A free labor market is a road with many exits. Noncompete clauses, no-poach deals, training-repayment traps (TRAPs), and occupational licenses turn those exits into tollgates. The toll is not paid only by a worker; it is paid by wages, by entrepreneurship, by the republic's promise that talent may move as quickly as ideas. This essay maps the choke points, tallies their costs, compares U.S. practice with Europe and Latin America, and outlines a compact that restores the oldest freedom in political economy: the freedom to leave.
I. The New Fences
The modern labor market restrains mobility with four distinct devices.
Noncompetes. Roughly one in five U.S. workers—about 30 million people—are bound by noncompete clauses, including many in ordinary jobs. The Federal Trade Commission (FTC) estimated that eliminating these restrictions would raise workers' earnings by $400–488 per year on average, with aggregate gains approaching $300 billion annually.
No-poach and wage-fixing pacts. After the Justice Department (DOJ) and FTC's 2016 guidance warned employers that naked no-poach and wage-fixing agreements are per se illegal, federal and state enforcers pursued high-profile cases and settlements (alongside some trial setbacks), making clear that the labor market falls within the scope of antitrust.
"Stay-or-pay" training repayment agreements (TRAPs). These contracts obligate workers to reimburse "training" costs if they quit "too soon"—sometimes for generic onboarding. The NLRB General Counsel in 2024 urged the Board to treat many such agreements as unlawful restraints; regulators also flagged TRAPs as potential consumer-finance abuses. (In early 2025, an acting GC rescinded certain memos, underscoring that the legal landscape here is live and contested.)
Occupational licensing. Licenses now cover about one quarter of U.S. workers, a sharp rise from the mid-20th century. Research finds licensing reduces labor supply into affected occupations by 17–27% and dampens geographic mobility; the Obama White House and Brookings called out these frictions years ago, and the evidence has only grown.
Each fence stands alone; together they create a maze. A noncompete narrows the set of jobs you can take. A no-poach agreement removes the firm most likely to bid for you. A TRAP turns resignation into a debt event. A license locks a city's worth of skills behind a state line. The result feels like air that is technically present but hard to breathe.
II. What the Evidence Says (and Why It Matters)
Wages and bargaining. Empirical work finds that stricter noncompete enforcement depresses wages, especially for lower-wage workers; easing enforceability raises pay. The FTC's own economic analysis projected a 2–3% wage increase from a national ban. Complementary studies document wage growth when noncompetes are constrained and when employers cannot coordinate to avoid bidding wars.
Entrepreneurship and innovation. When employees can move, they create firms; when firms fear leaks, they invest in trade-secret protection rather than blanket restraints. Classic natural experiments (from Michigan to Silicon Valley analogues) show noncompete enforcement lowers worker mobility and weakens spin-out formation and VC activity; loosening enforcement does the opposite.
Geographic mobility. Licensing requirements that do not port across state lines reduce interstate moves and raise vacancies in shortage occupations. The U.S. Treasury, Labor, and CEA flagged this in 2015; subsequent research quantified substantial entry reductions (again, 17–27%). Mobility is a wage in slow motion; when it falls, so do lifetime earnings.
Distributional stakes. Restraints rarely hit the insulated. They bind new grads, immigrants, caregivers re-entering work, and younger workers who must move to climb. The quit elasticity literature shows employers wield more monopsony power over women and Black workers—making mobility barriers a multiplier of inequality.
Low-wage misuse. Public outrage over sandwich-shop noncompetes (Jimmy John's) crystallized the problem: when firms restrict the lowest-paid from working for a rival deli, the clause is not protecting R&D; it is suppressing wages. State AGs forced retreats and settlements.
III. The Law as It Stands (November 2025)
Federal rulemaking. In April 2024, the FTC issued a final rule banning most noncompetes nationwide. A Texas federal court then vacated the rule; the FTC withdrew its appeal in September 2025, effectively ending that path for now. The agency continues to pursue case-by-case enforcement where noncompetes operate as unfair methods of competition.
Congress. Multiple bipartisan bills would curb noncompetes, from narrow low-wage bans to the broader Workforce Mobility Act; none has cleared both chambers, but the coalition remains unusual and instructive.
States. A wave of state reforms has moved faster than Washington:
- Minnesota enacted a near-total ban (effective July 1, 2023), with narrow sale-of-business exceptions.
- California doubled down in 2024: SB 699 and AB 1076 make it unlawful to use or attempt to enforce noncompetes, even if signed out-of-state, and created a private right of action plus mandatory employee notice.
- New York's broad ban passed the legislature but was vetoed in December 2023; narrower approaches remain under discussion.
- D.C. now bars most noncompetes but allows limited covenants for high-compensated workers.
Licensing portability. On the licensing front, states have joined interstate compacts (nurses, physicians, teachers) and adopted "universal recognition" statutes; federally, the Military Spouse Licensing Relief Act (January 2023) mandates cross-state recognition for servicemembers and spouses who meet simple criteria—a small revolution for a population that moves often.
TRAPs. The NLRB GC's 2024 memo targeted "stay-or-pay" clauses; the CFPB spotlighted employer-driven debt as a consumer-finance risk. (Subsequent leadership shifts at NLRB trimmed back GC guidance in early 2025, underscoring that stable statutory rules—not memos—are the durable path.)
IV. Comparative Law: What Europe and Latin America Teach
Spain (and much of the EU). Post-employment noncompetes are lawful only with compensation (Spain's Workers' Statute, Art. 21) and strict durational caps: typically ≤2 years for technical roles, ≤6 months otherwise. Uncompensated covenants are void. This "pay-to-restrain" logic prices the employer's fear and preserves exit as a norm. The UK, meanwhile, has proposed—but not yet enacted—a three-month statutory cap.
Latin America. Mexico generally treats post-employment noncompetes as unenforceable in light of the constitutional right to work; Brazil enforces them only with substantial compensation (often pegged near a month of salary per month of restriction) and narrow scope. Colombia permits tightly tailored covenants with safeguards. The region's bias is clear: protect trade secrets and clients, but resist broad restraints.
Lesson. Most advanced regimes do not treat broad restraints as a cost-free default. They either ban them or force firms to internalize their cost—a market test disguised as labor law.
V. The Political Economy of Staying Put
Mobility is not a slogan; it's the worker's strike without a picket sign. When you can credibly leave, your current employer must treat you as a market participant, not an inventory item. Robert Reich has long framed this as a question of power: when exit paths narrow, bargaining collapses and "merit" becomes a story told by incumbents to themselves. That lens helps explain why noncompetes, licensing walls, and TRAPs show up as lower wages, fewer start-ups, and quieter workplaces.
The politics follow the map of assets and paperwork. Those best served by the current equilibrium—incumbent firms; occupations with strong boards; workers with the time and counsel to navigate waivers—argue that restraints protect investment, safety, and training. Those on the outside hear a different music: your ladder is our moat. The task is not to deny legitimate interests; it is to price and narrow them so the moat no longer floods the road.
VI. A Policy Architecture That Restores Exit
1. A federal floor on noncompetes. Congress should enact a bright-line rule: ban post-employment noncompetes for employees and contractors nationwide, with a narrow, paid exception tied to the sale of a business. This would lock in the FTC's core economic insight while ending years of forum-shopping and litigation roulette. Pair it with mandatory transparency: if any restraint is attempted, disclose it conspicuously at offer, in plain language, with a cooling-off period. (State models in Minnesota and California already point the way.)
2. Use less-restrictive tools to protect real assets. Strengthen trade-secret protections (DTSA) and NDAs; require employers to show why those tools won't suffice before any restraint on post-employment competition is allowed. If a firm truly needs a post-employment restriction, compensate the worker during the restraint—the EU/Spain logic.
3. Outlaw "junk" TRAPs. Ban training-repayment clauses for ordinary onboarding or compliance lessons. Allow narrowly tailored agreements only for bona fide, portable credentials with proration, caps, and no repayment if the employer terminates without cause. Encourage the FTC/CFPB to treat abusive TRAPs as unfair practices; build Board-level remedies into statute so swings in administrative policy do not unsettle rights.
4. Licensing reciprocity by default. Make universal license recognition the presumption: if you are competent in one state, you are competent in another, subject to background and discipline checks. Write into federal law what Congress already guaranteed military families in 2023—license portability—and condition certain federal funds on adoption. Expand interstate compacts and fund the IT rails to verify credentials in real time.
5. Sunlight and measurement. Require large employers to file anonymized statistics on the use of noncompetes, NDAs, non-solicits, and TRAPs; require licensing boards to report processing times and denial rates. Publish mobility indices by state and sector. If we measure exits, we will widen them.
6. Antitrust at the job-to-job margin. Make enforcement against no-poach and wage-fixing a priority equal to product-market cartels. Update penalties to reflect labor-market harm, and create safe harbors for worker mobility pacts—e.g., a mutual promise among local employers not to use noncompetes on sub-$150k jobs, enforceable under state unfair-competition law.
7. Immigration and recognition. For high-demand licensed fields, recognize foreign credentials more rapidly (provisional licensing with supervised practice), or fund bridge programs that convert credentials without multi-year detours. Licensing should filter incompetence, not immigrants.
VII. The Editor's Rule: Price the Restraint
Other countries force employers to pay if they want to restrict a former employee's competitive freedom, and to limit the duration severely. That is not hostility to business; it is the market speaking: if you truly need a restriction, write a check and keep it short. Spain's model—compensation plus caps—offers a template for a U.S. compromise should Congress balk at a full ban. Mexico's stricter baseline reminds us of a more elegant principle: work is not a franchise.
VIII. Case Studies (Short, Sharp)
Fast-food noncompetes (Illinois and New York, 2016). Public exposure and state litigation forced a national chain to abandon sandwich-maker noncompetes and pay for worker education. The episode illustrates how shame plus state law can move faster than federal process—and how low-wage restraints collapse under scrutiny.
Nurse, doctor, teacher compacts. The Nurse Licensure Compact (now most states), the Interstate Medical Licensure Compact, and the Teacher Mobility Compact prove that portability can be engineered without sacrificing safety. Processing time—not principle—is the bottleneck.
Military spouses (2023 federal law). Congress already recognized that mobility is not optional for some families. The portability mandate works; civilians deserve a cousin of the same right.
The FTC rule that wasn't (2024–2025). The agency built a meticulous record and a sweeping rule; litigation killed it for now. The lesson is constitutional as much as economic: durable reform needs Congress—or fifty experiments that converge.
IX. Answering the Standard Objections
"We need noncompetes to protect investments in training and R&D." Protect the training by paying for it and prorating repayment; protect the R&D with trade-secret and NDA law. Broad restraints are a blunt subsidy for firms that refuse to invest in retention.
"Licensing protects consumers." So do targeted tools: insurance, inspections, certification, bonding, and discipline. Where risk is real (a pilot, a surgeon), use hard licensing and compacts; where it is not (natural-hair braiding, low-risk services), use certification and disclosure. The empirical record shows blanket licensing suppresses entry without commensurate safety gains in many fields.
"TRAPs keep workers from gaming sign-on bonuses." Narrow, prorated repayment for genuine, portable credentials is defensible; blanket "training" debts for onboarding are pay-to-quit schemes that chill lawful mobility. Regulators have already put such debts on notice.
X. Metrics for a Mobility Republic
If policy works, we should see:
- Higher job-to-job moves and faster wage growth at job change.
- Shorter vacancy durations in shortage occupations (a licensing-portability dividend).
- More firm births in non-compete-ban states relative to peers.
- Reduced TRAP incidence and average repayment amounts, especially in low-wage sectors.
- Convergence in mobility across demographic groups (closing the quit-elasticity gap that mirrors power differentials).
We should also see something subtler: an uptick in ordinary courage. Exit makes voice safer. When workers can leave, they can also speak—about safety, bias, and waste—without rehearsing the price of silence.
XI. Coda: The Turnstile Test
A healthy labor market is not a fortress; it is a station with turnstiles that click easily in both directions. Employers compete for people; people compete for employers; and the country grows through the friction of that traffic. Noncompetes that sprawl across job families, licensing that won't cross a river, debts that punish resignation—these are the small closures that add up to a slow economy and a cramped democracy.
The fix is neither romantic nor radical. Ban broad restraints. Price and cap the narrow ones. Port credentials. Shame the colluders. Publish the data. The rest will be done by workers and firms once we give them back what the market requires and the law once knew: the right to walk through an open door.
References
Federal regulation and enforcement:
- FTC, Final Rule on Non-Compete Clauses (April 2024); Texas court vacatur; FTC appeal withdrawal (September 2025).
- DOJ/FTC Antitrust Guidance for Human Resource Professionals (2016).
- FTC estimated wage gains from noncompete elimination ($400–488/worker; ~$300B aggregate).
Labor market restraints:
- NLRB General Counsel Memo on Training Repayment Agreement Provisions (GC 25-01, January 2024); subsequent rescission updates (2025).
- CFPB, Consumer Financial Protection Circular on employer-driven debt (2023).
- EPI, monopsony power and quit elasticity differentials by demographics.
Occupational licensing:
- U.S. Department of the Treasury, Council of Economic Advisers, and Department of Labor, "Occupational Licensing: A Framework for Policymakers" (July 2015).
- Blair, Peter Q. & Bobby W. Chung, "How Much of Barrier to Entry Is Occupational Licensing?" British Journal of Industrial Relations (2019); "Job Market Signaling through Occupational Licensing" NBER Working Paper (2018).
- White House/Brookings reports on licensing frictions.
State reforms:
- Minnesota Statutes § 181.988 (2023 noncompete ban).
- California SB 699 and AB 1076 (2024 expansions; private right of action; out-of-state enforceability bar).
- New York veto (December 2023); D.C. Code § 32-581.01 et seq.
- Military Spouse Licensing Relief Act (January 2023).
International comparators:
- Spain, Workers' Statute (Estatuto de los Trabajadores), Article 21 (post-employment noncompetes with compensation and durational caps).
- UK government consultation on three-month noncompete cap (proposal status as of 2025).
- Mexico constitutional right to work and enforceability doctrine.
- Brazil and Colombia compensation and scope requirements.
Case law and enforcement:
- Jimmy John's settlement with Illinois and New York AGs (2016).
- DOJ labor-market antitrust cases (hospital wage-fixing, tech no-poach settlements).
Academic and empirical evidence:
- Marx, Matt, et al., studies on noncompete enforcement and entrepreneurship/mobility.
- Starr, Evan, et al., noncompete incidence and wage suppression.
- Johnson & Kleiner, licensing and interstate mobility reductions.
Policy framing:
- Reich, Robert, public commentary on labor market power and exit.
- Interstate compacts for nurses, physicians, teachers (NCSBN, IMLC, NASDTEC).
(Figures and legal statuses reflect developments current to November 7, 2025.)