The license you need to braid hair and the jobs it quietly kills

Pop quiz. What do a hair braider, a florist, an interior designer and a person who wants to give guided tours of a city have in common? In some American state, at some point, each of them has needed the government's permission before earning a dollar at the trade. Not a permit for a loud noise or a health inspection for a kitchen, mind you. A license, with its fees, its exams and, often, its hundreds or thousands of hours of required training.

Let that sink in, because we are not talking about brain surgery. Braiding hair is a skill that people have passed down in families for generations. It involves no chemicals, no scissors and no heat. And yet in the not-so-distant past, a woman in one state or another who wanted to braid hair in her kitchen for a neighbor's daughter could be told she needed to attend a cosmetology school and learn to do things she would never do in her business, such as dye, perm and cut.

I think this is one of the great hidden scandals of the modern economy. It is not hidden because anyone conceals it. It is hidden because no single license looks outrageous, the people hurt by it are poor and have no lobbyist, and the people helped by it are well organized and know exactly what they are doing. If you want to understand why the job ladder in America has some missing rungs, begin here.

How a handful of professions became a quarter of the workforce

Licensing is old. For centuries governments protected certain callings, and in America the early licensed professions were the ones everybody understood: doctors, lawyers, and a short list of others where a mistake could kill or ruin you.

The Supreme Court spoke to the question in Dent v. West Virginia in 1889. West Virginia had required physicians to hold a certificate from a state board. Dr. Dent, who had practiced without one, argued that the requirement deprived him of his right to pursue his profession without due process of law. Justice Field, writing for the Court, upheld the law. The Court said that a state may prescribe the qualifications necessary to practice medicine, because the public is entitled to protection from "ignorance and incapacity." The case endures as the foundation of the modern licensing power. It also contains a line that cuts both ways: the right to pursue a lawful calling is real, and a state may regulate it only so far as is reasonable. For physicians, that was easy to see. The question is whether the same logic reaches a person who braids hair.

Over the twentieth century, the list grew. Economists who study the question, among them Morris Kleiner of the University of Minnesota and, in collaboration with him, the late Alan Krueger of Princeton, used survey data to estimate how large the licensed share of the workforce had become. Their work found that it rose from a small fraction of workers in the early 1950s to something on the order of a quarter or more of the workforce in the 2000s. A 2015 report prepared by the Obama administration's Treasury Department, Council of Economic Advisers and Department of Labor reached a similar conclusion, estimating that the share had risen from about 5 percent in the 1950s to nearly 25 percent. I am citing a Democratic administration on purpose. On this subject, the evidence has not been a partisan matter.

The braiding cases

The Institute for Justice, a public-interest law firm founded in 1991, has made hair braiding one of its signature fights. It has represented braiders in several states, arguing that requiring cosmetology training for a skill that does not use the dangerous tools of that trade violates the Constitution's guarantee of due process and equal protection. Two cases give the flavor.

In Utah, a federal court heard Clayton v. Steinagel in 2012. Jestina Clayton, a Utah mother, braided hair at home, and the state said she needed a cosmetology license. The legislature then changed the law. In Texas, a federal court ruled in Brantley v. Kuntz (2015) that applying the state's barbering regulations to African hair braiding was unconstitutional as applied. These cases did not make every state follow; many states had already carved out braiding, and others changed their laws afterward. The state-by-state situation moves, so please check your own state's board and statute rather than rely on what I say. My point is that it took lawsuits and legislative campaigns, over years, for a woman to be allowed to braid hair without attending a school for a different trade.

The same question has played out in the red states, where governors and legislators have been at the front of the reform. You Need the Government's Permission to Braid Hair? Licensing Gone Wild covers the state-level story, and the pattern is a good example of what the states-as-laboratories idea, discussed in Brandeis Called the States Laboratories and Conservatives Took Him at His Word, can do.

The best case for licensing

I promised to state the opposing argument at its strongest. Here it is.

Customers cannot judge quality in advance. When you hire a barber, an electrician or a dentist, you can't see the wiring in the wall or check the sterilization of the tools. In 1970 the economist George Akerlof described the logic in a famous paper about used cars, "The Market for Lemons": when buyers can't tell good from bad, bad sellers drive out good ones, and the whole market suffers. A license is a signal that the practitioner has met a standard. And in some trades the risks are real: a poorly wired house can burn, a careless stylist using strong chemicals can burn a scalp, and an unlicensed practitioner can spread infection. A license protects the public from the worst.

That is a serious argument, and in some professions it wins. I do not want an unlicensed man wiring my house, and I am not asking anyone to repeal the requirement that surgeons be trained. The real question is not whether licensing is ever justified. It is whether a specific license fits a specific risk. The reply to Akerlof is that there are many ways to solve an information problem that cost less than a ban on entry. Let me give you the reasons the license fails the test in a trade like braiding.

Five answers

One: the rule does not match the risk. If the worry is infection or burns, require what addresses those risks: clean tools, a posted sanitation standard, a health inspection, a complaint process. A cosmetology course teaches chemical treatments and cutting that a braider never performs. A rule that requires a braider to master a different, much larger trade is a license for the incumbents, not a safeguard for the customer.

Two: the cost lands on the people with the least money to spare. Think about what a license requires: tuition, exam fees, travel to a testing site, and the hours themselves. Here is a hypothetical calculation. Suppose a state requires 1,500 hours of training, the school charges $15,000 in tuition, and the person could otherwise earn $12 an hour. The lost wages alone are 1,500 hours times $12, which is $18,000. Add the tuition and the person has put $33,000 of cost, in cash and forgone pay, in front of the first paying customer. A wealthy applicant can absorb that. A single mother with two children, or an immigrant with no savings, simply cannot. The barrier does not shut out the unqualified. It shuts out the poor.

Three: the people writing the rules are the people already in the trade. Licensing boards are usually dominated by working members of the licensed profession. They decide who may join their own competition. The Supreme Court recognized the danger in North Carolina State Board of Dental Examiners v. FTC (2015), holding that a board controlled by active market participants cannot claim immunity from the antitrust laws unless the state actively supervises it. The case involved dentists moving to stop non-dentists from whitening teeth. You do not need to believe anyone is wicked to see the incentive: every practitioner has a reason to prefer fewer practitioners.

Four: the evidence of benefits is thin. In a well-known study of dentistry, Kleiner and Robert Kudrle compared states with stricter and looser licensing and found that tougher rules raised the price of dental care for consumers but did not produce better dental health outcomes. More broadly, the research Kleiner and others have done finds that licensing is associated with higher wages for those inside and higher prices for customers, with no consistent improvement in the quality of service. I would not claim the evidence is airtight for every occupation. I would say that the people who defend a license should be asked for the evidence that it improves safety, the way a drug company is asked to prove its drug works.

Five: it is a barrier to mobility. A licensed nurse, cosmetologist or electrician who moves across a state line, say with a spouse who serves in the military, may find that the new state does not recognize the old license. She has to start over. That hurts families, and it has an oddly large effect on the labor market in a country where people move for jobs.

The reform that is spreading

You do not have to throw out every license to fix this. There are ways to reduce the barrier while keeping the protection that matters.

1. Universal recognition. In 2019 Arizona became the first state to enact a law recognizing occupational licenses from other states for people who move there. In general, the law lets a worker who has held a license in good standing elsewhere for a period of time get an Arizona license without repeating the training. A number of states have adopted versions of it since. This is exactly the kind of idea the states-as-laboratories theory predicts: one state tries it, the economy does not collapse, and others copy it.
2. Certification instead of licensure. The state can let anyone work but give a formal title to those who pass a test, so that customers have the signal Akerlof wanted without a ban on everybody else.
3. Registration or bonding. Requiring a worker to register, carry insurance or post a bond protects customers if something goes wrong.
4. Narrow the scope. Allow the braider to braid without a cosmetology license. Carve out the tasks that carry no real risk.
5. Regular review. Subject every license to periodic review, with the burden on the profession to show what harm the license prevents.

Milton Friedman raised the whole question in his 1962 book Capitalism and Freedom, which includes a chapter on occupational licensure and argues that licensing is often a means by which a trade restricts competition. You can read his case, written in the plain language of a man talking to his neighbors, in Milton Friedman Took Free Markets Out of the Classroom and Into the Kitchen.

What you can do about it

The good news is that this is a fight ordinary citizens can win, and they have. Licensing is a state and local matter, which means a letter to a state legislator carries weight, and a person who testifies at a hearing for three minutes is often the only unpaid voice in the room. how a bill becomes a state law, and where a citizen can step in explains the path of such a bill. And if you have been stopped by a licensing rule yourself, tell your story. A real braider in a real kitchen is worth more at a hearing than an expert in a suit.

The same instinct that says "count the cost" in Regulation protects us, so why count the cost? Because someone pays it applies here: ask what the rule prevents, what it costs, who pays, and whether a smaller step would do. And for the larger question of rules that gate access to housing rather than work, zoning may be the biggest price control you never got to vote on picks up the same thread.

The bottom line

• A license is the right tool for work that can seriously harm a customer. It is the wrong tool when the rule bears no relation to the risk, as with a braider forced to learn a different trade.
• Licensing has grown from a small share of workers in the 1950s to roughly a quarter of the workforce, by estimates from economists and from an Obama-era administration report. The fight over it is not strictly partisan.
• The people most hurt are those with the least money and time. In our hypothetical, a 1,500-hour requirement costs $33,000 in tuition and lost pay before the first customer.
• Support universal recognition, certification in place of licensing where it fits, narrow scopes and regular review. Arizona's 2019 law shows it can be done.
• Check your own state's board and statute, and speak up at a hearing. This one is won locally.