broadly under the FAA in order to permit states greater flexibility.
Part IV offers two case studies for how states can develop arbitration-limiting procurement requirements appropriately tailored to proprietary needs. First, transportation gig work in state contracts offers an archetype of a public-private partnership in which limiting forced arbitration ensures safe and lawful service delivery. Second, much like private entities that excised arbitration clauses under public pressure during the #MeToo movement, states, too, should be free to prohibit service providers from imposing arbitration against consumers and employees in analogous contexts, such as civil rights claims.
Absent congressional action, few viable pathways remain for states to limit forced arbitration. But states that seek to protect employees and consumers have a potentially powerful tool in the market participant doctrine, with valuable precedent and lessons from its application to other federal statutes. In the face of the Supreme Court’s growing skepticism toward broad preemption regimes,11 the market participant exception to the FAA is emerging as a constitutionally rooted, politically feasible tool that states should experiment with to continue testing the boundaries of the FAA preemption regime.
I. The Rise and Rise of the FAA
The FAA states that arbitration agreements “shall be valid, irrevocable, and enforceable, save upon such grounds as exist at law or in equity for the revocation of any contract.”12 In 1925, Congress passed the FAA to ensure enforcement of arbitration agreements between merchants and facilitate efficacious dispute resolution outside of a purportedly crowded court system.13 Corporate proponents emphasized two key limitations in congressional hearings14: first, the Act was intended to apply only in federal courts, not state courts; and second, the Act did not apply to employment contracts.15 These limitations, however, were not embodied in the text of the statute because Congress did not foresee how dramatically the Commerce Clause would expand post–New Deal.16
But in the 1980s, the Court disregarded these limitations and began expanding the FAA well beyond its intended breadth to declare the statute a “liberal federal policy favoring arbitration.”17 Although the statute was intended to apply to “merchant-to-merchant arbitrations, never merchant-to-consumer arbitrations,”18 the Court ignored the FAA’s pre–New Deal context and upheld arbitration clauses even in civil rights19 and employment20 cases. The Court also ignored congressional intent that the FAA should apply only in federal court — as Justice Thomas has repeatedly criticized in his dissents.21
While expanding the FAA’s substantive reach, the Court also built an airtight preemption regime. In Southland Corp. v. Keating,22 Chief Justice Burger held that a California law prohibiting franchise agreements from imposing arbitration was preempted because it frustrated the FAA’s purpose.23 The Court then began preempting any state policies that “place arbitration clauses on an unequal ‘footing’” with other contract clauses.24 While Justice Scalia initially criticized Southland as an “unauthorized eviction of state-court power,”25 he later delivered a significant blow to states by upholding waivers of class actions, which prohibit claim aggregation in favor of one-on-one arbitration26 — even when the cost of bringing an individual claim far outweighs the potential recovery.27 Under the Court’s precedent, then, an arbitration agreement may deprive a plaintiff of a remedy altogether, since “[t]he realistic alternative to a class action is not 17 million individual suits, but zero individual suits, as only a lunatic or a fanatic sues for $30.”28
Empirics belie the Court’s claim that arbitration is merely a more “efficient, streamlined” option for dispute resolution.29 From 2010 to 2012, the Consumer Financial Protection Bureau estimated that consumers only initiated an average of 411 arbitration claims per year, and of these, around 25 claims a year sought $1,000 or less — compared to the more than 30 million Americans eligible for class action relief per year.30 In employment, arbitration clauses are estimated to have eliminated up to 98 percent of claims altogether.31 Arbitration, then, is not so much about efficiency as it is about evading accountability.32 There is no empirical evidence to suggest that arbitration’s purported efficiencies have passed cost savings to consumers, either.33
Even if a plaintiff successfully files for arbitration, disturbing democratic legitimacy concerns remain.34 Arbitral fora may have limited rules of discovery and evidence, minimal appeals processes, and potentially biased private arbitrators,35 and the private, confidential setting enables corporations to hide wrongdoing, dodge public accountability, and isolate plaintiffs.36 Arbitral decisions do not set legal precedent, threatening to signal “the end of law” altogether in consumer protection and employment.37 The social cost is further compounded through the loss of the expressive, disciplining function of public adjudication.38
The Court brushes these criticisms aside as “generalized attacks on arbitration” better directed to Congress.39 Corporations, meanwhile, have taken advantage of this wall of precedent: the share of private-sector, nonunion workers subject to arbitration grew from 2 percent in 1992 to nearly 60 percent in 2018 and is projected to reach 80 percent this year.40 That arbitration agreements enable corporations to steal wealth with relative impunity41 is all the more disturbing when juxtaposed with data showing they are more common among low-wage workers, and thus disproportionately target women and African Americans.42
State courts and legislatures have tried and failed to limit arbitration clauses by imposing disclosure requirements,43 permitting workers to reject clauses without fear of job loss,44 and declaring the pairing of forced arbitration with class waivers unconscionable.45 Few paths remain: states may use their own finite resources to bring suit,46 regulate the procedure of arbitral fora,47 and protect workers outside the FAA’s reach.48 But under the heavy burden of precedent, states have limited power to pass affirmative laws limiting the reach of arbitration. This invites the question: is there a way to circumvent preemption altogether?
II. The Origins of the Market Participant Exception
The market participant exception holds that preemption applies only to state regulatory action — unless Congress manifests a clear intent to preempt state proprietary action.49 This doctrine first arose in the dormant commerce clause context in an era of limited Commerce Clause power. But even as Congress’s powers under the Commerce Clause expanded, the clear statement requirement to preempt state market action has persisted in the Supreme Court’s NLRA precedent. Circuit courts, taking cue from the federalism principles in which the exception is rooted, have applied the doctrine to various other federal statutes.
A. Dormant Commerce Clause Roots
The Supreme Court first articulated the market participant doctrine in a series of dormant commerce clause cases in the 1970s and 1980s. The dormant commerce clause restrains states from enacting economic regulations that discriminate against or burden interstate commerce absent congressional approval.50 In Hughes v. Alexandria Scrap Corp.,51 the Court considered a Maryland statute that doled out a subsidy to in-state processors for hauling junked cars but made it significantly more difficult for out-of-state processors to access the subsidy.52 The Court upheld the statute, despite its pro-Maryland tilt, because the state itself had entered the scrap market as “a purchaser, in effect, of a potential article of interstate commerce.”53
The Court later refined this doctrine to a succinct principle: a state participating in a market, as opposed to regulating it, is free from the Commerce Clause’s negative implications because “[t]here is no indication of a constitutional plan to limit the ability of the States themselves to operate freely in the free market.”54 Under this exception, the Court upheld a South Dakota policy to sell state-owned cement only to state residents55 and a Boston requirement that publicly funded construction projects employ a certain share of bona fide city residents.56
The market participant exception faced a shaky start. Justices divided sharply over the doctrine’s limits.57 The limits of the Commerce Clause, too, were in flux. On the same day that Hughes expanded state power, the Court limited federal power by declaring in National League of Cities v. Usery58 that Congress lacked the power to mandate a minimum wage for state employees engaged in “traditional government functions.”59 Usery was overturned only nine years later60 — but the market participant exception remained.61
B. Shift to Federalism in NLRA Case Law
Shortly after Garcia was decided, the Court made clear that the market participant exception not only survived the overturning of Usery but functioned as a general principle that a state does not offend “preemption principles” when acting as a proprietor.62 In the following twenty years, the Court clarified the hazy line between regulatory and proprietary action as applied to the NLRA.
The NLRA is the only context in which the Court has considered whether the market participant exception applies to affirmative federal legislation. In Wisconsin Department of Industrial, Labor & Human Relations v. Gould,63 a Wisconsin statute prohibited the state from doing business with entities that had violated the NLRA at least three times within five years.64 Wisconsin argued preemption did not apply because its bill was a proprietary exercise of state spending power, and, referencing the dormant commerce clause cases, “evenhandedness as a matter of federalism . . . requires that states operate as free from federal constraint as a private company.”65 A unanimous Court hinted that market participation might possibly escape NLRA preemption, while also cautioning that “government occupies a unique position of power” and cannot be treated exactly the same as a private party.66 Regardless, the statute at hand could not “plausibly be defended as a legitimate response to state procurement constraints or to local economic needs,”67 as Wisconsin conceded that it had a regulatory purpose of enforcing labor law — a role strictly reserved for the National Labor Relations Board (NLRB) as “a comprehensive regulat[or] of industrial relations.”68
The Court definitively confirmed that the market participant principle applies to the NLRA in Building & Construction Trades Council v. Associated Builders and Contractors of Massachusetts/Rhode Island, Inc.69 (Boston Harbor), which upheld a Massachusetts procurement requirement that all contractors on Boston’s $6.1 billion, ten-year harbor cleanup project sign a labor peace agreement requiring all work be done under a union contract featuring a no-strike clause.70 The Court emphasized that “pre-emption doctrines apply only to state regulation”71 — and state proprietary action “with no interest in setting policy” is not regulation at all.72 As evidence that the bid specification was more proprietary than regulatory, the Court agreed it was tailored to a particular job (albeit a very large one), with an interest in completing the job “quickly and effectively.”73
On the other hand, the Court rejected that a California statute prohibiting private recipients of state funds from using those funds to encourage or prevent union organizing was proprietary rather than regulatory.74 The state’s argument that it merely sought to promote neutrality with regards to unionization via a narrow use restriction was unavailing.75 The strict reporting conditions, civil liability for commingling funds, and carve-outs for certain pro-union activities all indicated that the statute reached beyond California’s sovereign, proprietary interest to instead impose punitive sanctions — much like the Wisconsin statute in Gould.76
Put together, the cases impart three lessons for distinguishing regulatory and proprietary action. First, the determination is functional, not formalist: leveraging police power versus spending power or use versus receipt restrictions is irrelevant. Second, states must demonstrate a credible nexus between their actions and a proprietary purpose tailored to issues they face in the relevant market; blanket legislation that applies across markets may be more difficult to defend as proprietary.77 Third, although the Court does not accept the state’s stated purpose at face value, the focus is on purpose rather than motivation. Put differently, Boston’s bid specification may have been motivated by a pro-union tilt, but its stated purpose of labor efficiency survived the Court’s scrutiny because its actions were narrowly tailored to that purported purpose.78 In the nearly forty years since it began applying the market participant exception to the NLRA, the Court has not considered its import to other federal legislation.
C. The Doctrine in the Circuit Courts
Taking cue from the Court’s statement that “[i]n the absence of any express or implied indication by Congress that a State may not manage its own property when it pursues its purely proprietary interests, and where analogous private conduct would be permitted, this Court will not infer such a restriction,”79 circuit courts have applied the market participant doctrine broadly. The Sixth and Ninth Circuits have definitively held the exception applies to the Employee Retirement Income Security Act of 197480 (ERISA), while the Third Circuit has assumed as much without deciding.81 Circuit courts have also applied the exception to the Telecommunications Act of 199682 and the Federal Aviation Administration Authorization Act of 1994,83 though they have never considered arguments that the exception should shield against FAA preemption.
To determine whether state action is proprietary, circuit courts leverage the two-part framework from Cardinal Towing & Auto Repair, Inc. v. City of Bedford.84 First, the state action must “essentially reflect the [governmental] entity’s own interest in its efficient procurement of needed goods and services, as measured by comparison with the typical behavior of private parties in similar circumstances.”85 Second, the “narrow scope of the challenged action” must “defeat an inference that its primary goal was to encourage a general policy rather than address a specific proprietary problem.”86
States can offer an array of evidence to respond to these inquiries. On the first prong, states might point to private parties that do, in fact, leverage similar contracting requirements.87 This is not a strict requirement, though, as courts consider, hypothetically, whether a private entity might invoke similar interests should it enter an analogous private contract.88 States may also offer expert testimony regarding the proprietary benefits they anticipate such requirements will offer.89 For example, even in Boston Harbor, where the state made only general assertions about convenience and lacked record-specific findings that a labor peace agreement would result in greater efficiency, the Supreme Court nonetheless upheld the requirement.90 And states need not get a “good deal” out of the particular procurement requirement, as this would unduly “impose economic rationality on state contracting decisions.”91
The Cardinal Towing test is applied with varying rigor. The Sixth Circuit, for example, found a Cincinnati ordinance preferencing “contractors who provide health care and retirement benefits” was proprietary because such contractors may be “less likely to experience significant employee turnover, improving the stability and overall quality of a project.”92 A court taking a harder look might have found this connection between benefits and project quality as somewhat tenuous. In the Sixth and Ninth Circuits, satisfying either prong of Cardinal Towing is sufficient, whereas the Fifth Circuit examines both prongs.93 And other circuits may recite the Cardinal Towing test as more of a general principle without explicitly considering whether each individual prong has been satisfied.94
If the state is found to be acting as a market participant, then the court turns to the broader legal question of legislative intent: Did Congress intend that the federal statute at issue preempt state proprietary action, looking to statutory text and purpose?95 Circuit courts have found that express statutory provisions claiming to preempt “[s]tate laws” that “hav[e] the effect of law” do not apply to proprietary action — since such action by definition lacks the force of law.96
A particularly pertinent example of the market participant exception arose in a case concerning the Clean Air Act97 — and it is also the closest the Supreme Court has come to acknowledging the doctrine’s application outside the NLRA. A California agency mandated that vehicle fleet operators comply with certain stringent emissions requirements, and diesel-fueled engine manufacturers challenged the rules as preempted by the Clean Air Act.98 The Supreme Court rejected the state’s argument that the rules were outside the statute’s scope altogether — but writing for an 8–1 Court, Justice Scalia cautioned that this did not necessarily mean the state regulations were “pre-empted in toto”99: the district court was to consider arguments not presented, including “whether some of the Fleet Rules (or some applications of them) can be characterized as internal state purchase decisions (and, if so, whether a different standard for preemption applies).”100
On remand, California argued its actions were not subject to preemption under the market participant exception — and the Ninth Circuit agreed, insofar as applied to “the procurement behavior of state and local government entities.”101 Even after the Supreme Court clarified the Clean Air Act’s wide preemptive scope, California, like UPS and FedEx — both of which the court analogized to as private entities taking measures to reduce pollution — could set purchasing requirements, circumventing preemption.102