Home Antitrust and Competition The Ties that Bind Workers to Firms: No-Poach Agreements, Noncompetes, and Other...

The Ties that Bind Workers to Firms: No-Poach Agreements, Noncompetes, and Other Ways Firms Create and Exercise Labor Market Power

0
Illustration by erhui1979, via Getty Images

Collusive no-poach agreements are per se illegal, but noncompete clauses are not. Recent research casts doubt on the rationale for this legal distinction and the Biden administration is considering banning noncompetes. But if the Biden administration does so, it must also be ready to consider the potential anticompetitive effects of alternative arrangements, which can also effectively and in some cases more strongly bind workers to firms. 


Does antitrust have a labor market problem? The last few years have seen growing interest among academic scholars in the causes and effects of concentration in US labor markets. Concurrently (and not unrelated), there has been an explosion of interest among policymakers and the general public in the impact that firms with market power may have on wages and working conditions. What do the data say regarding employer concentration and its effect on workers? Is antitrust in its current form equipped to address issues related to labor market power? In an attempt to answer these questions and more, we have decided to launch a series of articles on antitrust and the labor market.

In 2009, the US Department of Justice sued several Silicon Valley tech companies for agreeing not to poach each other’s workers. The case, which eventually settled for more than $400 million, marked a turn for US antitrust agencies towards long-neglected anticompetitive behavior in labor markets. This turn was further highlighted both by the DOJs 2016 Antitrust Guidance for Human Resource Professionals—which alerted HR professionals to the fact that agreements between competitors to limit or fix the terms of employment for potential hires may violate antitrust laws—and several recent cases involving similar “no-poach” agreements, including between medical schools and between railway firms. But the implications of these cases raise antitrust concerns about anticompetitive conduct in labor markets that go well-beyond collusive “no-poach” agreements.

In the wake of the Silicon Valley no-poach case, Arijit Mukherjee and Luís Vasconcelos wrote an article arguing that employer collusion not to hire and exclusive employment agreements such as noncompete clauses are substitute arrangements. Their argument suggests that, were it not for California’s longstanding ban on noncompetes, the tech firms would not have needed to collude because they could have stemmed such labor market competition with noncompete clauses. This argument is important because it highlights a key distinction between noncompetes and no-poach agreements: while bilateral no-poach agreements, which cover just one other employer, are per se illegal, noncompete clauses, which prohibit departing workers from joining or starting companies within an entire industry, are legal in most states, subject to a reasonableness test that balances the protection needed by firms with the harm done to workers and society.

However, this legal distinction is on the verge of changing. Spurred by evidence documenting that 18 to 28 percent of US workers are bound by noncompetes, including 14 percent of workers paid below median earnings, in the last five years eight states have taken action to ban low-wage noncompetes. And just this year, more than 60 state or federal noncompete bills were proposed, including the Uniform Restrictive Employment Agreement Act promulgated by the Uniform Law Commission. Moreover, in July, President Biden encouraged the Federal Trade Commission to “curtail the unfair use of non-compete clauses,” and a recent FTC workshop on promoting labor market competition considered the issue.

Considering the policy discussions currently taking place around noncompete clauses, it is worth asking what explains the discrepancy between the legal treatment of no-poach agreements by antitrust authorities and that of noncompetes, given that both are both restraints of trade in the labor market. While arguments justifying no-poach and noncompete clauses are often similar—that they are needed to protect firm investments in trade secrets, training, goodwill, etc. from being misappropriated—one primary distinction revolves around who “agrees” to them. No-poach agreements are bilateral, organization-level agreements; workers are generally unaware that they exist, and thus cannot be compensated for working under one. In contrast, noncompetes are written into the worker’s employment contract, such that, at least in theory, workers can be compensated for giving up their post-employment freedom. 

Although this distinction undergirds (at least in part) the disparate treatment of no-poach and noncompete clauses under the law, recent research casts doubt on its validity. For example, workers bound by noncompetes earn 5 percent less on average than those bound by only nondisclosure agreements, and several studies show that where noncompete clauses are more enforceable worker mobility and earnings are lower, including for low-wage workers, the average worker, and high-tech workers. For example, in the study of high-tech workers researchers found that high-tech workers starting their job in a state that enforces noncompete agreements earned, cumulatively, 5 percent less eight years later than comparable high-tech workers starting their job in states that did not enforce noncompete clauses. To contrast these effects with the effects of no-poach agreements, a recent working paper by Matt Gibson re-examines the Silicon Valley no-poach case and estimates that workers at firms with just a single no-poach agreement experienced 2.5 percent lower annual salaries as a result.

While a few recent studies of executives and physicians document positive wage effects related to noncompetes, this recent literature largely rejects the idea that the average worker is compensated for agreeing to a noncompete clause. Instead, as Balasubramanian et al. put it, noncompete clauses appear to “lock workers in.” These results should be concerning to antitrust authorities because they suggest that an apparently core distinction between noncompetes and no-poach agreements does not hold for most workers. 

Aside from these wage effects, however, recent research reveals another reason why noncompetes are troubling for antitrust authorities: that restraints of trade can generate negative spillovers that unfairly affect labor and product market competition. 

Regarding labor markets, one study finds that where enforceable noncompetes are used en masse, the whole labor market is slower moving and lower earning, including for those not bound by noncompetes. Another finds that state increases in noncompete enforceability reduce earnings both in the focal state and in neighboring labor markets in other states. One reason for these spillovers is that noncompete clauses make it harder for firms to hire, such that firms that would otherwise benefit from hiring an employee are worse off as a result. In the case of executives, a recent working paper finds this externality is so large that the optimal enforcement policy is close to a ban, even though the executives benefit when agreeing to a noncompete.

Even if one is willing to overlook the negative wage effects and negative spillovers on workers and other firms—and even if one thinks workers are rationally agreeing to the terms in their employment contract (which seems dubious for most workers given the survey evidence)—a recent paper suggests that noncompete clauses should still be of concern to antitrust authorities. This is because noncompetes are also restrictions on firm entry and can thus blunt product market competition, increasing prices and reducing consumer welfare. If competition reduces rents to the firm, then a rationale worker may decide to agree to a noncompete clause in exchange for a share of the rents that they would have achieved via starting a competitor. In this way, just like a no-poach agreements represents a form of collusion between competitors, a noncompete clause reflects a form of collusion between potential competitors (i.e., the firm and its employees). These effects can be particularly pernicious in markets where noncompetes are very common, such as in health care and high-tech, because if all workers are bound by noncompetes, then who will start the new competitor, and who will any new firm be able to hire?

“If the FTC does ban noncompete agreements, as some have encouraged, the antitrust problems posed by postemployment-type restrictions on workers won’t end.”

Recognizing how noncompetes disadvantage workers, other firms in the labor market, and consumers, the Federal Trade Commission might consider noncompetes to be either unfair methods of competition or an unfair and deceptive practice—which they could use to promulgate rulemaking. Interestingly, the only occupation for which noncompetes are unenforceable in every state states are lawyers, who justified such a ban based on the idea that restrictions on an attorney’s professional autonomy are tantamount to restrictions on a client’s freedom to choose a lawyer. That is, noncompete clauses are unenforceable for lawyers because of concerns about negative spillovers onto customers for legal services. 

If the FTC does ban noncompete agreements, as some have encouraged, the antitrust problems posed by postemployment-type restrictions on workers won’t end. In fact, if there are no penalties in the law for using unenforceable noncompetes, it is likely that firms will still use unenforceable noncompetes, that workers will be largely unaware that their noncompetes are unenforceable, and that the unenforceable noncompetes will still chill employee mobility. Alternatively, if there are some teeth to any such rule, firms may resort to no-poach-type arrangements like in the Silicon Valley case. Firms may also shift towards fixed-term employment contracts, as suggested by the California-based lawsuit in Twentieth Century Fox Film Corp. v. Netflix. In this case, Fox sued Netflix for poaching Fox executives midway through their fixed-term contracts. While the judge rejected Netflix’s arguments that such contracts were against California’s public policy, the practice of renewing fixed-term employment contracts just a few months before their expiration can make it difficult for workers to time any labor market transitions—thereby making it difficult for firms to compete for labor to resolve short-run hiring needs.

More likely, however, is that firms will bolster their use and enforcement of other restrictions firms regularly use alongside noncompetes. These include, among others, non-disclosure agreements, client and coworker non-solicitation agreements, training repayment agreements, arbitration agreements and liquidated damages clauses. Indeed, in light of recent policy movements, attorneys are already encouraging firms to consider adopting these other tools. While in theory, these terms are less restrictive than a noncompete—because they do not directly prohibit workers from taking a job—they raise similar and sometimes more severe antitrust concerns. 

For example, training repayment agreements, which typically require workers to pay back training expenses if they leave before some predetermined time, may have dramatically inflated training repayment costs. Similarly, liquidated damages clauses can impose direct fines on workers for trying to leave. If the costs imposed by these restrictions are substantial, then these restrictions are broader than noncompetes because a departing worker would bear these costs when leaving for any other employer, not just a competitor. 

While the other restrictions may not impose direct fines on workers, they can impose such broad restrictions that they effectuate a noncompete. For example, in TLS Mgmt. and Mktg. Ser. LLC v. Rodriguez-Toledo, 19-1104P (1st Cir. 2020), a former subcontractor started a competing tax and accounting and was sued for taking trade secrets and violating his nondisclosure agreement. In considering these claims, the court determined that the nondisclosure agreement was unenforceable because the terms were so “astoundingly” broad as to effectively be a noncompete. 

And just like noncompete agreements can have negative spillovers for others in the market, the same is true for non-disclosure and non-solicitation agreements. For example, recent research shows that non-disclosure agreements, which cover 57 percent of workers, prevent workers from sharing negative information about their firm, making it harder for “high-road” employers to stand out in the labor market. And broad client non-solicitation agreements also directly impose negative externalities directly on clients since a worker changing jobs may then prevent clients from accessing services from their preferred provider. The same is true for co-worker non-solicitation agreements, where the departure of one employee may foreclose a job opportunity for a coworker. 

Currently, no state or federal law addresses these myriad employment restrictions and their potential anticompetitive effects. The only policy that does so is the Uniform Law Commission’s (ULC) recently drafted Uniform Restrictive Employment Agreement Act, which imposes limits on each of these restrictions and more. The final draft of the act was just released in December 2021, however, and no state has adopted it so far. However, the ULC hopes that many states will enact this act, just as many of them adopted the ULC’s Uniform Trade Secrets Act. 

The Silicon Valley no-poach case stirred antitrust authorities to consider anticompetitive behavior in the labor market. However, the similarity between noncompetes, which are generally legal, and no-poach agreements, which are not, raises questions about what distinctions between them justify their different treatments under the law. A growing body of evidence suggests that such distinctions are largely unfounded, and that, if anything, noncompete clauses raise further antitrust concerns about unfair competition. If the FTC uses its authority to engage in rulemaking on noncompetes, however, it must also be ready to consider the potential anticompetitive effects of substitute arrangements, which can also effectively bind workers to the firm. 

Disclosure: Professor Evan Starr is an Associate Professor at the University of Maryland Robert H. Smith School of Business. He regularly consults on issues related to postemployment restrictions for government agencies, think tanks, and within the legal community. Professor Starr acknowledges financial research support from the Ewing Marion Kauffman foundation, as well as data-sharing agreements with Payscale.com. Dr. Starr also received financial support from the Economic Innovation Group to write a 2019 policy brief on noncompetes, and received compensation for delivering lectures in the George Mason University Attorney General Education program. Dr. Starr also acknowledges support for his services as an expert witness in several recent and ongoing cases related to noncompete clauses.

Learn more about our disclosure policy here