Locked In: How Noncompetes Affect Wages and Mobility


Research by Professor Jagadeesh Sivadasan shows tech workers have lower wages and less job mobility in states that strongly enforce noncompete clauses.

When it comes to noncompete agreements, legal opinions vary as much as the patchwork of state laws governing them.

Some argue noncompete agreements — which limit when and how a departing employee can work for a competitor — unfairly inhibit worker mobility and innovation. Others say companies are more willing to invest in people when assured they won’t skip right over to the competition. This disagreement over the impact of noncompetes has led to intense and ongoing debates about reforms to noncompete policy in several states.

The real effects are hard to measure, but new research co-authored by Michigan Ross Professor Jagadeesh Sivadasan suggests noncompete agreements, when enforced more strictly, decrease worker mobility and lower wages for tech employees.


“Companies use noncompete agreements to protect their intellectual capital, but we find statistically significant effects for lower wages, and lower mobility, meaning they stay in a given job longer,” says Sivadasan, professor of business economics and public policy.

Sivadasan and his co-authors — Ross PhD student Jin Woo Chang, Natarajan Balasubramanian of Syracuse University, Mariko Sakakibara of UCLA, and Evan Starr of the University of Maryland — focused on tech workers because of the prevalence of noncompetes in that industry.

They analyzed individual worker-level data from the U.S. Census Bureau, which collects it from states. They then ranked each state according to how strongly it enforces noncompete agreements, drawing on work by Michigan Ross Professor Norman Bishara. California, for example, doesn’t enforce them at all, while Florida’s state law is more generous for employers.

Wages were persistently lower throughout the job tenure for tech workers in states with higher enforceability. For example, if a non-enforcing state such as California were to enforce noncompetes to the maximum level of Florida, tech job earnings would be 2.5 percent to 3.5 percent lower.

Sivadasan and his co-authors found a similar effect for mobility. If California started enforcing noncompetes like Florida, the average length of a tech job stay would increase by 7.5 percent, compared to the average length of a non-tech job.

“This provides strong evidence that noncompete enforcement locks employees into their jobs, with little evidence that any productivity benefits are shared with the employees,” he says.

Noncompetes also play into the broader discussion of dynamism in the labor market — that is, how much workers move around and how many jobs are created and destroyed. Data suggests dynamism has slowed down and that noncompete agreements are a policy governments can influence.

“We can’t make the correlation that dynamism is slower because companies are using noncompetes more aggressively, but we can say they slow mobility for the highest-value knowledge workers,” Sivadasan says. “If the government wants to promote mobility, then our research suggests that one policy lever it can use is to limit enforcement of noncompetes.”

The paper is titled “Locked In? Noncompete Enforceability and the Mobility and Earnings of High-Tech Workers.”



Media Contact: michiganrosspr@umich.edu