Why a State Move Can Break Career Continuity
This series examines the relationship between state-based occupational licensing and interstate career mobility. The specific question is not simply whether licensed people move less often than everyone else. It is whether crossing a state line can make it harder for a worker to continue the same occupation—and how large that effect is once the other costs of moving are taken seriously. The answer matters to anyone whose career depends on a license, but also to employers, households, and policymakers deciding whether professional credentials should travel more easily.
A move can look like a change of address in a population table while being a much larger transition in a worker's life. A nurse, electrician, teacher, pharmacist, or lawyer may arrive in a new state with the same education and experience but still face a new board, exam, application, or period of uncertainty before being allowed to practice. In that interval, a job offer can expire, savings can run down, or a household can decide that the move is no longer worth it. Yet it would be just as misleading to treat every lower migration rate among licensed workers as proof that licensing is the cause. Some occupations depend on local clients or seniority. Some workers are more attached to their home state. And moving a household is expensive even when no credential is involved.
The useful problem, then, is one of career continuity. What has to travel with a worker for a move to preserve access to work, and what has to be rebuilt at the destination? The evidence below supports a way to answer that question without turning a conditional finding into a universal rule.
The border can change access to work
Occupational licensing is a government requirement that a worker meet specified conditions before practicing an occupation legally. In the United States, most licensing is administered at the state level. The Johnson–Kleiner study describes licensing as covering about 5 percent of the workforce in 1950 and roughly 25 percent in the modern period. The Bureau of Labor Statistics article reports that more than a quarter of workers are in occupations regulated by state law and emphasizes how widely requirements can vary.
Those differences are not merely administrative trivia. The BLS article gives a striking comparison: Michigan requires three years of education and training to become a licensed security guard, while some other states require only 11 days. It also reports that average training requirements among low-income licensed occupations range from 724 days in Hawaii to 113 days in Pennsylvania. The exact examples concern initial licensing, not necessarily the requirements imposed on an experienced worker transferring a license. They nevertheless show why a state border can change the value of a worker's existing preparation.
Licensing can serve legitimate purposes. Minimum qualifications may protect clients and patients, and a common standard can make a service more trustworthy. The question here is not whether all licensing is harmful. It is whether a system designed around state jurisdictions can make a worker's qualification less portable than the work itself. A worker may know how to perform the job, yet still need permission to perform it in the new place.
That distinction separates two kinds of mobility. A person can move within a local labor market and keep the same clients, employer network, and professional standing. A person can also move far enough to enter a different labor market, where the old relationships no longer produce income. Crossing a state border often adds a legal question to that economic transition: can the existing license be recognized, or must the worker qualify again?
The difference is easy to miss when migration is measured only as a change in residence. A table that counts everyone who moved between states does not reveal whether a worker delayed the move because of licensing, changed occupations after arriving, or accepted a lower-quality job while waiting for approval. Nor does it reveal the people who never moved because the combined cost seemed too high. Career continuity is partly an observed outcome and partly a missing counterfactual.
A move has more than one career cost
A simple way to understand the problem is to separate three costs that can arrive together.
The first is the ordinary cost of relocation: transportation, housing, job search, and the psychological and family costs of leaving familiar surroundings. These costs affect licensed and unlicensed workers alike. A destination with higher wages may still be unattractive if housing is expensive or a partner cannot find work.
The second is the loss of local career capital. Lawyers, real-estate professionals, barbers, veterinarians, and some health-care professionals may depend on reputation, referral networks, repeat clients, or knowledge of local institutions. Teachers may lose seniority or benefits tied to a school district. A long-distance move can destroy some of that value even when the worker's formal qualification transfers perfectly. The worker is not merely carrying a skill to a new market; they are rebuilding the conditions under which the skill produces income.
The third is re-licensure. A new state may recognize an existing credential, require an endorsement application, demand an additional exam, or impose other conditions before the worker can practice. The burden can be low, high, or uncertain. It can also vary by occupation and by the worker's experience. A national examination may reduce the need to repeat a test, but it does not mean that every other part of the process is identical across states.
The researchers use this distinction to classify occupations rather than to claim that the categories capture every rule. Among 22 occupations licensed in every state and identifiable in the American Community Survey, some require exams whose content or difficulty varies by state. These are treated as state-specific occupations. Others require national exams with a common passing standard and are treated as quasi-national occupations. Lawyers, for example, take a state bar examination; nurses take a national licensing examination. Pharmacists and veterinarians can have both a national exam and a state-specific exam, so the presence of the latter places them in the state-specific group.
This classification is useful because it gives researchers a concrete way to compare likely transfer burdens. It is also limited. The exam does not capture every training, experience, endorsement, or board-discretion rule. A quasi-national license is not necessarily frictionless, and a state-specific exam is not the only reason an occupation may be locally rooted.
The three costs also explain why a raw migration gap is hard to interpret. Suppose state-specific occupations show a lower interstate migration rate than occupations with national exams. That pattern could reflect the additional exam. But it could also reflect greater dependence on local clients, a stronger preference for stable employment, or a tendency for people who want to remain near family to choose an occupation whose training is tailored to their home state. If these characteristics are not separated from licensing, the observed difference is a mixture of mechanisms.
The location of the move matters for the same reason. A short move may leave an occupation's labor market and clientele largely intact. A 50-mile move is not a perfect definition of a new labor market, but it gives researchers a way to focus on people who have already accepted a substantial relocation cost. Comparing whether that move crossed a state line is different from asking whether licensed workers are willing to move at all.
The model is therefore not a claim that licensing is always decisive. It is a map of where licensing can enter a decision that already contains other risks. It also predicts why an effect might be concentrated in particular occupations rather than spread evenly across the labor market.
The evidence does not ask just one question
Two studies make the distinction between individual careers and aggregate migration especially important. Johnson and Kleiner examine 22 universally licensed occupations and use the difference between state-specific and quasi-national exams as a proxy for the difficulty of transferring a license. They focus on people who made a long-distance move and ask how often that move crossed a state line. In their preferred comparison, workers in state-specific occupations were about 7 percent less likely to make an interstate move than workers in quasi-national licensed occupations. That is a conditional relative estimate, not a claim that licensing prevents most moves.
The BLS article by Thomas J. Cooke, Mark Ellis, and Richard Wright asks a broader question. It models the volume of migration from each state to every other state using overall state licensing rates, distance, contiguity, labor-market conditions, income, amenities, and state-specific factors. Across several population groups, the aggregate licensing measure has no detectable effect on interstate flow volume in that model.
At first glance, the findings seem to point in opposite directions. One says that higher transfer costs are associated with less interstate mobility; the other says that differences in state licensing rates do not explain state-to-state flows. But they do not measure the same thing. The first study compares workers in selected licensed occupations after they have already made a substantial move. The second compares aggregate flows among broad groups and asks whether a state's overall licensing penetration helps predict where people move. One is close to a career-portability question; the other is a question about the relative importance of licensing among many forces that shape migration.
That difference is the central discipline for reading the evidence. A licensing rule can be a real obstacle for a teacher, pharmacist, or lawyer deciding whether a particular opportunity is worth pursuing, while being too heterogeneous or too small to move the national migration rate. Conversely, a near-zero aggregate coefficient cannot establish that no worker faces a costly transfer process. It shows that aggregate licensing rates are not a dominant explanatory factor in the specified flow models.
The next step is to examine how the individual-level study turns this broad mechanism into a comparison, why its preferred estimate is much smaller than the raw migration gap, and which careers show the clearest remaining penalty. Only then can the aggregate result be placed beside it without losing the worker whose career is the subject of the question.