Non-Compete Clauses Quietly Trap Millions of Low-Wage Workers

The non-compete clause was invented for a specific, narrow problem: an executive walks out the door with the customer list, the pricing model, and the product roadmap, and sets up shop across the street. Courts tolerated these contracts as a grudging exception to the general rule that people are free to work where they choose, because the harm they prevented was concrete and the people who signed them had lawyers.

That is not who signs them anymore. Non-compete clauses now reach deep into the hourly workforce — sandwich makers, security guards, hair stylists, home health aides, dog groomers. According to research compiled by the Economic Innovation Group, roughly 18 percent of the American labor force, on the order of 28 million people, currently works under a non-compete, and around 38 percent of workers have signed one at some point in their careers. The Federal Reserve Bank of Minneapolis reports that about 13 percent of workers earning less than $40,000 a year are bound by one, as are 14 percent of workers without a bachelor's degree. These are not people carrying trade secrets. They are people carrying name tags.

A Contract Most Workers Never Negotiate

To understand why the system works, start with when the paper appears. Survey research on non-competes consistently finds that a large share of workers are asked to sign after they have already accepted the offer — sometimes on the first day, in a stack of onboarding documents between the tax forms and the direct-deposit authorization. By that point the worker has usually declined other offers, given notice at the old job, maybe relocated. The realistic options are to sign or to walk away from a job they have already reorganized their life around.

There is no negotiation because there is nothing to negotiate with. A software architect fielding three offers can strike the clause or price it in. A cashier cannot. The U.S. Government Accountability Office, reviewing the evidence in 2023, found that employers routinely apply non-competes far beyond any plausible trade-secret rationale, that the agreements restrict job mobility, and that they likely suppress wages — not only for the workers who sign them, but for others in the same labor market, because every trapped worker is one fewer competitor an employer has to outbid.

That last point is the quiet center of the whole arrangement. The Economic Policy Institute describes non-competes as part of a broader pattern of employers requiring workers to sign away rights as a condition of employment. Pay rises when workers can credibly leave. A non-compete is a device for making leaving incredible.

Why a Sandwich Shop Wanted a Two-Year Non-Compete

The case that dragged this practice into public view involved sandwiches. Until 2016, the fast-food chain Jimmy John's required its hourly employees to sign agreements barring them, for two years after leaving, from working at any business within two miles of a Jimmy John's store that earned more than 10 percent of its revenue from sandwiches. Given the chain's footprint, the clause functioned as a ban on food-service work across much of urban America — imposed on people assembling subs for wages near the legal minimum.

The attorneys general of New York and Illinois investigated, and Jimmy John's settled, agreeing to stop using the clauses, to rescind the ones already signed, and to pay $100,000 toward outreach in Illinois. New York's attorney general at the time called non-competes for low-wage workers "unconscionable," noting that they "bully" vulnerable workers into staying under threat of a lawsuit.

What the case revealed was not one company's overreach but a business model. No court would likely have enforced a two-year non-compete against a sandwich maker. It didn't need to be enforceable. It needed to be believed.

The Chilling Effect Is the Point

This is the mechanism that makes low-wage non-compete clauses effective even where they are legally void. A worker earning $15 an hour who receives a letter from her former employer's law firm does not research the enforceability of restrictive covenants in her state. She cannot afford the consultation that would tell her the letter is a bluff. She turns down the competing offer, or the new employer — also unwilling to buy a lawsuit — quietly withdraws it.

Economists call this the in terrorem effect: the contract does its work through fear rather than litigation. It helps explain a finding that would otherwise be puzzling — non-competes suppress mobility even in states where they are unenforceable. The paper doesn't have to hold up in court. It only has to sit in a drawer and be remembered.

The same logic has spawned substitutes. As some states restrict non-competes, employers have shifted to training repayment agreement provisions — TRAPs — which bill departing workers thousands of dollars for "training" of often negligible value, and to broad non-solicitation and confidentiality clauses that mimic a non-compete's effect. Federal consumer regulators have opened inquiries into this kind of employer-driven debt. The instrument changes; the leverage is the constant. It is the same asymmetry that lets wage theft flourish as the biggest property crime nobody calls crime: the cost of asserting your rights exceeds the value of the rights themselves.

A Federal Ban That Rose and Fell

For a moment, this looked like a solved problem. In April 2024, the Federal Trade Commission finalized a rule banning nearly all non-competes nationwide, estimating it would raise wages across the economy. The rule never took effect. In August 2024, a federal district court in Texas set it aside in Ryan LLC v. FTC, holding that the agency had exceeded its statutory authority.

What followed was quieter and more decisive. In September 2025, the FTC voted to abandon its appeals and accede to the rule's vacatur, and effective February 2026 the agency formally removed the rule from the Code of Federal Regulations. The commission says it will still pursue individual enforcement actions against the most abusive covenants, particularly those binding low-wage workers. Case-by-case enforcement against a practice covering tens of millions of contracts is, by arithmetic alone, a different kind of promise than a ban.

The State Patchwork That Decides Who Goes Free

With the federal rule gone, whether a worker is trapped depends on their zip code. California, Minnesota, North Dakota, and Oklahoma ban non-competes outright. Washington has gone further than most, enacting legislation that will void nearly all non-competes in mid-2027; until then it enforces a salary floor set well above six figures. Illinois voids non-competes for workers earning under $75,000. A growing list of states has adopted similar low-wage thresholds, and others have tightened their rules this year.

The patchwork produces strange geography. A line cook in Minneapolis is free; the same cook across the border in Wisconsin can be bound. And thresholds only help workers who know they exist — which returns us to the chilling effect, because an unenforceable clause signed in Illinois still sits in the same drawer, remembered the same way.

Who Benefits From Keeping Workers in Place

Why does the practice persist against decades of evidence? Because it works, and for the employer it is nearly free. A non-compete lowers turnover without raising pay — it substitutes a legal restraint for a retention raise. In labor markets where a handful of employers dominate hiring, the clause deepens an existing power imbalance. The same calculation shows up wherever employers can shift costs onto workers who lack the leverage to refuse, from restrictive covenants to the injury rates of high-speed warehouse work: the savings are booked by the company, and the cost is carried privately, in narrowed options and forgone wages.

The evidence on the other side is just as consistent. Research on state-level bans and thresholds finds that restricting non-competes for low-wage workers raises their earnings and increases mobility, with little sign of the harms employers predict. Trade secrets remain protected by trade-secret law, which exists in every state and requires no contract at all.

What Would Actually Fix This

The honest answer is that the problem is already legally solvable and remains unsolved for reasons of power, not doctrine. A durable fix would ban non-competes for workers below a meaningful wage threshold everywhere, require any lawful clause to be disclosed with the job offer rather than after acceptance, render void clauses unenforceable and penalize employers for asking, and close the TRAP loophole before it becomes the next standard form. Several states have adopted pieces of this. None of it requires new theory — only the decision that a sandwich maker's next job is not a business asset.