Occupational Licensing and the Earnings Premium in the United States: Updated Evidence from the Current Population Survey
| Date | 01 December 2019 |
| DOI | http://doi.org/10.1111/bjir.12469 |
| Published date | 01 December 2019 |
| Author | Samuel J. Ingram |
British Journal of Industrial Relations doi: 10.1111/bjir.12469
57:4 December 2019 0007–1080 pp. 732–763
Occupational Licensing and the
Earnings Premium in the United States:
Updated Evidence from the Current
Population Survey
Samuel J. Ingram
Abstract
Using survey data from 2015 to 2018, this article analyses the occupational
licensing wage premium in the United States. The estimates show a robust 4–6
per cent wage dierential forlicensed workers. This premium is robust to careful
control for location/local labour market eects and occupation eects. The
premium is also positive for the majority of individual occupations and groups
of occupations estimated. Similar results are found using additional techniques,
including a matching estimator and an analysisof border metropolitan statistical
areas.
1. Introduction
Occupational licensing has a broad impact in the United States, covering one
in four workers. Research on this topic has become increasingly important,
not only because of the extent of licensing, but also because the fraction of the
workforce licensed has been significantly increasing over time. Kleiner (2006)
documents that in the 1950s approximately 5 per cent of the U.S. workforce
was licensed. This increase has drawnthe attention of policymakers and policy
researchers. The White House (2015) has issued a policy brief on licensing
and several policy research groups have published recent licensing analysis,
including the Institute for Justice (Carpenter et al. 2017), the Kauman
Foundation (Wiens and Jackson 2015) and the Center for the Study of
Occupational Regulation (Timmons et al. 2018).
Occupational licensing, the legal requirement that individuals get
government permission to work, has many potential consequences on the
Samuel J.Ingram is at the University of Kentucky.
C
2019 John Wiley & Sons Ltd.
Occupational Licensing and Earnings 733
labour market. This article explores the relationship between occupational
licensing and higher earnings at the national level in the United States.
Profitable work on the licensing premia for specific occupations has been
gradually extended since the seminal work by Friedman and Kuznets (1945).
As the breadth of licensing has increased, however, interest has grown in the
potential impacts of licensing regulation on the entire labour market. Due
to this interest, individual response data on licensing has become available.
Kleiner and Krueger (2010) present the first national estimates using licensing
responses, and two government-conducted surveys have since addedlicensing
questions.
As new data become available, updates and new investigations into the
averagenational licensing premium in the U.S. can be completed. This article
extends the literature by utilizing recent survey data from the 2015–2018
Current Population Survey (CPS). This is the largest survey on licensing
to date, providing opportunities for updated estimates, better controls,
investigations into licensing response and improved estimation approaches.
This article begins by estimating the national licensing wage premium with
the updated data, and then expands the analysis along these dimensions.
The estimates here show an average licensing premium of 4–6 per cent.
Estimates control for location and occupation in more detail than previous
studies, which is important in this literaturegiven that regulations for specific
occupations and jobs are frequently instituted atthe state level. I also estimate
the premium for each occupation and each metropolitan statistical area
(MSA) to explore potential patterns. Binning the occupations, I also find
most occupation groups, such as the education group, have a positive and
significant licensing premium. The magnitudes of these binned estimates are
positive, significant and comparable to the national estimates, ranging from 2
to 14 per cent. The computer and mathematical premium is an exceptionwith
5 per cent lower wages associated with licensed workers, as discussed in the
results.
The new data also allow for an investigation into licensing response by
occupation and location. Since occupations can be licensed at the national,
state or local level, data with small national samples cannot verify if
there is variation in licensing response within an occupation and within
fine geographic areas. When analysing the degree of licensing within an
occupation, workersmay only be licensed in certain states or cities.It may also
be the case that only a portion of the workers within a licensing jurisdiction
are required to be licensed. This results in ‘partial licensing’ of an occupation
code within a licensing jurisdiction as discussed in Gittleman and Kleiner
(2013). The extent of partial licensing documented below is important for
both analysis using licensing response data and analysis using regulations for
licensing status, as discussed in the next section and Section 4.2.
Lastly, this article explores potential advancements in the estimation of
the national premium. As licensing data improves, estimation techniques can
bridge the gap between the national cross-section estimates and the cleaner
identified occupation specific premia found in the case study literature. Two
C
2019 John Wiley& Sons Ltd.
734 British Journal of Industrial Relations
estimators are used to achieve this goal. A matching estimatoris used to take
advantage of the finer detail regarding the worker’s location and occupation.
Then, a simple border state framework is used to take advantage of the
potential variation of licensing laws across state lines. While not perfect
given the discussions regarding the variation in licensing response, these two
additional techniques improve our understanding of the national premium.
The national earnings premium found in previous studies, as well as in this
study using finer controls,is robust to these techniques, which utilize the finer
details in the data to construct more parallel comparison groups.
In the next section, I discuss the literature and provide some background on
these topics. The data follows in Section 3 and the results are presented and
discussed in Section 4.
2. Background
Licensing regulation can aect the national economybeyond altering product
characteristics for specific markets of production. The long-term growth of
licensing across markets may lead to less dynamic labour markets, alter the
allocation of workers and impact earnings. While earnings premia do not
capture the welfare eects of regulation, they are an indicative measure of
distortions in the labour market and are more comparable across markets
than changes in product characteristics or consumer welfare. These earnings
outcomes can be compared across occupations and when combined, give
an informative statistic about the potential impact of licensing on the entire
economy.
Potential Mechanisms for Higher Earnings
Licensing policies have the potential to alter earnings through several
channels. The first possible channel, which is the stated intent of these
regulations, is to improve the quality and safety of consumer products. Thus
far, empirical evidence has been inconclusive as to whether occupational
licensing improves quality. The product quality mechanism will drive higher
earnings if higher quality products are correlated with higher entry costs or
increased consumer demand. This may alter the workers’ earnings profile and
may be one explanation for higher earnings in some professions. See Kleiner
(2006) for a discussion of occupational licensing and quality.
Another channel for higher earnings of licensed workers is through labour
market restrictions and entry costs.Licensing may directly limit labour supply
by restricting the number of licenses awarded or indirectly decrease it through
fees, tests and additional education requirements. In the United States, this
mechanism is more direct than the quality mechanism because licensing
boards typically implement regulation through entry restrictions and higher
standards on new applicants, instead of on final products through quality
inspections.
C
2019 John Wiley& Sons Ltd.
To continue reading
Request your trial