Methods

Sampling Lower-Income Occupations

This third edition of License to Work, like the first and second, examines the burdens and scope of licensing. We continue to observe requirements across all 50 states and the District of Columbia for 102 lower-income occupations. The State and Occupation Profiles summarize the results for each state and occupation. In addition to providing summary results, the Occupation Profiles also provide definitions for each occupation drawn from the U.S. Department of Labor-sponsored O*NET OnLine (www.onetonline.org), except where noted. ¹

The 102 occupations are the same as in the second edition of License to Work, published in 2017. Originally drawn for the first edition, published in 2012, the sample comprises occupations that in 2012 were licensed by at least one state and recognized by the U.S. Bureau of Labor Statistics as ones in which practitioners earned incomes below the national average. One of the 102 occupations, psychiatric aide, is no longer licensed by any state.

The occupations come from diverse industries, ranging from construction (e.g., carpenter and crane operator) to education (e.g., public preschool teacher and head coach for public high school sports). Many of the occupations are suited to people entering or reentering the economy, with some offering opportunities for entrepreneurship. Those offering entrepreneurial avenues include many of the construction trades, auctioneer, taxi driver and interior designer.

While they come from a range of industries and attract a variety of different workers, many of the occupations are ones members of the public come into contact with regularly. One example is transportation occupations, such as city and school bus drivers, taxi driver, and truck driver. Others are barber, cosmetologist and other beauty occupations. These occupations also tend to be licensed commonly and often universally—though requirements vary considerably.

Other occupations are more obscure, such as dairy equipment still machine setter, log scaler, milk sampler, packer and weigher. These less-familiar occupations also tend to be licensed less frequently.

Table 1 provides a complete list of the 102 lower-income occupations included in this report, along with the number of states that license each.

Data Collection Timeframe

We collected the data for this report occupation by occupation over a period of two years—from February 2020 to March 2022. To allow time to prepare this report for publication in November 2022, we ended data collection on March 18, 2022. 

We made every effort to ensure our data are current as of that date. However, as this report goes to press, we cannot guarantee our data are perfectly up to date. Over time, licensing requirements can change, meaning some data we collected earlier in the data collection period could have become outdated by the end. Some data could also have become outdated after we completed data collection.

To hedge against our data becoming outdated, we attempted to capture changes that occurred while data collection was still ongoing. Specifically, when we learned of pending legislation that would affect relevant licensing requirements, we made note of it and checked the status at the end of our data collection period. Though it is possible that some changes escaped our attention, our data are likely to be highly accurate overall given the slow-moving nature of the legislative and regulatory processes. Nevertheless, our data are best thought of as a snapshot of licensing requirements at one point in time. 

Identifying Licenses

Every occupation in this report is licensed by at least one state. The one exception, as noted above, is psychiatric aide. 

We consider an occupation to be licensed when people must get a permission slip from their state to legally practice it, regardless of what that permission slip is called. This means we count as licenses some permission slips that states refer to as “registrations” or “certifications” so long as people must hold them to legally work in an occupation. We do not count as licenses restrictions on the use of an occupational title, such as “certified interior designer,” as such restrictions still allow for open occupational entry. Nor do we count any of the voluntary certifications offered by many state governments.

Two occupations, direct entry midwife and optician, are treated somewhat differently in our data. In some states, these occupations can only be practiced with a higher-level license than we observe in this report (i.e., a nurse midwife license for midwives and an optometrist license for opticians). In these states, we treat the occupation as prohibited, and do not record license requirements in order to avoid conflating different license types. By contrast, other states do not require any license to practice these occupations; in these states, the occupations are treated as unlicensed. 

Measuring Burdens

In addition to measuring how widespread licensing is, this report measures the burdens states impose on aspiring workers through licensing.

We measure licensing burdens by looking at five common types of licensing requirements: fees, education and experience, exams, minimum grade completed in school, and minimum age. States also impose many other types of requirements, such as bonding, insurance, character references and minimum net worth. Such requirements are not common enough to serve as consistent measures of burden across the 102 occupations, so we have not sought to capture them.

Education and experience requirements take many different forms—hours, years, college credits and so forth—making it difficult to compare across occupations and states. To make comparisons possible, we converted all education and experience requirements into a common measure of “estimated calendar days lost.” For information on how we performed these conversions, see Appendix A.

As many of the licenses we observe have multiple kinds of associated fees, education and experience requirements, and exams, we combined these requirements by type. For fees and exams, we simply summed the fee amounts and the number of exams. ²  For education and experience, we first converted each requirement into days lost and then added the days lost from each requirement together to get the total days lost to education and experience for licensure.

States sometimes require multiple licenses to fully practice an occupation. ³  In such cases, we recorded each license’s requirements separately and then summed them across licenses by requirement type.

States also sometimes offer multiple pathways to licensure. When this was the case, we favored the pathway with the least burdensome requirements. ⁴  This conservative approach helped us avoid overstating burdens. That said, because we are interested in the effects of licensing on entrepreneurship, we observed, wherever possible, the requirements to independently operate within an occupation, even though licenses for sole proprietors are typically stricter than those for employees. ⁵

Because we consistently collected a common set of licensing requirements, our data allow us to compare how difficult the 50 states and the District of Columbia make it for workers to enter the 102 occupations.

Data Improvements

This third edition of License to Work observes the same occupations and uses the same methodology as the second edition. However, to ensure our data would be comparable over time, we corrected errors in the data from the two prior editions.

For example, in the public preschool teacher occupation, we now observe SAT or ACT exam fees in states that allow aspirants to take those exams instead of more expensive ETS Praxis exams. This option was available when we collected data for prior editions. Not accounting for it was an oversight.

Importantly, because of these corrections, 2017 data presented in this edition may not match data presented in the second edition. Of note, corrected 2017 average burdens reported here are higher than those previously reported. Specifically, while the second edition reported averages of 360 days lost to education and experience and $267 in fees across all 102 occupations, the 2017 figures from our corrected dataset are higher: 384 days lost and $280 in fees. We urge readers to rely on the revised data in this edition.

The revised data for the three editions of License to Work can also be viewed side by side in our comparison dataset, available at https://ij.org/report/license-to-work-3/ltw3-data/. The data for the second and third editions are comparable across all 102 occupations. However, as indicated in the dataset, only 45 occupations are comparable across all three editions due to systemic differences in data collection methodology between the first and second editions. ⁶  

Taken together, the three editions allow us to track licensing requirements for a broad array of low- and middle-income occupations over time—specifically, from 2012 to 2017 to 2022—making License to Work a one-of-a-kind resource.

Undercounting Licensure and its Burdens

While this third edition of License to Work provides a much-needed window into occupational licensing requirements, our approach undercounts licensing’s scope and burdens. First, the 102 occupations we study are only a sample. Nationwide, there are many other occupations and job responsibilities that require a license.

Second, we consider only state licenses, not municipal, county or federal ones. In some states, certain occupations may be licensed by cities and counties, not state governments. ⁷  For example, Denver licenses contractors, but the state of Colorado does not. ⁸  We do not observe Denver’s licenses. Other occupations are subject to federal regulations, such as mobile home installers. In states without a license for mobile home installation, installers must get a federal license. Because we count only state licenses, we treat those states as unlicensed. ⁹  There are also some occupations that are licensed only or primarily at the federal level, though none of the 102 occupations we study fit that description. ¹⁰

Third, as noted above, there are many types of licensing requirements we do not record, such as bonding and insurance. In addition, we do not count the costs associated with required schooling or other forms of training, such as tuition or forgone income. These costs undoubtedly present hardships for lower-income workers hoping to find jobs in licensed occupations. However, these costs are highly variable and often indirect—and thus impractical to record.

For further details on our data collection and methods, please see Appendix A and Appendix B.