Skip to content

Eight Reactions to the FTC’s Proposed Ban on Non-competes


Suresh Naidu (@snaidunl) is Professor of Economics and Public Affairs at Columbia University.

Catherine Fisk (@CatherineFisk1) is the Barbara Nachtrieb Armstrong Professor of Law at U.C. Berkeley School of Law.

David Seligman (@daveyseligman) is the Executive Director of Towards Justice.

Leah Samuel (@lmsamuel777) is a recent graduate of Yale Law School.

Sandeep Vaheesan (@sandeepvaheesan) is the legal director at the Open Markets Institute, and author of Democracy in Power: A History of Electrification in the United States.

Sanjukta Paul (@sanjuktampaul) is Professor of Law at Michigan Law School.

Najah A. Farley (@NajahFarley) is a senior staff attorney at the National Employment Law Project (NELP).

Basel Musharbash (@musharbash_b) is the Founder & Managing Attorney at Basel PLLC, an antimonopoly and community development law firm.

Earlier this month, the FTC proposed a new rule that would ban non-compete clauses in labor contracts. This rule, which aligns with the agency’s renewed commitment to enforce the federal ban on unfair methods of competition, has received ample coverage and support in the mainstream press. Yet it also raises a host of questions – about the relationship between labor law and antitrust, about the legal basis of the FTC’s rulemaking authority, about the enforcement of such a ban, and much else – that have received far less attention. To begin to sort through these issues, we invited eight friends of the blog to share their initial reactions to the proposed rule and to reflect on what might be missing from the current discussion.

Suresh Naidu

The FTC has been rightfully lauded for this announcement. Those worried about the exercise of private power in the labor market, like labor unions and their allies, can cheer the loosening of bonds to particular employers. Those attached to economic welfare can celebrate potential increases in output, employment and wages. Everybody, except maybe overzealous human resource managers, can leave feeling happy.

But behind the no-non-compete consensus lies an unsettling division. The rule comes as a blanket ban from the FTC, under the aegis of increasing competition. In a recent WSJ op-ed, Chair Khan reiterated that antitrust law stipulated preserving competition, regardless of what other sources of countervailing power exist (e.g., union neutrality).

Competition itself should not be the primary desiderata for the labor market. For one thing, employers have considerable power even in decentralized, competitive markets, so the scalpel of antitrust is poorly suited, compared to either labor law or macroeconomic policy, to tackle the bulk of the monopsony power in the labor market. Regulators can police the behavior of large employers with bureaucratic modes of worker control, but have a harder time neutralizing the diffuse workplace despotism of many small businesses.

Second, competition is often the source of many labor market harms (including, historically, racial conflict), and can easily destroy existing norms of fairness and networks of solidarity, as was recognized by the Clayton act. Other sources of countervailing power can be broken by an activist antitrust regime exclusively focused on competition as the pre-eminent property of functioning markets.

Finally, by relying too much on competition, we forgo other, deeper and more democratic, principles that could undergird an expansive notion of economic non-domination. Ironically, owing to our history of coerced labor, American law has deep resources for fighting unfree labor contracts. One of the earliest uses of the 13th Amendment came with the 1867 anti-peonage statute, which outlawed the widespread indentured servitude in the newly added New Mexico territory. We could have gone after non-competes under the banner of “no contractual surrender of liberty,” encompassing binding arbitration, non-disclosure, no-strike clauses, and yellow-dog contracts as well as non-competes, rather than the narrower mandate of “more competition.” Although labor rights generally have been won as forms of economic regulation, rather than as entrenched freedoms, there’s no reason to concede to that. And in the context of wider interest in re-thinking the Constitution and ideas of economic freedom, it might be time to bring those ideas back to the fore as a way of thinking about the labor market generally.

Catherine Fisk

Agreements not to compete are just that—anti-competitive. It thus is squarely within the authority of the FTC to ban them.

The use of non-compete agreements to restrain workers from taking a job in the same industry as a previous employer is of relatively recent vintage. Courts across the United States generally treated workplace knowledge as an attribute of the worker, rather than as property of the employer, and rarely enforced agreements not to compete against employees before the early twentieth century. They were deemed illegal efforts to restrain competition, a hindrance on economic growth, and an unjust restriction on the ability of workers to profit from their skill. A frequently cited English case of 1711, Mitchel v. Reynolds, explained the policy against using them: they cause “mischief, 1st to the party, by the loss of his livelihood and the subsistence of his family; 2ndly, to the publick, by depriving it of a useful member.” They are, the court continued, subject to abuse “from corporations, who are perpetually labouring for exclusive advantages in trade.” For these reasons, California has banned non-competes for employees since 1872.

This history suggests that the Federal Trade Commission’s broad statutory authority to prevent “unfair” methods of competition, which supplements the prohibition on agreements in restraint of trade of the Sherman Act of 1890, extends to banning non-compete agreements. While business groups that use non-competes to keep labor costs down may be expected to challenge the FTC rule on the grounds that it is novel and exceeds the FTC’s statutory authority, if courts let their interpretation be guided by history, they should uphold the rule.

David Seligman

Especially in the face of a tight labor market, the FTC’s proposed rule, on its own, is unlikely to put a stop to unfair employer efforts to strip workers of a critical foundation of their bargaining power—the leverage of being able to seek out work that will pay them more or treat them better. Even in states like California, which have long held traditional non-compete agreements to be unenforceable and illegal, around nineteen percent of workers are required to sign illegal agreements as a condition of doing their jobs. Other employers have consciously shifted away from non-competes to other coercive contracts, like the use of predatory employer-driven debt, including so-called Training Repayment Agreement Provisions (TRAPs), that threaten workers with financial penalties if they leave their jobs. 

To deter employers from skirting the law through agreements that act like non-competes, the FTC’s rule should set even clearer prohibitions against employer-driven debt and no-hire agreements between employers. While the rule itself bans certain de-facto non-competes, including the requirement to repay training costs not “reasonably related” to the employer’s costs, the rule should be expanded to cover other forms of employer-driven debt and clarified to prohibit debts that are plausibly related to an employer’s purported costs, especially where those purported costs are for the employer’s benefit. It is the employer’s responsibility to train their workers, and they should compete fairly to retain workers after they are trained through decent pay and treatment, not through debts that coerce continued employment. To be sure, the DOJ, CFPB, and DOL may all have authority to address those problems, but they also clearly fall within the FTC’s Section 5 authority. 

We’ll also need real enforcement from workers and state attorneys general to crack down on employers that continue to push the limits through unfair contracting. 

And, just as importantly, the legal profession will need to do some real self-policing against lawyers encouraging clients to enter into illegal and unenforceable non-compete agreements. Behind every unfair and illegal employment contract is a lawyer who drafted it. It’s unethical for a lawyer to file a frivolous complaint, and it should be unethical for a lawyer to encourage a client to enter into an unenforceable contract. But it happens all the time. 

In the FTC’s enforcement action against a Michigan security guard employer, the employer continually required workers to sign non-compete agreements that Michigan courts had already refused to enforce. If a lawyer was involved in counseling the employer to continue presenting these contracts to workers, wasn’t the lawyer deceiving their client’s workers by presenting them with an unenforceable fine print term? The suggestion to workers, of course, is that a contract with their employer has the force of law—in many cases, the employer and its lawyer know this isn’t true, but an unenforceable contract term can exert a powerful chilling effect on workers who will unlikely ever challenge it. State professional ethics boards should clarify that it is unethical for lawyers to encourage clients to present consumers and workers—especially those who are unlikely to have lawyers—with unenforceable contract terms.

Leah Samuel

The FTC’s authority to make rules like the non-compete ban under its “unfair methods of competition” (UMC) rulemaking authority will inevitably be challenged for allegedly running afoul of the newly minted major questions doctrine. When it is, many will look to 1914, the year that the FTC Act was passed. Yet, as I will briefly explain, the 1970s is the more relevant period.

The FTC’s ability to conduct substantive rulemaking under both its UMC and UDAP (consumer protection) mandates was upheld in National Petroleum Refiners (D.C. Cir. 1973). Shortly after the Supreme Court denied cert in Nat’l Petroleum, Congress passed the 1975 Magnuson–Moss Warranty Act, which implicitly ratified the Nat’l Petroleum decision as it applied to UMC rulemaking, while raising procedural burdens for UDAP rulemaking. For the textualists out there, the ordinary meaning and legal background of the statute very clearly leaves UMC rulemaking intact under the Nat’l Petroleum standard. However, after last term’s West Virginia v. EPA, judges with a deregulatory bent can opt for a sort of heightened scrutiny approach to look outside the text when a statutory authorization isn’t crystal clear, and when the agency is exercising an “unprecedented” power in contravention of “unbroken” agency practice.

So is this exercise unprecedented? The common law roots of Section 5 and its early enforcement by the FTC did not distinguish between antitrust and consumer protection—fairness and truthfulness in advertising was understood to be an integral part of fair competition, and it was only in later years that consumer protection and antitrust were distinguished in administration and academia. When the FTC first began to promulgate Section 5 rules, neither the agency nor the courts sharply distinguished between the UMC and UDAP authority. Critically, the Octane Rule at stake in Nat’l Petroleum not only invoked both powers but conceptually linked the information-forcing priorities of consumer protection with the competition-enhancing priorities of antitrust. Though driven by Naderite consumer protection advocates, the Octane Rule was a valid exercise of UMC power, in substance and in text.

Both the Octane Rule and the later Standards and Certification Rule (never finalized) make clear that the exercise of UMC rulemaking power is not “unheralded” and shouldn’t be treated as a major question.

Sandeep Vaheesan

In light of the FTC’s proposed ban on non-competes, it’s worth taking a moment to reflect on how we arrived at this moment. In 2014, reporting by the Huffington Post’s Dave Jamieson revealed that fast-food chain Jimmy John’s was imposing non-competes on its sandwich makers. This story helped put non-competes on the national radar and overturn the assumption that only highly paid workers (think CEOs) had non-competes. Subsequent research by the prolific economist Evan Starr and the Economic Policy Institute found that non-competes bind tens of millions of workers across nearly all lines of work and up and down the income scale. Further, empirical analysis showed that these contracts, which have a weak justification, generally have adverse effects on wages and wage growth.

Based on this and other research, a labor and public interest coalition led by the Open Markets Institute (where I work) petitioned the FTC in 2019 for a rule banning non-compete clauses for all workers. Although the FTC during the Trump years did not act on our petition, it held a public workshop on non-competes in early 2020 and invited comments from the public, helping develop the evidentiary case for a future rulemaking. 

With the election of Joe Biden (who has long condemned the use of non-competes by employers) and appointment of Lina Khan (who had previously supported FTC regulatory action on non-competes), the stage was set for this month’s announcement. Indeed, given the work done prior to President Biden’s inauguration in January 2021, an outside observer might wonder why the FTC took so long just to propose a ban, especially since earlier regulatory action would not have faced an anti-labor Republican House of Representatives and potential threats such as the Congressional Review Act.

What comes next? A big fight lies ahead. During the 60-day comment period on the FTC’s proposal, opponents of regulatory action will seek to weaken the final rule. Some friendlier voices might recommend the FTC enact a more targeted regulation, one for instance that protects only low-wage workers. They will cite the risk of judicial invalidation of a blanket ban, undoubtedly a threat, in favor of a more modest measure.

The FTC presented ample evidence that a full ban is the right policy. The factual record in the proposal suggests line-drawing—such as protecting workers making less than $100,000 a year but not those above that threshold—might be the “arbitrary and capricious” choice. For decades, the FTC has targeted the use ofvertical restraints” (a category that includes non-compete clauses), and in the pre-Sherman Act common law, restraint of trade “nearly always referred to limiting or prohibiting someone from engaging in a particular trade or business.” This history weakens the claim that the FTC is doing something novel. Nonetheless, the conservative courts have flexible authority to do what they want to do. Rather than water down its own policy in anticipation of a poorly reasoned court decision, the FTC must do all it can to protect workers in the United States and defend the strongest policy before the judiciary.

Sanjukta Paul

I write to make a quite narrow and limited point, yet one I have seen crop up in the developing debate among antitrust lawyers over the FTC’s proposed rule and one that implicates fairly deep aspects of the relationship between antitrust law and the legal regulation of the labor relationship.

Despite some action at the level of federal antitrust law in recent years, worker noncompete agreements at present are largely regulated by state law. That state law often has a statutory basis, but as a practical matter (outside specific exceptions and carve-outs, or outright bans), it has a recognizable structure and elements inherited from common law, and is still typically worked out in common-law fashion by state judges.

This common law treatment of worker noncompete agreements is often referred to in terms of the “rule of reason”—and this seems to lead to some confusion given the specific and different development of the “rule of reason” in federal antitrust law. They are similar only in that they involve the consideration of multiple factors and entrust judges with substantial discretion in applying and balancing them in particular cases. They do not, however, share the same structure or elements.

As a result of this potential confusion, antitrust lawyers may be inclined to think that removing the “market power” requirement from the regulation of worker noncompetes would be a significant reform. On this line of thinking, one might expect it to lead to substantially different outcomes than the results we have seen documented in empirical research over the past couple of decades (during which state law has been the essential factor). Indeed, the absence of a market power requirement would be a significant reform to the rule of reason used in the application of federal antitrust law, but not to the state law governing noncompete agreements, which generally has no such market power requirement. See, e.g., Hopper v. All Pet Animal Care Clinic, 861 P.2d 531 (Wyo. 1993) (following the Restatement, and noting that the initial evidentiary burden is on employer to show that the restriction is reasonable on its own as well as necessary for the protection of legitimate business interests).

There seems to be a near-universal consensus that status quo regulation is not doing enough to deter the adoption of noncompete agreements, even among those commentators who oppose an outright ban. If that is the case, though, the absence of a market power requirement is unlikely to change current levels of deterrence—since that’s already the law in states following a standard formulation.

Antitrust lawyers should keep in mind that the relative permissiveness toward worker noncompetes in recent decades is not the product of a hoary common law antitrust tradition but rather of encroachment by the master-servant aspects of employment law at the intersection of the two areas of law. As judges routinely note, the common law policy against covenants not to compete is one of the oldest. The relative toleration of noncompete agreements in more recent times was worked by effectively extending the penumbra of the duty of loyalty a servant owes to the master beyond the end of the term of employment, not by principles inherent in either current or past antitrust law.

At the broadest level, current debates in antitrust and competition circles over labor issues too often tend to assume a ‘neutral’ baseline for the legal treatment of labor relationships—instead of recognizing that that treatment has always been specific to labor in many ways—and then proceed to theorize any antitrust reforms that actually favor workers in some way as special anti-competitive exceptions. In the end, of course, we must recognize that competition law never just promotes healthy competition, but also channels competition in particular directions and chooses among the coordination mechanisms that are present in every market. The law of master and servant was one coordination mechanism, and the fairer contracts workers may be able to bargain under this rule are simply another.

Najah Farley

The FTC’s proposed rule, if enacted, will completely ban all noncompete clauses, which management lawyers have justified as a tool through which companies can “protect their investments.” A rule of this breadth will ensure that all workers—especially low-income workers—can move freely to increase their wages and leave abusive work environments without engaging a lawyer to defend their right to move to another job. On first blush, it might seem strange to regulate such a key dimension of the employment relationship through antitrust law. But I see this as an important manifestation of how antitrust and labor law can work together to increase worker power.

The first conference where I was invited to speak about noncompetes was the ABA antitrust conference. I presented my work with a group of lawyers from the EU, Canada, and Great Britain, and nearly all of them worked in the antitrust field. (I discussed some of these issues again in the Regulatory Review later that year.) Prior to the invitation to that conference, I had viewed noncompetes from the vantage point of a labor and employment attorney. My takeaway was that internationally, many countries outlaw such clauses altogether except as applied to corporate executives—and antitrust played an important role in structuring the rules concerning noncompete agreements. If finalized, the new FTC rule will bring US law closer to international norms; this is fitting because antitrust law considers aspects of competition and the labor market beyond each individual worker, and thus has the capacity to address non-competes as a structural economic problem.

Tackling these agreements through antitrust can demonstrate how employers have explicitly used these agreements to lower wages across industries. Antitrust investigations can expand our understanding of the power that employers have to keep wages stagnant and also may reveal key information about the relationship between noncompetes and wage inequality (i.e., that noncompetes contribute to increasing wage and wealth inequality by decreasing a worker’s power to change jobs and bargain for higher wages).

Importantly, though, the FTC should collaborate with the Department of Labor to communicate information about the impact of noncompetes on wages. It should also work with DOL, the EEOC, and the NLRB to develop best practices for broad-based enforcement across industries. In a perfect world, the FTC could collaborate with these worker-focused agencies to strategize about which industries may be affected by noncompetes and structure remedies so that they benefit the greatest number of workers. In so doing, the FTC’s rule could invigorate and develop the synergy between the fields of labor and antitrust law to protect workers and increase worker power.

Basel Musharbash

Some critics of the proposed ban will argue that it is too broad or aggressive. However, based on my experience litigating noncompetes, I’m skeptical that anything less than a summary ban would have a meaningful effect on the behavior of employers, who will use any justification, however thin, to defend their use. For example, I’ve seen employers argue that even a noncompete against a janitor is necessary to protect their “confidential information,” “customer goodwill,” and “investment in specialized training” — and use that argument to drag out their case for years. That janitor was lucky to find a law firm like Pollard PLLC (where I worked at the time), which specialized in worker-side noncompete litigation and could take her case on contingency. But that’s not the case for most workers.

Most workers don’t have the resources to challenge a noncompete; they also generally don’t know that a noncompete is not “just like any other contract” — i.e., subject to antitrust law. (Remarkably, this is a point that we routinely had to explain to federal judges). As a result, it’s almost never necessary for an employer to file an actual suit; they can usually get compliance with a noncompete just by threatening one. If that doesn’t do the trick and the worker starts a new job with a competitor, all their former employer usually has to do to get them fired is send a demand letter to their new employer. After litigating a couple dozen — and working with people who’ve litigated a few hundred — of these cases, I can confidently say that the number of companies that will *not* fire a non-executive worker after getting slapped with a tortious-interference-with-noncompete-agreement claim from their former employer is . . . next to nil.

The upshot of all this is to say that, as long as the law leaves a colorable argument for the lawfulness of a noncompete against a given worker, employers will likely be able to exploit the asymmetries of power between them and the worker to secure compliance — legal formalities notwithstanding. I also have little faith that lawyers will faithfully implement any type of factorial analysis to tell an employer they can’t use a noncompete. If the FTC intends to do something meaningful, the FTC needs to ban noncompetes clearly and categorically. Maybe then, lawyers would be open to malpractice liability for writing a noncompete into an employment contract, and maybe then they’d stop doing it.