FTC Evidence That Non-Competes Reduce Earnings Is Inconclusive
Stephen Bronars of Edgeworth Economics analyzes the Federal Trade Commission’s stance that non-compete clauses reduce worker earnings across industries, saying that research doesn’t support this conclusion.
On Jan. 19 the Federal Trade Commission proposed a rule that would prohibit firms and employees from agreeing to non-compete clauses. These contract provisions prevent employees from working for competing businesses in certain geographic areas within a set time after leaving a firm.
The rule claims recent research shows NCCs have “reduced wages for workers across the labor force—including workers not bound by non-compete clauses,” even though studies indicate only 18% of employees work under an NCC.
The rule claims NCCs lower pay by limiting mobility and reducing competition. However, NCCs have an ambiguous impact on earnings, because firms requiring NCCs must offer higher compensation to attract workers, all else equal. In addition, by limiting mobility, NCCs encourage employer-provided training that benefits workers.
Most cited studies cannot identify which employees work under an NCC but infer the effect on earnings by comparing pay in states (or time periods) with different amounts of “NCC enforceability.” An exception is a study showing CEOs earn 18.4% more, all else equal, if their contract includes an NCC.
The FTC’s conjecture that the rule will increase earnings is based on inappropriate extrapolations from six cited studies that correlate NCC enforceability and pay. The research examines either CEO compensation, two state-specific NCC bans limited to certain job groups or the impact of small differences or changes in NCC enforceability.
Empirical models in the cited research cannot identify workers with similar skills. The rule acknowledges the limitations of studies of specific states and job groups, including CEOs, so I focus on other limitations of the research.
A problem common to all the cited research is the inability to adequately measure employees’ skills and relevant prior work experience. More experienced workers receive higher pay, so higher earnings in these studies may be greater skills and experience and not increased competition.
For example, cited research finds average earnings among new hires in Hawaii’s tech sector increased by 4% relative to other benchmarks—tech jobs in other states and other jobs in Hawaii—after Hawaii’s tech sector NCC ban in 2015.
The FTC interprets this as evidence that a new employee with the same skills is paid 4% more after the NCC ban than before the ban. However, because the study is based only on average salaries, it cannot compare job qualifications of new hires before and after the NCC ban.
The study cannot distinguish between an increase in competition and firms hiring slightly more experienced and skilled tech workers after the NCC ban.
Main FTC Focus
The study given the most weight by the FTC covers all workers in all states and examines within-state changes in NCC enforceability. It explains only 36% of pay variation using age and educational attainment as worker characteristics.
Workers with the same age and schooling appear identical in the study but can differ by company tenure, time in current position, relevant prior work experience, and college major.
Higher earnings after a drop in NCC enforceability may be because firms hire slightly more experienced workers after NCCs are less enforceable.
The study finds a one standard deviation decline in NCC enforceability is associated with 0.3% higher earnings. This effect size is small because within-state changes in NCC enforceability are small during the study period.
Predicting the rule’s impact from this study is inappropriate because it requires unreasonably large extrapolations. The rule mandates a change of more than 10 standard deviations in NCC enforceability in many states.
A change this large, for representative businesses and employees, has not been studied. For comparison, if a pharmaceutical study found a small benefit from a small dose of a proposed drug, it would be inappropriate to conjecture that 10 times the dose would yield 10 times the benefit.
A relative decline in hiring of less experienced workers, which slightly increases average pay, is expected if NCCs are prohibited because firms will invest less in employer-provided training if workers can leave for competing businesses after being trained.
By limiting worker mobility somewhat, NCCs encourage businesses to hire less-experienced workers, at lower initial pay, and provide training. A cited study finds where NCC enforceability is greater the incidence of training is 14% higher.
A nationwide NCC ban will reduce some firms’ incentives to hire and train less experienced workers, which harms workers who lack the qualifications for higher-wage jobs.
This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law and Bloomberg Tax, or its owners.
Stephen G. Bronars is a partner with Edgeworth Economics. He analyzes employment discrimination, wage and hours, and the use of criminal background checks in hiring through the lens of complex economic data.