What new data tell us about the growth of occupational licensure

Many of us are required to maintain licenses in order to practice our occupations legally. In recent years, 22 percent of the employed have reported holding a government-issued license, in professions ranging from physical therapy to cosmetology to public school teaching. Because occupational licenses can be costly in time and money to obtain, licensure matters for how the labor market performs and who can access economic opportunity.

According to new data and research on the shifting landscape of occupational licensure, both the share of workers who have a license and the number of licensed occupations appear to have stabilized in recent years, after several decades of growth. That earlier growth in worker share was powered both by the rising share of employed people who are in already licensed occupations and the expanding list of occupations that states are licensing—delicensure, by contrast, being quite rare.

The data underlying these findings are part of an expansion of the Occupational Licensing Dashboard, an interactive tool from the Federal Reserve Bank of Minneapolis. From its initial release in 2023, the dashboard has enabled users to explore the characteristics of licensed workers and their occupations but not the details of licensure policy. Newly revamped and extended, the dashboard now makes it easy to see where and when specific occupations were licensed. This extension is only possible because of pioneering work by Nicholas Carollo, a senior economist at the Board of Governors of the Federal Reserve System who assembled the policy data and conducted much of the research described in this article.

Information about licensure is useful not just for researchers but for practitioners, policymakers, and the public. Information about licensure policy is especially valuable because it can help groups such as state legislators understand the broader regulatory environment as they assess the need for new or existing licensure in a particular occupation. For example, states often assess proposals for new licensure, a process typically referred to as “sunrise review.” Important context for such a decision includes the licensure landscape outside the state and the labor market outcomes of workers in relevant occupations. Ultimately, this kind of information can help policymakers evaluate and balance licensure’s potential benefits, such as improved public health and safety, with its costs, such as higher prices for services and fewer opportunities for unlicensed workers.

Making this information easily accessible to policymakers and members of the public supports the Minneapolis Fed’s mission to pursue an economy that works for all of us, and it aligns with our Community Development and Engagement team’s specific mission to advance the economic well-being of low- to moderate-income individuals, households, and communities in the United States.

How to navigate the licensing dashboard

When a user launches the Occupational Licensing Dashboard, they’ll find information about a random occupation presented onscreen, with a map of the United States on the left side and a chart on the right side. The U.S. map shows the state-by-state breakdown of occupational regulation in the most recent year for which data are available. Dark blue represents licensure—that is, it designates states that require a government-issued credential in order to legally practice the occupation in question—while other colors represent registration and certification. These other types of regulation are less stringent than licensure and can be alternatives for policymakers to consider. On the right side of the dashboard, the chart shows the number of states using a particular regulation in a given year.

Consider physical therapist assistants, for example. As of recent years, all 50 states and Washington, D.C., license the occupation; if a user selects it from the “Choose an occupation” drop-down menu in the upper left of the dashboard, the map will show dark blue coloring of every state.¹ But the chart shows that widespread licensure of this occupation has only recently been the case. In 1967, Florida became the first state to license the occupation, as a user can discover by hovering over that state on the map. In 2014, Hawaii became the last state to license physical therapist assistants.

Many occupations are currently licensed in some but not all states. For example, 35 states and Washington, D.C., license electricians. That count is up from 30 in the 1990s and fewer than 10 at midcentury. However, licensure data can be messy. In the case of electricians, this is true in a couple of different ways. First, an occupation might be licensed in some narrow ways but not across the board. For example, Ohio requires licensure for electricians who are acting as commercial contractors, but not more generally. For that reason, we exclude Ohio from the group of states that require licensure. Second, states are not the only governments that can require licenses. Municipal governments sometimes have their own licensure requirements, for instance. While we do not track local government licensing requirements in the dashboard, we do provide information on federal licenses, which can be accessed through the “Federal Policies” tab at the top right of the dashboard.

Next to the tabs for “State Policies” and “Federal Policies” is a tab labeled “Workers.” Here a user can again see a U.S. map on the left side of the dashboard, but one with a different purpose. Rather than showing policy information, this part of the dashboard shows survey-based data about licensed workers. These data are intended to provide context for licensure policy: How many workers report holding a license in a given occupation? How much are workers in an occupation earning? What is the demographic breakdown of an occupation?

By clicking through a drop-down menu on the top left, a user can again choose from a variety of occupations.² Just below that menu is another drop-down that enables the user to select a metric of interest. This menu defaults to displaying the licensure rate but offers two other metrics as well. On the right side, a user can click the “Licensed worker characteristics” and “Occupation characteristics” bars to display charts that show how the selected metric has evolved since 2016, which is shortly after the U.S. Census Bureau’s Current Population Survey began featuring questions about occupational licensing. An additional drop-down menu in the top center of the dashboard enables the user to adjust which states’ metrics the charts display.

How and why licensure has evolved over the years

By one estimate, the share of workers who were licensed in the 1950s was only about 5 percent. At the time, most people worked jobs that tended to be unlicensed: almost half of the employed were working in agriculture, mining, construction, or manufacturing. In the decades that followed, the share of licensed workers rose to 22 percent, and it’s stayed roughly at that level in recent years (2016−2025), when it’s been measured most accurately and consistently. The rise in the share of licensed workers over time was driven by two factors: occupations became newly licensed in many states, and employment shifted toward the service sector (Carollo et al. 2025).

Figure 1 shows the first of those two factors. It displays the number of times states newly licensed (or certified, or registered) an occupation that had not been regulated in the same way the previous year.³ The figure shows that new licensure accelerated in the 1970s and remained high until the 2010s, subsequently falling to a much lower level. Certifications, whether state or private, and registration were introduced in much smaller numbers through the whole period.

When deciding whether to license an occupation, states appear to be looking to each other: the decisions of adjacent states and a few bellwethers like California, New York, and Texas all have predictive power for when a given state licenses an occupation. Professional associations also matter. Once an association is organized in a state, licensure or other forms of regulation (registration or certification) become dramatically more likely. When the tasks in an occupation become more complex, the existence of professional associations themselves may become more likely; greater task complexity is associated with a higher likelihood of regulation. Interestingly, yet another factor is whether an occupation is exposed to competition from immigrants, with licensing more likely to occur in states that were more exposed.⁴ Prior work by Minneapolis Fed researchers found that licensing disproportionately reduces employment of foreign-born workers.

Regardless of the reason for the enactment of occupational licensure, one pattern in the data stands out: delicensure is rare. An occupation that started out licensed in any given year from 1950 to 2020 remained licensed 99.9 percent of the time in the next year. By contrast, the annual probability that an unlicensed occupation became licensed in the subsequent year was between 1.1 percent (for occupations initially subject to registration) and 3.6 percent (for occupations initially certified by the state) over the same time frame. When interpreting these calculations, it’s important to understand that our data do not contain every occupation in the labor market. Rather, the data only include occupations that at some point in history were regulated by at least one state.

Figure 2, which pools all instances of state-level licensure or delicensure by decade, shows a similar pattern to that shown in Figure 1. Throughout the second half of the twentieth century and into the 2000s, the share of occupations (in a given state) that transitioned into licensure kept rising, to a high of 3.6 percent in 2001–2010. That share declined somewhat in the 2010s, when the rate of new licensure was 2.4 percent. By contrast, the highest-ever delicensure rate was only 0.2 percent, in the 1970s. The rate fell to a low of virtually zero in 2001–2010.

Informing the future of licensure

As policymakers and researchers grapple with licensure policy, more and better data are crucial. The Minneapolis Fed will continue to regularly update worker-survey data in its licensing dashboard. Researchers at the Fed will also look to answer the important questions that arise about licensing and its implications for workers. Particularly for low- and moderate-income workers, licensure requirements can constitute barriers to employment and economic success, making it essential that policymakers have all the information they need to balance these costs with the public health and safety benefits.

The authors thank Nicholas Carollo for reviewing an earlier draft of this article.

Endnotes

1 As of this writing in December 2025, the data presented in the dashboard and used in this article extend only through 2020.

2 The occupations shown on the “Workers” tab are different in many cases from the occupations shown on the “State Policies” and “Federal Policies” tabs. This is because the former are harmonized Census occupations (in other words, they are adjusted for changes in occupational definitions over time), devised by IPUMS and based on underlying Census occupation codes. By contrast, the latter are occupations that track definitions in state and federal law.

3 Each line shows five-year pooled samples, which help to smooth through year-to-year variation in the data that is due to sampling variability.

4 With the exception of the last sentence, the findings in this paragraph all draw from Carollo et al. (2025).