How occupational licensing limits access to jobs among workers of color

Occupational licensure—the legal permission to work in certain occupations—is typically justified in terms of its effects on public health and safety, but it also matters for workers’ access to jobs and economic opportunity. Nearly one in every four workers across the labor force has obtained a license, which is necessary for individuals pursuing occupations as varied as plumbing, dentistry, nursing, and practicing law. The process requires (to a greater or lesser extent, depending on the job and the state) substantial investments of time and money.

Workers who secure a license tend to receive higher wages than similar unlicensed workers and may also enjoy more job security (as suggested by longer tenure and lower unemployment). However, these benefits result in large part from the exclusion of other workers from licensed work. This is especially true among workers of color, who are less likely to be licensed than White workers, even after taking into account differing levels of education. Unnecessarily strict licensure requirements that adversely affect immigrants and people with criminal records may contribute to the lower levels of licensure among workers of color. Policy reforms could help reduce racial exclusion from licensed work and economic opportunity.

Workers of color are much less likely to be licensed

In the United States, nearly one in every four workers is licensed. However, this rate is not constant across groups: workers of color are substantially less likely to be licensed relative to White, non-Latino/a workers. (See Figure 1.) The disparity is largest for Latino/a workers, who are about half as likely (11 percentage points less) to be licensed than White, non-Latino/a workers. Asian, Black, and American Indian and Alaska Native workers are also less likely than White workers to be licensed, by 6, 5, and 4 percentage points, respectively.¹

Across most of the states in the Ninth Federal Reserve District, workers of color similarly have lower rates of licensure than White workers. (See Figure 2.) In some cases, the disparities are even larger than those nationwide. For example, the disparities in licensure between Asian and White workers in Minnesota and North Dakota are nearly twice as big as those at the national level. In Montana and South Dakota, the disparities between Black and White workers are again more than twice as large as the disparities nationally. Latino/a workers in Minnesota see an even larger gap (16 percentage points) in their licensure rates compared to White workers. One exception to these trends is Wisconsin, where Black and White workers have similar rates of licensure.²

Differences in educational attainment don’t account for the gaps

One possible factor that can account for racial disparities in licensing is differences in educational attainment. The type of work that is licensed often comes with educational standards that would exist even in the absence of licensing. For example, if licensure for attorneys were abolished, aspiring lawyers would likely still need to complete undergraduate and professional degrees to be hired or to attract clients. However, the educational or training requirements for licensure in some occupations go beyond what would be typical in the absence of licensure and beyond what is necessary to protect public health and safety. (See Greenberg 2021, Kleiner and Soltas 2019, and White House 2015 for more on this.) The interaction between licensing and education is therefore complicated and can vary from occupation to occupation.

With that said, we find that White workers are overrepresented in licensed work across every level of educational attainment.³ (See Figure 3.) For example, among those with only a high school degree, White workers make up 52.0 percent of all unlicensed workers but 64.0 percent of licensed workers. Conversely, Latino/a workers with only a high school degree or less make up a smaller share of licensed workers (17.1 percent) than of unlicensed workers (29.1 percent) in that education group. Asian workers are particularly underrepresented at higher education levels; those with more than a four-year degree make up 16.6 percent of all unlicensed workers with that level of schooling but only 7.5 percent of licensed workers.

Workers of color remain less likely to be licensed compared to similar White workers, even after controlling for educational attainment, age, and gender. After those adjustments, the rates of licensure are 10 percentage points lower for Asian workers, 5 percentage points lower for Latino/a workers, 3 percentage points lower for Black workers, and under 1 percentage point lower for American Indian and Alaska Native workers (though this last difference is not statistically significant) than for comparable White workers.⁴

Differences in education, age, and gender account for a little over half of the Latino/a-White disparity in licensing but less of the disparity for Black workers. The Asian-White disparity actually grows after adjusting for those differences, implying that the gap would be larger if Asian and White workers had the same observable characteristics. However, we do not know to what extent particular licensing barriers are the causes of the licensing-rate disparities that remain after the adjustments we make. We investigate two such barriers below.

Credentials of foreign-born workers often don’t count toward licensure

If education differences cannot account for racial disparities in licensure, what can help account for them? One possibility is that foreign-born workers—82.4 percent of whom are people of color—face an uphill struggle in becoming licensed. Training, education, and credentials obtained abroad are often not recognized by state licensing authorities (White House 2015). For example, a dental hygienist trained abroad is barred in some states from licensed practice even if they hold equivalent experience and passing exam scores (Little Hoover Commission 2016).

With no estimates available of how many unlicensed foreign-born workers have equivalent preparation for licensed work, another way to assess the issue is to estimate licensure-rate differences by foreign-born status and race, across levels of educational attainment.⁵ (See Figure 4.) Overall, foreign-born workers—who make up 18.7 percent of our sample of workers 16 and older—have lower licensure rates, with the largest difference among workers who have more than a bachelor’s degree. The gap in licensure rates between foreign-born and native-born workers with more than a bachelor’s degree is 16.8 percentage points for workers of color and Indigenous workers and 17.6 percentage points for White workers.

Workers with criminal records are often excluded from licensure

Analysis of occupational licensing has emphasized the role of criminal records prohibitions (i.e., limitations on whether applicants with criminal records can obtain licensure). Because roughly one-third of younger American Black and Latino men (in their late 20s and early 30s) without a high school degree have been incarcerated, criminal records prohibitions can substantially limit employment opportunity for workers of color (Doleac 2016). The restrictions can boost the wages of those workers who are able to obtain licenses despite the prohibitions: research has explored how licensing wage premiums vary by race and gender, and how these premiums relate to criminal records prohibitions (Blair and Chung 2020 and 2021). But they can also limit access to opportunity for workers whose criminal records are in the distant past or have little connection to the licensed work they seek to perform.

White workers are generally overrepresented in the largest licensed occupations

Since licensing happens at the occupation level, a look across all occupations can miss important detail that is easier to see in an occupation-specific analysis. Among the ten largest licensed occupations in the country, White workers are overrepresented (relative to their share of the total workforce) in all of the most common licensed occupations, with the one exception of licensed practical and vocational nurses. (See Figure 5.) Lawyers and secondary school teachers have the largest shares of workers who are White, at 83.3 percent and 82.1 percent, respectively. Some occupations feature contrasting patterns: for example, pharmacists and physicians—with median hourly wages of $55 and $48, respectively—have an overrepresentation of both White and Asian workers but underrepresentation of Black and Latino/a workers. In the one large occupation in which Black workers are overrepresented (licensed practical and vocational nurses), the median hourly wage is only $22.

Making licensing more inclusive

Stringent licensure requirements hinder entry into licensed occupations. Even after adjusting for education and earnings, White workers are more likely than workers of color to successfully navigate the burdens of licensing requirements and thereby benefit from the exclusivity these barriers to entry create. What kinds of licensing reforms would help workers of color better access labor market opportunities? Three policy options could make a substantial difference:

• Licensing requirements that are not needed to protect public safety (e.g., some portions of an occupation’s training curriculum) could be removed. This would improve access to jobs for all workers but would have disproportionate benefits for those (including many workers of color) who find it especially difficult to procure the required training.
• To streamline the path to licensure for qualified immigrants, experience and training acquired abroad could be more fully recognized, with clear paths made available to acquire any additional training that might be necessary. For example, through its International Medical Graduate program, Minnesota has employed a variety of strategies, including new licensing alternatives and residency positions, to make it easier for immigrants to join the medical workforce. As part of the latter strategy, Minnesota funded residency positions dedicated to foreign-born physicians who were willing to work in underserved or rural areas (Minnesota Department of Health 2015). Such residencies were previously out of reach for international medical graduates because their clinical experience was gained outside of the United States.
• For people with criminal records, blanket prohibitions on licensure could be replaced with tailored prohibitions that address public safety needs. While relevant convictions would still be considered (e.g., a conviction for driving under the influence of alcohol or drugs on the record of someone seeking a commercial driver’s license), irrelevant convictions would not disqualify workers from securing licenses.

Collectively, these reforms would improve access to jobs for all workers and have disproportionate benefits for those workers—including many workers of color—who find it difficult or impossible to overcome the barriers to licensed employment.

The authors are grateful to Nick Carollo, Morris Kleiner, Gabriel Scheffler, and Daniel Takash for their insightful feedback on a previous draft.

Endnotes

1 A well-known issue with the survey questions we rely upon is that they might not always reliably distinguish occupational licenses (which are legally required to practice) from professional certifications (which can signal quality but are not legally required). However, it is relatively unusual for a government-issued credential to function as a certificate rather than a license; our focus on government-issued credentials therefore gets us quite close to the formal definition of a license.

2 The difference in the disparity for Montana relative to the national disparity is not statistically significant.

3 Adjusting for racial differences in educational attainment properly adjusts for the fact that some licensed work requires education even in the absence of licensure—but it over-adjusts for the additional education prompted by the licensing requirement. That is, overly stringent educational and training requirements may cause racial disparities in licensure if White workers are in a better position to surmount them. This analysis therefore presents adjustments for education but may underestimate racial gaps attributable to licensure. To a first approximation, comparing licensure among workers with the same education levels allows us to compare more similar workers. But an issue arises when the license itself requires education above and beyond what would otherwise be necessary if the occupation were not licensed. In that instance, the education itself becomes a part of the barrier to licensure. When we compare licensure of workers of color to White workers at similar education levels, we have already accounted for the educational-requirement portion of licensure costs and may underestimate racial gaps in licensure to the extent that they arise from educational requirements.

4 When we extend these same regressions with hourly wage controls, we find that racial gaps in licensure rates are only slightly attenuated relative to the estimates above.

5 See Cassidy and Dacass (2021) for a detailed investigation of licensure and foreign-born workers.