The Permission Slip You Need Before You Can Braid Hair
Roughly a quarter of American workers need a government license to do their job, up from about one in twenty in the 1950s. The stated purpose is safety, and the measurable effect is a barrier to entry.
A Quiet Expansion
In the early 1950s roughly one in twenty American workers required a state license to work legally in their occupation. That figure now sits near a quarter of the workforce. The change did not come from a single decision. It accumulated occupation by occupation, state by state, over seventy years.
Licensing is distinct from two adjacent things it is often confused with. Certification is voluntary and signals competence without restricting practice. Registration requires notifying the state but does not gate entry. Licensing makes it illegal to perform the work without permission, which is why it is the version with economic teeth.
The Economic Mechanism Is Not Complicated
Restricting who may supply a service reduces supply. Reduced supply raises the price of the service and raises the wages of those permitted to provide it. This is not a controversial prediction, and empirical work generally finds it.
Estimates of the wage premium associated with licensing commonly land in the range of ten to fifteen percent relative to comparable unlicensed workers, after controlling for education and experience. Prices for licensed services rise correspondingly. The disagreement in the literature is not about whether these effects exist but about what is purchased with them.
| Group | Effect of Licensing |
|---|---|
| Incumbent practitioners | Higher wages, less competition |
| Consumers | Higher prices, fewer providers |
| Aspiring entrants | Training cost and time before earning |
| Workers moving between states | Re licensing burden |
The Quality Question
The justification for licensing is consumer protection in markets with severe information asymmetry. A patient genuinely cannot evaluate a surgeon before the operation, and the consequences of error are irreversible. Almost nobody argues against licensing physicians.
The difficulty is that licensing expanded far beyond that category. Requirements exist in various states for interior designers, florists, shampooers, auctioneers, and travel guides, occupations where the information asymmetry is mild and the harm from poor service is a bad haircut rather than a fatality.
Where researchers have looked for quality improvements from licensing in lower risk occupations, results have been weak and inconsistent. Studies of licensed and unlicensed markets for similar services frequently find no detectable difference in complaint rates or outcomes, while consistently finding higher prices.
The case for licensing rests entirely on the severity of the harm a bad practitioner can cause. Where that harm is small and observable, the license is mostly a price. Where it is large and hidden, the license is doing real work.
Who Writes the Rules
Licensing boards are typically composed largely of practitioners from the licensed occupation. This is defensible, since expertise is required to judge competence, and it also places the decision about how many competitors to admit in the hands of the existing competitors.
The Supreme Court addressed this directly in a case concerning a state dental board composed mostly of practicing dentists that had acted to exclude non dentists from teeth whitening services. The Court held that a state board controlled by active market participants is not automatically immune from antitrust law unless it is actively supervised by the state. That ruling did not end self regulation, but it established that a board of incumbents restricting competitors is subject to the same scrutiny as any other trade association doing so.
The Mobility Cost
The most concrete harm, and the one with the broadest political agreement, is interstate mobility. Licenses are issued by states and requirements differ, so a licensed cosmetologist, nurse, or electrician moving across a state line may face additional coursework, examinations, fees, and months of lost earning.
This falls hardest on populations who move involuntarily and frequently, most visibly military spouses, and it degrades labor market flexibility generally. It is also the area where reform has actually advanced, through interstate compacts recognizing licenses across member states in nursing, medicine, and several other fields, and through universal recognition laws under which a state accepts an out of state license outright. These reforms have drawn support across the political spectrum, which is rare.
The Reform Debate Is About Calibration
The productive version of this argument is not licensing versus no licensing. It is about matching the instrument to the risk. Serious proposals include applying a genuine cost benefit analysis before creating a license, requiring periodic review that sunsets requirements failing that test, substituting certification where the harm is reputational rather than physical, reducing training hour requirements that exceed any demonstrable safety need, and removing licensing barriers for people with unrelated criminal records where the connection to public safety is absent.
The obstacle is straightforward political economy. The benefits of a license are concentrated on an organized group of practitioners who will show up to a hearing, and the costs are spread thinly across consumers and hypothetical future entrants who will not.
The Bottom Line
Occupational licensing is the clearest everyday example of concentrated benefits and diffuse costs producing durable policy. It works where the information asymmetry is severe and the harm is irreversible, and it functions mainly as a price increase where it is not. The reform with the strongest evidence and the least opposition is interstate recognition, because it reduces the deadweight cost without touching the safety rationale at all.