The Federal Trade Commission should develop a nuanced approach to employee noncompete agreements.
The Federal Trade Commission (FTC) has proposed a rule that would ban nearly all employee noncompetition agreements as unfair methods of competition. These agreements prevent an employee from taking a new job in a competing business for a certain period after his or her employment ends. The duration and geographic scope of the agreements varies.
Today, purely vertical noncompete agreements are analyzed under antitrust’s rule of reason and most are lawful under federal law. Several states have enacted stronger laws. The FTC’s proposed rule would apply the FTC Act, a statute that reaches further than the Sherman Antitrust Act but cannot be enforced by private parties.
This essay does not consider the FTC’s power to make and enforce a rule like the one proposed, but rather it addresses the wisdom of the proposed rule itself. Should employee noncompete agreements be invariably unlawful or should the law reach only a subset? If the latter, how should the law distinguish between different agreements?
The effects of employee noncompete agreements vary with the type of job. Recent attention has turned to noncompete agreements covering lower-wage employees in industries such as fast food. These are unjustified and anticompetitive, and there is growing support for prohibiting them. In these situations, employers have not invested a great deal in specific employees. Most of their employment is “at will,” which means that either employer or employee can ordinarily terminate employment on reasonable notice. A noncompete agreement in such situations is nothing more than an anticompetitive restraint on employee mobility, limiting an employee’s right to seek a better job by moving or threatening to move elsewhere.
Other employees are different. They may receive specialized training that is costly to their employer and “portable” in the sense that the training can be carried to a different employer. They might also hold confidential information, trade secrets, customer lists, or other knowledge that would be of value to a different employer. For many of these employees, a noncompete agreement is the only effective way an employer can protect its investment from the risk of “free riding,” which occurs when one person is able to commandeer another person’s investment. For example, although an employer may be able to prohibit an employee from purloining a trade secret, there is no way to prohibit a former employee from using skills provided by the first employer in a second employer’s business. Here, noncompete agreements are socially valuable, provided that they are reasonable in scope and duration. Workers who sign such agreements often receive higher wages than those who do not. Noncompete agreements can protect the employer’s investment. A rule categorically banning them would harm these workers as well as their employers.
A well-designed policy on noncompete agreements should distinguish these situations. The important questions are (1) whether harmful free riding is likely, (2) whether protection requires a noncompete agreement, and (3) whether the particular agreement is no broader than necessary to protect the employer’s legitimate interests. For all these questions, the burden of proof should be on the employer, because it is in the best position to provide evidence showing its own investment and justifications for the agreement.
In its present form, antitrust law’s rule of reason is not effective for addressing this issue. Its requirements are too onerous, including that the employer possess significant market power in the labor market. Today, nearly all plaintiffs lose rule of reason cases.
Considering only antitrust law, one can see why an antitrust enforcer such as the FTC would prefer per se illegality. A per se rule simply condemns noncompete agreements without allowing for justifications or requiring proof of market power. Given the current sad state of antitrust’s rule of reason, policymakers may have no effective choice between per se illegality or de facto legality under the rule of reason.
But FTC rulemaking need not be hitched to that wagon. It should attempt to repair defects in the rule of reason, not simply plow over them. The FTC can develop its own, more nuanced approach to distinguishing beneficial from harmful noncompete agreements. That approach need not require a showing that an employer possesses market power. But it should also permit defenses that antitrust law’s per se rule does not permit.
Within this framework, any noncompete agreement that operates against a broad range of employees without accounting for these free rider concerns should be struck down without further inquiry. Restraints contained in agreements between upstream parties, such as a franchisor and its numerous franchisees, and then secondarily imposed on employees should also be treated as highly suspicious. A franchisor has no obvious interest in preventing the employees of its various franchisees from inter-franchisee movement. The most likely reason for such arrangements is that the franchisor is responding to what is essentially a cartel—or no-poaching agreement—of its own franchisees, although one that may be technically difficult to prove. These agreements are substantively overbroad and should be condemned as unreasonable restraints on worker mobility.
The FTC rule also applies to noncompete agreements with independent contractors. Such workers do not even receive some of the minimal wage and hours protections that apply to employees. At the same time, such agreements can occupy a position more akin to exclusive dealing in business services, particularly when they involve nonlabor elements, or the hiring firm commits dedicated and specific resources.
One legitimate concern is inter-brand free riding, which might occur when an independent contractor working for a firm can employ an asset or other costly resource in service provided for another firm. Hypothetically, a firm such as Uber might supply its drivers with cars, which the driver then might use to carry passengers for Lyft, a competitor. Independent contractors often use tools, databases, or other assets provided by their hirer. In many such cases, a rule prohibiting sharing of the asset is a reasonable and less restrictive alternative than a noncompete agreement. They should thus be presumptively, not absolutely, unlawful. The proposed FTC rule acknowledges this possibility.
Given the expertise and flexibility that the FTC can apply, one would have hoped for a rule about employee noncompete agreements that appreciates more fully the variety of such agreements, the varying degrees of harm or benefit that they might offer, and weaknesses in the now-existing rule of reason that make it inadequate to evaluate them. Many noncompete agreements offer little or no social benefit, and full rule-of-reason analysis is seriously underdeterrent against these inefficient agreements. Other agreements, though, offer benefits for employees as well as the economy.
The FTC should consider issuing a final rule that will recognize a distinction between these categories of agreements and improve on the process for distinguishing between them, rather than obliterating the distinction altogether.