The Tyranny of the Noncompete Clause

California's enforcement ban has worked out pretty well for Silicon Valley.

Feel free to compete.

Photographer: Justin Sullivan/Getty Images

In “Regional Advantage,” her classic 1994 explanation of why Silicon Valley became the center of the tech universe and Route 128 outside Boston did not, AnnaLee Saxenian of the University of California at Berkeley argued that a key difference was that in Silicon Valley people jumped from company to company, while along Route 128 they stayed put.

In Silicon Valley, Saxenian told me in a 2014 interview:

People start companies, they fail, they succeed, they move on. And that seeds new companies, and those people carry on the knowledge and the know-how but it gets recombined with other skills and technology. Whereas you can think about the 128 company as being autarkic. The company was the family was the unit, and everything stayed within the company.

Saxenian wrote that in Silicon Valley, “Early efforts to take legal action against departed employees proved inconclusive or protracted, and most firms came to accept high turnover as a cost of business in the region.” Along Route 128, which rivaled Silicon Valley in the 1960s and 1970s before falling behind in the 1980s, it was a different story. One leading firm, Data General, “repeatedly sued competitors and former employees to prevent the loss of proprietary corporate information.”

Saxenian, a political scientist who’s now dean of Berkeley’s School of Information, depicted this difference in litigiousness as the product of broader cultural differences between Massachusetts and California. Writing a few years later, Stanford Law School’s Ronald J. Gilson suggested that maybe the legal differences came first. In Massachusetts, as in 46 other states, it’s possible for employers to enforce post-employment covenants not to compete -- aka noncompete clauses. In California it generally isn’t.

Noncompete clauses usually ban employees from going to work for a competitor or starting a competing firm for some pre-determined period of time. Such agreements have been around since at least the 1400s, with proponents defending them as a way to encourage employers to develop new technologies and invest in worker training (because they have less reason to fear losing their secrets and their valuable employees to a competitor) and critics depicting them as an unfair restraint of trade that hurts workers.

California’s ban on the enforcement of noncompetes doesn’t seem to have been the reasoned product of this debate. Wrote Gilson:

The California prohibition dates to the 1870s, a serendipitous result of the historical coincidence between the codification movement in the United States and the problems confronting a new state in developing a coherent legal system out of its conflicting inheritance of Spanish, Mexican, and English law. The existence of this anachronistic legal rule at the time that Silicon Valley developed solved the collective action problem associated with encouraging employee mobility within the district.

Thanks to dumb luck, then, California ended up with a legal framework that encouraged the creation of Silicon Valley. And thanks to Saxenian and Gilson, there was a big new argument against noncompetes: They make it harder to create another Silicon Valley.

Following on Saxenian’s and Gilson’s work, other scholars have in recent years found that noncompetes reduce job mobility and entrepreneurship and cause brain drain from regions that enforce them.  Some even argue that noncompetes are self-defeating for individual companies. This is Orly Lobel of the University of San Diego School of Law and On Amir of the University of California at San Diego’s Rady School of Management:

Subjects in simulated noncompete conditions showed significantly less motivation and got worse results on effort-based tasks. Why? We believe that limits on future employment not only dim workers’ external prospects but also decrease their perceived ownership of their jobs, sapping their desire to exert themselves and develop their skills. The resulting drop in performance may be more damaging to companies than the actual loss of the employees would be.

Not all the research findings have been bad news for noncompetes: One recent study found that they do appear to stimulate investment in worker training. Overall, though, academic research since the 1990s has shone an increasingly unflattering light on noncompete clauses. Yet actual use of noncompete agreements has grown since 2000, at least according to studies of chief executive officer contracts and legal rulings.

Noncompetes are most common in engineering and computer and mathematical occupations, but they also seem to have made their way into less-technical, less-remunerative jobs. In 2014, the New York Times wrote about a summer-camp counselor and a hair stylist who had been prevented from working for competitors, while the Huffington Post discovered that sandwich chain Jimmy John’s required some prospective sandwich-makers and delivery drivers to sign noncompetes that prevented them from working for two years for any business that “derives more than ten percent (10%) of its revenue from selling submarine, hero-type, deli-style, pita and/or wrapped or rolled sandwiches and which is located within three (3) miles" of a Jimmy John’s (which covers a lot of territory).

I’m guessing that your average Jimmy John’s worker can’t afford to take two years off while preparing to seize a better opportunity at Subway. The clause seems intended less to protect training investments or competitive secrets than to reduce the negotiating power of low-wage employees. And sure enough, the Jimmy John’s revelation has led to a bill being proposed in the U.S. Senate to ban noncompetes for low-wage workers. It hasn’t gone anywhere, but there has been some action on noncompetes at the state level. The Idaho and Utah legislatures last month approved laws that restrict the duration of noncompete clauses (to 18 months in Idaho and one year in Utah). In Massachusetts, where lawmakers have repeatedly proposed bans on noncompetes in recent years but haven’t been able to muster enough votes, a new bill that would stop short of banning but put some serious crimps in the practice appears to stand a better chance.

The U.S. Treasury Department, meanwhile, came out with a report last month  that offers new research findings on noncompete clauses’ effect on wages (they depress them), and proposes that firms be required to provide “consideration” to workers when a noncompete agreement keeps them from working:

For instance, a worker who quits may receive 50 percent of her previous salary in exchange for abiding by the terms of the non-compete.

That sounds a bit like the “gardening leaves” common in the upper reaches of the financial sector. It’s certainly more humane and employee-friendly than current practice, and it would probably result in the end of noncompetes for low-wage workers. Still, I think more states should consider outright bans. It sure seems to have worked out OK for California.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.