In recent research, Johnathan S. Hartley and Morris K. Kleiner find that occupational licensing is globally pervasive among both developed and developing nations. However, higher national licensing rates are associated with lower GDP per capita, larger informal sectors, and weaker governance scores.
In the 1950s, only about 5% of American workers needed a government license to do their job. By the early 2000s, that figure had climbed to roughly 25%, and it has stayed there ever since. Almost every occupational licensing study of the past 20 years highlights an important point: licensing, which was once a narrow tool reserved for doctors, lawyers, and a handful of skilled trades, now extends to cover a wide range of positions, from genetic counselors to Uber drivers.
This is all while union membership has declined from roughly a third of the workforce in the 1950s to about 10% today, almost the mirror image of the rise in licensing over the same stretch. When unions were pervasive, they facilitated workers’ bargaining power and enabled unionized employees to obtain higher wages. Deindustrialization and political opposition has undermined labor unions power. The relationship between decreasing unionization and rising licensing requirements is not necessarily causal, but licensing can restrict competition and extract higher pay, potentially from consumers. Unlike a union, associations representing an occupation that oversee occupational licensing requirements do not have to bargain with an employer or win a certification election. Instead, it has to persuade the state legislature and governor that the public needs protecting. Licenses are often specific to a state: a lawyer licensed in one state cannot necessarily practice in another, further limiting employment.
The increasing licensing requirements reflect an array of American legal and labor institutions, but are licensing requirements unique to the United States? Europe has completed one survey of occupational licensing showing about 22 percent of the EU has attained an occupational license. However, most of the developing world has never been assessed. We attempt to address that in a new paper, using existing surveys and fielding new nationally representative ones in 44 countries, including several, like India, Argentina, and Nigeria, where no comparable licensing data had ever existed.
We find that 42.5% of Indian workers hold a job that legally requires a government-issued license, the highest rate in our entire sample, ahead of every country in Europe and well ahead of the 28% figure for the U.S. South Africa is close behind at 40.2%. In contrast, licensing-averse Denmark sits at just 14%, whereas wealthier nations such as Germany have almost 33% of their workforce requiring an occupational license.
The original case for occupational licensing was about bridging information asymmetries and minimum quality standards (if you’re seeing a doctor, you want to know that they have the requisite skills and experience, so they don’t put you or your community at risk of, for example, a communicable disease). However, the growth of occupational licensing suggests that it has expanded to occupations such as hair braiding and interior design that do not threaten the public.
Across our 44-country sample, countries with heavier licensing burdens are correlated with lower GDP per capita, larger informal sectors, and weaker scores on major World Bank governance indicators, from rule of law to control of corruption. India’s and South Africa’s licensing regimes did not emerge from careful cost-benefit analysis of consumer protection. They are legacies of dense, discretionary state control over economic life: India’s post-independence license raj and South Africa’s post-apartheid effort to regulate a fractured labor market through formal rules. If licensing were mainly about protecting patients and consumers, we would expect it to track state capacity and good governance. Instead it tracks weak governance and a large informal sector, which is associated with a pattern of rent-seeking theories of regulation.
Germany, Australia, and Israel complicate the story, and are worth taking seriously rather than waving away. They license heavily, and are wealthy, well-governed economies. But their licensing is embedded in coordinated systems of apprenticeships, sectoral bargaining, and vocational credentialing that long predate the more recent, occupation-by-occupation expansion of licensing that we have seen in the U.S.
Our finding on informality sharpens the point, as countries with more extensive licensing also tend to have larger informal economies. India and Peru combine high licensing coverage with informal employment rates above 70%. When the price of entering a formal occupation—the fees, the exam, the required schooling—exceeds what a worker can pay, the rational response is to skip the formal economy altogether. A licensing board that prices out workers is not protecting consumers; it is potentially pushing economic activity into a shadow market where consumers have no protection at all, and where the incumbents who lobbied for the license face that much less competition.
The evidence we find on wages is also consistent with the rent-seeking story. Licensed workers in our pooled sample earn about 6-19% more than comparable unlicensed workers, in line with the roughly 15% premium past work has found in the U.S.
This certainly does not mean we should get rid of licensing altogether. Most people don’t want an unlicensed physician, and some minimum quality standards are worth their cost for critically important tasks. But the burden of proof should sit with those seeking to expand licensing requirements, not the consumer of the service. A statute giving a license should be the exception, granted where there is a demonstrated public-safety rationale, not the default setting for entering one quarter of the labor market. American states have begun to internalize this, moving to recognize out-of-state licenses and, in some rare cases, sunsetting boards that cannot justify their own existence. Developing countries with the highest licensing burdens in the world have had no comparable reckoning, largely because until now they lacked the data to have one (even if it was something that might be obvious with a Licensing Raj). Our paper’s results suggest they should start having meaningful conversations about the potential negative effects of their licensing regimes and develop the data to analyze these institutions.
Author’s Disclosure: The authors report no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
