“Unfair method of competition.” That is the legal category the Federal Trade Commission placed noncompete clauses in on Tuesday, and the phrase is doing precise work: it is the statutory hook the agency has, and it is the thing the courts will be asked whether it reaches.
The rule bans new noncompetes for all workers and renders most existing ones unenforceable, with a carve-out for existing agreements with senior executives in policymaking positions above a defined earnings threshold. It was adopted three to two.
The belief worth testing here is the one that has defended these clauses for thirty years: that they exist to protect trade secrets and investment in training. The record shows something narrower and more mundane.
What the defense claims
The argument is coherent and it is not made in bad faith.
A firm that invests heavily in developing a person, or that gives an employee access to genuinely valuable confidential information, is exposed if that person can walk to a competitor the following week. Without some protection, the reasoning goes, firms underinvest in training and are more guarded with information, and both of those are losses to everybody.
That describes a real problem. It describes it for a particular kind of employee: senior, technical, with access to something specific that a rival would pay for.
What the record shows about who actually signs
The clauses are not confined to that population and have not been for a long time.
Noncompetes have been applied to hourly staff in food service, to hairdressers, to warehouse workers, to security guards, to camp counselors. Those are not people with trade secrets. There is nothing in the job that a competitor could learn from hiring them beyond what is visible from the street.
Which means that for a large share of the agreements in force, the trade secret justification cannot be the reason, because the predicate is absent. Something else is being purchased.
What is actually being purchased
Wage suppression, and not as an accusation. As a mechanism.
A worker’s main leverage over pay is the credible ability to leave. That leverage does not require actually leaving; it requires the employer to know that a better offer down the road is available and would be taken. A noncompete removes it, and it removes it whether or not the employer ever enforces the clause.
That last point is the one that makes the trade secret defense hardest to sustain. Enforcement is rare, expensive and uncertain. The clause works anyway, because a worker who believes they are bound behaves as though they are, and the cost of finding out is a lawsuit they cannot afford.
So the instrument does most of its work in the minds of people who will never see a courtroom, which is a strange property for something defended as a remedy for a specific harm.
What the existing law already did
Here is the part that decides whether the reversal holds, and it is the strongest argument for the rule.
Trade secrets are separately protected. There is a body of federal and state law about them, and it is available to a firm whose former employee has taken something. Confidentiality agreements are not affected by this rule. Neither are nonsolicitation clauses covering a departing employee’s approaches to clients.
A firm with a genuine secret therefore has several instruments that remain entirely available. What it loses is the one instrument that does not require it to identify what was taken, or to show harm, or to prove anything at all: the one that simply prevents the person working.
That asymmetry is the reversal in a sentence. Every tool that requires proving a harm survives. The tool that requires proving nothing is the one being removed.
What the evidence base actually shows
A commission writing a rule of this reach has to build a record, and the record is the part that will be picked over in court.
The rule follows a proposal from January 2023 and a review of more than 26,000 public comments, which is a large response by the standards of agency rulemaking and is itself evidence of how many people believe themselves affected.
The empirical literature the agency relies on is genuine and it is not unanimous. Studies exploiting differences between states, several of which already prohibit these clauses, find effects on wages, on job mobility and on new business formation. They also find that the effects vary by occupation and by how enforceable the clauses were to begin with, which is what you would expect from an instrument whose main channel is what workers believe rather than what courts do.
The honest summary is that the direction is reasonably well established and the magnitude is contested. That is a normal state for economics and an awkward one for a rule that has to survive judicial review on the strength of it.
The case against the rule that is not about noncompetes
It should be said that the strongest objection has nothing to do with whether noncompetes are good.
It is about whether an agency can do this. The commission is asserting authority to write a substantive competition rule governing tens of millions of private contracts, and the dissenting commissioners argued it does not have that power. Challenges were filed the same day, and the courts have recently been notably unsympathetic to agencies claiming broad authority from old statutes.
A person can believe the clauses are indefensible and that the agency lacked the power to ban them. Those are separate questions and only the second will be decided this year.
What to watch
The rule has an effective date in September, and the first thing to watch is whether it survives to reach it. A court in Texas issuing a stay in August would be the least surprising development of the year.
The second is what employers do in the meantime, because the interesting behavior happens before any ruling. A firm that responds by tightening confidentiality agreements and nonsolicitation terms is telling you it had a real interest to protect. A firm that responds by extending notice periods and garden leave is telling you it wanted the person off the market, which is the same purchase by another name and is not covered by this rule.
Nothing has changed yet for the person who has one of these in a drawer, and that is worth stating precisely rather than hopefully. The rule is not in effect. The clause is exactly as enforceable as it was on Monday. The first real information about it will arrive from a courtroom in Texas rather than from anybody’s personnel department.
