Employers say they need noncompete clauses in order to protect their trade secrets and confidential information. But do they put their money where their mouth is? Research we conducted suggests that they don’t.
Noncompete clauses prohibit workers from being able to take a job with a company that competes with their current employer. In January the Federal Trade Commission proposed banning them based on evidence of the harm they do markets and to low- and high-wage workers; a Senate bill would similarly limit the use of noncompetes. Some trade groups and business associations are opposed to the FTC’s proposal, however, defending employers’ need for noncompetes as a means to protect trade secrets and other confidential information. We decided to put these claims to the test.
For a recent paper we wrote with Michael Lipsitz, we tested whether firms give small raises to workers to get them over a minimum earnings thresholds to enforce noncompetes. Our study leveraged a 2020 Washington state law that retroactively invalidated noncompetes for workers earning below a threshold of $100,000 per year (tied to inflation), covering approximately 80 percent of Washington workers. If employers valued the ability to enforce noncompetes for workers earning just below the threshold, then after the law came into effect employers should have given such workers small raises to reach or exceed the threshold. In the aggregate data we should have then seen more workers at or just above the earnings threshold, and fewer workers just below.
Using data covering nearly every worker in Washington from 2001 to 2021, our study found no evidence that employers were giving workers raises to reach the noncompete thresholds established in the law ($100,000 in 2020 and $101,390 in 2021). This was true even in industries where the justifications for noncompetes seem the strongest, such as manufacturing or professional services.
A survey of employment attorneys in Washington reinforced these findings. The median attorney indicated that just 10 percent of their clients would raise the wages of just-below-threshold workers in 2022 to reach or exceed the threshold. When asked why employers wouldn’t give workers such raises, attorneys responded that employers did not need to enforce noncompetes in court for workers near the threshold, and that employers had other tools, such as nonsolicitation and nondisclosure agreements, to protect their interests.
Employers also claim that without noncompetes they will lose value. However, we found no evidence that Washington’s noncompete ban reduced the value of publicly traded companies operating in Washington.
What conclusions can policymakers draw from this study? No matter what employers might say, or the industry they are in, we did not find any evidence that employers value the ability to enforce noncompetes enough to give workers even small raises to get them above an earnings threshold corresponding to approximately the 80th earnings percentile. Perhaps at some higher earnings threshold employers would be willing to pay to enforce a worker’s noncompete. We can’t say where that threshold is, or if it exists. Indeed, Microsoft recently dropped its noncompetes for everyone but highly paid executives, implying that they, too, don’t find noncompetes necessary for most workers.
As for the Federal Trade Commission’s proposed rule (and other state and federal legislation regarding noncompetes), our findings suggest that policymakers probably shouldn’t be too concerned about banning noncompetes for at least 80 percent of workers, because firms aren’t either. Rather, the debate over noncompetes should probably be limited to the top wage earners, perhaps just the top 5 to 10 percent.
Amid this noisy debate over regulating noncompetes, employers have indeed spoken—not with their voices, but with their pocketbooks.