Removing noncompete clauses from short-term recruiting contracts increased moves between competing employers and raised workers’ earnings from the two participating firms, according to a field experiment. The results put a measurable cost on restrictions that companies often use to protect business information.
The study in The Quarterly Journal of Economics involved about 14,000 offers to freelance recruiters working with two finance firms. All contract versions included a nondisclosure agreement, which restricts sharing confidential information.
Recruiters offered contracts without a noncompete were 36% to 52% more likely to move between the competing employers. Their total earnings from the two firms increased by 12% to 17%. Those figures concern earnings within the experiment, not a rise in annual salaries across the labor market.
Separating information protection from job restrictions
A noncompete limits work for a competitor after an engagement ends. A nondisclosure agreement, often called an NDA, limits the use or disclosure of protected information. An employer can require confidentiality without imposing the same restriction on where someone works next.
Researchers Bo Cowgill, Brandon Freiberg and Evan Starr randomized contract terms and wages. Initial work involved screening résumés, with pay capped at one billable hour. The work exposed recruiters to confidential recruiting information; a competing firm later offered another engagement.
The researchers found no evidence that removing noncompetes increased leakage beyond the protection supplied by the NDA. Participants faced no actual legal risk under the experiment, although uncertainty about enforcement could still affect their decisions.
The setting limits how far the result can be taken. A brief freelance engagement is different from a senior engineer spending years developing a product. The evidence concerns these recruiters and the information they encountered, rather than every trade secret or industry.
A clause can affect hiring before anyone leaves
The University of Maryland’s account of the research, published on September 30, draws attention to a second business cost: attracting people in the first place.
Workers did not receive higher pay for agreeing to the restrictions in the experiment. Making a noncompete prominent also made the job less attractive. Many participants appeared unaware of a restriction when it was placed further inside the contract.
For an employer, hiring costs and retention need separate accounting. A restriction might discourage an existing contractor from accepting a rival’s offer while discouraging another candidate from joining at all. Counting only departures misses the second group.
Our earlier reporting on employee retention and workplace conditions examines the reasons people choose to stay. Contractual limits on leaving address a different problem from creating a job that employees want to keep.
The wider question is how to protect an investment
The OECD Employment Outlook 2026 examines the competing arguments. Employers may want to protect proprietary knowledge and spending on training. Restrictions on mobility can also weaken workers’ bargaining position and prevent skills from reaching another business.
The OECD finds that countries differ substantially in how they regulate such clauses. It also stresses that workers’ perceptions of restrictions can matter alongside formal legal rules. A clause can influence a decision even when the worker is uncertain whether it could be enforced.
Training is another operating decision, as we reported in our coverage of how companies organize employee development. Assessing the investment requires knowing which skills are being developed and how the business expects to recover its cost.
The recruiter study does not decide whether a particular contract is legally valid. It gives employers evidence to examine a narrower choice: whether restricting future work adds protection beyond confidentiality terms, and what the restriction costs before and after recruitment.
