What the policy is
A non-compete clause is a contractual provision in which an employee agrees not to work for a competitor (or start a competing business) for a specified period and within a specified geography after leaving a job. They are distinct from non-disclosure agreements (which restrict sharing of specific confidential information) and non-solicitation agreements (which restrict poaching clients or colleagues).
Non-competes are historically associated with protecting legitimate business interests: trade secrets, client relationships, and proprietary processes that require significant investment to develop. The classic case is a senior software engineer who has access to unreleased products and customer data.
But research over the past decade has found non-competes far more widespread than this rationale would justify. Approximately one in five American workers is currently bound by a non-compete, including workers in low-wage industries like food service and retail, where trade secrets are not a plausible concern. California, North Dakota, and Oklahoma ban non-competes almost entirely, and comparison with other states offers a natural experiment into their effects.
How it works
Non-competes affect labor markets through three channels. First, they reduce worker outside options: if a worker cannot join competitors, their employer faces less wage competition and can pay less. This monopsony effect is the primary wage channel.
Second, non-competes reduce labor market mobility. Workers who want to advance their careers or escape bad employers are constrained by geography and industry scope. Research shows regions with enforced non-competes have lower rates of job-to-job transitions and less wage growth from mobility.
Third, non-competes slow knowledge diffusion. When skilled workers cannot move between firms, the spillover effects of their expertise (which generate broad economic gains) are suppressed. Silicon Valley's early advantage over Route 128 in Massachusetts has been partly attributed to California's ban on non-competes, which allowed engineers to move freely and spread knowledge across firms.
Who wins and who loses
| Group | Effect | Detail |
|---|---|---|
| Workers bound by non-competes | Cost | Direct wage suppression (estimated 4–8% for covered workers); reduced ability to change employers, negotiate raises, or relocate for opportunities. |
| Employers using non-competes | Benefit | Reduced turnover costs; protection against competitors recruiting trained employees; monopsony wage savings. |
| Competitors and new entrants | Cost | Reduced access to experienced talent; entry barriers that protect incumbent firms from competition. |
| Regional economies | Cost | Slower knowledge spillovers; reduced entrepreneurship as workers cannot leave to start competing businesses; less dynamic labor markets. |
| Consumers | Mixed | Ambiguous: less competition in product markets may raise prices, but firms may invest more in training if they can retain workers. |
What the evidence shows
Higher non-compete enforceability is consistently associated with lower wages for workers across income levels, with the largest absolute effects on mid-wage workers.
Arguments for and against
- Protects genuine trade secrets: Firms invest in developing proprietary processes, software, and customer relationships. Without protection, rivals could free-ride on that investment by hiring away employees. Non-competes preserve the incentive to invest in training and R&D.
- Supports employer-sponsored training: If employers fear workers will immediately leave to competitors after being trained, they will underinvest in workforce development. Non-competes may partially address this hold-up problem.
- Used far beyond their legitimate scope: Workers in minimum-wage jobs, retail, and food service regularly sign non-competes with no trade secrets to protect. This reflects employer bargaining power, not legitimate business justification.
- Wage suppression through monopsony: By eliminating competing offers, non-competes give employers wage-setting power over their workers. The evidence shows wage effects that cannot be explained by anything other than reduced competition for labor.
- Chills entrepreneurship and innovation: California's non-compete ban correlates strongly with its disproportionate share of venture-backed startups. Workers who cannot found competing companies take fewer entrepreneurial risks.
- NDAs are sufficient protection: Non-disclosure agreements can protect specific confidential information without restricting where workers can work. They are a more precise instrument that does not impose the same labor market costs.
Non-compete agreements serve a narrow legitimate function: protecting genuinely proprietary information in senior technical and executive roles. But they are wildly overused, applied to workers who have nothing secret to protect, as a blunt instrument to suppress wages and reduce worker mobility. The FTC's proposed rule to ban most non-competes reflects this evidence. The right reform is not elimination of all protection for trade secrets, but a narrow, clear standard that limits enforcement to cases where genuine intellectual property is at stake, not a catchall tool for wage suppression.