AdobeStock_122614358.jpeg
VoxEU Column Labour Markets

Occupational licensing has a sizeable impact on job mobility in the US

More than a fifth of American workers are required to hold an occupational licence to do their job, usually with the aim of protecting public health and safety. However, secular declines in job mobility, business dynamics, and productivity growth have raised concerns over the costs of licensing and its potential influence on these trends. Using novel administrative data with nearly complete employment coverage, this column presents suggestive evidence of sizeable effects of licensing on job mobility, especially on job-to-job flows across states. 

For some time, occupational licensing reform has been high on the political agenda in the US (White House 2015, NCSL 2019). However, those within licensed professions are often strongly opposed to reform (Kilmer 2018). Regulators license an occupation by establishing a set of requirements (e.g. education, work experience, exams) needed to carry out certain activities and use a protected title. The aim is usually to protect the health and safety of consumers, and to ensure the quality of services. However, licensing has expanded well beyond ‘traditional’ professions such as doctors, lawyers and teachers, and now also applies to security guards, home inspectors, interior designers, and even florists in some states. As a result, more than 20% of American workers are licensed (Figure 1), resembling the levels seen in both the EU and Japan (Koumenta and Pagliero 2017, Morikawa 2018).

Figure 1 Many occupations require a licence and covers more than 20% of all workers

Percentage of workers with an occupational licence by occupation, 2015 (EU) and 2018 (USA)

Note: The proportion of total employment in each occupation is reported in parentheses for the United States. For comparison, occupational classification codes used in the EU (ISCO-08) have been converted to occupational codes used in the United States (SOC 2010). In cases when the ISCO-08 code links to more than one main SOC group, the group with the highest employment share is used.
Source: Current Population Survey, BLS; Calculations produced by Maria Koumenta (Queen Mary University of London) based on the EU Survey of Regulated Occupations.

Research on the costs and benefits of occupational licensing has struggled to find evidence of better quality of services (Kleiner 2017), except for some of the first professions to become licensed more than a century ago, such as midwives and doctors (Anderson et al. 2016, Law and Kim 2005). By contrast, evidence of higher prices and wages, lower mobility, and weaker firm productivity is growing (Gittleman et al. 2018, Blair and Chung 2019, von Rueden et al. 2020). This poses the question of whether (and by how much) occupational licensing contributes to the secular decline in job mobility, interstate migration, and productivity growth. 

In a recent paper (Hermansen 2019), I attempt to provide macro-level evidence of the link between licensing and job mobility by making use of administrative data from a relatively novel Job-to-Job Flows database from the Census Bureau. Previous studies have usually had to rely on survey data with limited sample sizes or have been forced to study only a few specific occupations. Comprehensive data on licensing regulation over time is, however, still incomplete. As a result, I have to rely on licensing variation across states to make inference. 

Coverage of occupational licensing

The state level governs most of the occupational regulation in the US, with substantial variation in coverage and licensing requirements. Fewer than 50 occupations are licensed in all states, while more than 400 occupations are licensed in at least one state. My estimates suggest that the share of licensed employment varies from around 15% in Hawaii and Mississippi to more than 27% in Illinois and New Jersey. This is similar to the variation seen across the EU (Koumenta and Pagliero 2017). 

Differences in licensing coverage matters for job mobility. A simple scatter shows lower job-hire rates in states with a higher share of licensed employment (Figure 2, Panel A). Job hires can be from non-employment, or directly from another employer (job-to-job hire). It should be noted that the new employer may be located in the same state, or in another state altogether. Breaking up the job-to-job hire flows suggest that occupational licensing depresses interstate hire in particular (Figure 2, Panel B). States with high licensing coverage attract substantially fewer job-to-job hires from other states compared to low-licensing states. This is not surprising since a licence obtained in one state is not automatically recognised in other states, effectively imposing non-tariff trade barriers and mobility constraints between states.   

Figure 2 High occupational licensing coverage depresses job-to-job hire between states

Note: Licensed employment by state is computed by mapping information on licensing regulation to occupational employment statistics and aggregating across states (Hermansen 2019). Panel B shows simple averages across 25 low- and high-licensing states; see Hermansen (2019) for an employment-weighted decomposition.
Source: OECD calculations based on Job-to-Job Flows database, Census Bureau; careeronestop.org and Occupational Employment Statistics, BLS.

I scrutinise these correlations by formal empirical analysis, exploiting the cross-sectional state variation. It is important to bear in mind that the available data do not allow me to infer causal effects. Still, the analysis finds significantly negative associations between licensing coverage and all available measures of job mobility, incluing job hires, job separation, transitions in (and out) of non-employment, as well as job-to-job moves. These associations are economically important, which is illustrated with a simulation exercise in Figure 3, taking estimates at face value. If overall licensing coverage had been gradually reduced in the 2000s (to five percentage points lower in 2018 compared to the observed level; see Panel A), job hire and separation rates could have been substantially higher (Panels B-D). The job-hire rate could have increased by as much as 0.6 percentage points, corresponding to a quarter of the decline observed since 2000 (Panel B).

Figure 3 How would job mobility look like if licensing coverage had been reduced in the 2000s?

Note: Upper and lower bound estimates reflect the estimates from the cross-sectional estimation with control for sex/age or sex/education using the NCSL and COS indicator, respectively.
Source: OECD calculations based on Job-to-Job Flow database from the Census Bureau.

Strictness of occupational licensing regulation

Licensing (or not licensing) an occupation is just one aspect of regulation. The requirements to obtain a licensure for the same occupation in two different states can be quite different. Training requirements for cosmetologists range from 500 hours in South Carolina to 2,100 hours in Iowa, Nebraska, and South Dakota (Figure 5). Virginia requires ten years of experience to acquire a HVAC (heating, ventilation and air conditioning) contractor licensure, whereas six states license this occupation without any training or experience requirements. Fifteen states do not license HVAC contractors at all. 

Figure 4 Training and experience requirements vary substantially across states

Note: All training and experience requirements are converted to hours (1 week = 40 hours, 1 month = 2000/12 hours; 1 year = 2000 hours). For some occupations, a few states with missing information are not recorded
Source: Occupational Licensing Database from the National Conference of State Legislatures.

I construct a new indicator for the strictness of regulation (on  scale of 0-6) to summarise the variation in licensing requirements across states. The indicator has four components: (i) entry restrictions (such as minimum age and out-of-state licensing recognition), (ii) education and training requirements, (iii) renewal requirements, and (iv) restrictions for ex-offenders (such as blanket bans and limitations on background checks). This information is available for 31 occupations only (NCSL 2017), which are all licensed in at least 30 States. 

Using the strictness indicator in similar empirical analysis as above, I again find significant and economically important associations between licensing strictness and job mobility. Back-of-the-envelope calculations suggest that if the most regulated state (Washington) reduced the strictness indicator to the level of the median tates (North Carolina), the job hire rate could increase by 0.4 percentage points (Figure 5, blue bars). This would correspond to one-fifth of the decline since 2000. Interestingly, education and training requirements are found to have a positive association with job-to-job hire and job-separation in a joint estimation (with all four components of the indicator). This may suggest that human capital investments from licensing improve job opportunities, all else being equal. 

Figure 5 How would job mobility change if strictness of regulation was reduced from maximum to median state level?

Simulated effect of the most regulated state in each dimension moving to the median state regulation level

Note: The figure shows five separate policy experiments based on the constructed strictness indicator and its subcomponents (see Hermansen, 2019). Licensing strictness reflects Washington reducing the indicator from 2.9 to the median level of 2.4 in North Carolina. Entry restrictions reflects Nevada reducing the sub-indicator from 2.6 to the median level of 1.7 in Virginia. Education and training reflects Washington reducing the sub-indicator from 2.4 to the median level of 1.8 in Minnesota. Renewal requirements reflects Washington reducing the sub-indicator from 3.4 to the median level of 2.5 in Utah. Restrictions for ex-offenders reflects Alabama reducing the sub-indicator from 4.2 to the median level of 3.6 in Georgia. Insignificant estimates at the 5% level are set to zero. 
Source: Hermansen (2019).

Policy implications

Regulating occupations plays a legitimate role in protecting consumers and ensuring that markets work effectively. But emerging evidence points to excessive and costly occupational licensing, possibly reflecting successful lobbying from professionals seeking to protect their market or gain status. My findings support the conjecture that the licensing of more and more occupations may have made an important contribution to the secular decline in dynamism. 

States could delicense or use less restrictive alternatives such as voluntary certification to revive job mobility. The problem of interstate barriers for job-to-job moves by ‘out-of-state’ licenses not always being recognised is of particular concern. Recently, Arizona was the first (and so far, only) state to automatically grant occupational licences to anyone who moves there with a licence from another state. Interstate compacts (a binding contract between states) is another alternative that has been found to increase the mobility of nurses (Abdul Ghani 2019). Yet, forming compacts for every occupation would be cumbersome and time consuming. The US may want to look to Canada as an example of a system that facilitates mobility of licences through an internal free trade agreement, or to the EU’s efforts to increase cross-country mobility for regulated professions.

Author’s note: The opinions expressed in this column are those of the author and do not necessarily reflect the views of the OECD.

References

Abdul Ghani, A (2019), “The Impact of the Nurse Licensing Compact on Inter-State Job Mobility in the United States”, OECD Economic Survey of the United States: Key Research Findings, Paris: OECD Publishing.

Anderson, D M, R Brown, K K Charles and D I Rees (2016), “The Effect of Occupational Licensing on Consumer Welfare: Early Midwifery Laws and Maternal Mortality”, NBER Working Paper, No. 22456.

Blair, P and B Chung (2019), “How Much of Barrier to Entry Is Occupational Licensing?”, British Journal of Industrial Relations 57(4): 919-943. 

Gittleman, M, M Klee and M Kleiner (2018), “Analyzing the Labor Market Outcomes of Occupational Licensing”, Industrial Relations 57(1): 57-100.

Hermansen, M (2019), “Occupational Licensing and Job Mobility in the United States”, OECD Economics Department Working Papers, No. 1585.

Kilmer, M (2018), “A Look at Occupational Licensing Reform across the United States”, Arkansas Center for Research in Economics.

Kleiner, M (2017), “The influence of occupational licensing and regulation”, IZA World of Labor, No. 2017:392.

Koumenta, M and M Pagliero (2017), “Measuring Prevalence and Labour Market Impacts of Occupational Regulation in the EU”, European Commission.

Law, M and S Kim (2005), “Specialization and Regulation: The Rise of Professionals and the Emergence of Occupational Licensing Regulation”, Journal of Economic History 65(3): 723-756.

Morikawa, M (2018), “Occupational licenses and labor market outcomes in Japan”, Japan & The World Economy 48: 45-56.

National Council of State Legislatures (2019), “The Evolving State of Occupational Licensing: Research, State Policies and Trends”.

National Council of State Legislatures (2017), The National Occupational Licensing Database

von Rueden, C, G Nicoletti, I Bambalaite (2020), “Occupational Entry Regulations and their Effects on Productivity in Services: Measurement and Firm-Level Evidence”, OECD Productivity Working Papers, forthcoming.

White House (2015), Occupational Licensing: A Framework for Policymakers, report prepared by Department of the Treasury Office of Economic Policy, Council of Economic Advisors and Department of Labor.

1,155 Reads