On July 11, the Harvard Business School's Working Knowledge blog published an article by Senior Editor Carmen Nobel entitled "Non-Competes Push Talent Away." The research described in the article is being used to influence public policy on non-competes, yet the accuracy of the research and the soundness of its premises raise significant questions. To clarify these issues, we turned to Robert Schnell for a commentary
Q: What is the thesis of the article?
Robert Schnell: The article purports to establish through the use of statistics a "brain drain" of inventors from states where non-compete agreements are enforceable to states where such agreements cannot be enforced. The article does not clearly explain what it means by "non-compete" agreements, but the context indicates that the focus is on post-employment restrictions on working in a similar, competing business, rather than addressing any impact of non-solicitation of customer or employee covenants or employee assignment of invention agreements.
Q: How does the article go about establishing its case for the inventor "brain drain"?
Schnell: The article focuses on Michigan, where there was a change in the law in 1985 to make non-competes enforceable, and compares the emigration rate of inventors from Michigan before and after the law changed to the emigration rate of inventors from other states which supposedly did not enforce non-competes. The article states that emigration of inventors from Michigan increased in the period after the law changed, while it dropped in states the article says do not enforce non-competes. The article uses California, Minnesota, Connecticut, and seven others as states "where non-compete agreements are substantially restricted by law."
Q: So what's wrong with the analysis?
Schnell: The fundamental error is that the article specifically identifies seven states as ones that substantially restrict non-competes when, if fact, non-competes are robustly enforced in several of those states. Take Minnesota, for example, with which I am very familiar. Big employers with non-competes try to get Minnesota law to govern those agreements, because Minnesota law is favorable to enforcement. There was a highly-publicized series of decisions some years ago in courts in Minnesota and California involving a company called Advanced Bionics over where a non-compete would be litigated. The parties must have spent huge sums of money on litigation over that one issue, because it was outcome-determinative. The notion that Minnesota and California are alike in terms of non-compete enforcement is just wrong.
In fact, if you read the chart attached to the article, there are only two other states that are like California in their categorical rejection of post-employment non-competes—North Dakota and Oklahoma. And the article fails to address non-solicitation of employee or assignment of invention restrictions, both of which also are relevant to the mobility of inventors and the portability of their inventions, and are generally enforceable across the United States, including California.
Q: Well, why not just compare California, North Dakota and Oklahoma inventor emigration to Michigan inventor emigration, since those three states are the only similar states?
Schnell: The problem is obvious. California's economy is so huge that adding tiny economies like Oklahoma and North Dakota would just be a rounding error. And we can all imagine reasons why inventors left Michigan, but not California, in the late 80's and 90's that have nothing to do with non-competes.
Q: You're right. The problem is obvious. How does the article deal with it?
Schnell: The article addresses the unique nature of California by taking it out and comparing inventor emigration from the other "non-enforcement" states. But since those states are much more like Michigan in terms of enforcing non-competes, the analysis proves nothing. If anything, the analysis might prove that inventors left Michigan for reasons OTHER THAN a non-compete rule, since they left Michigan in higher numbers than they left other states with similar non-compete rules.
Q: OK, so the statistics are flawed. Are there other problems with the approach in the article?
Schnell: Yes. The article deals exclusively with movement of inventors. That is a unique subset of those who might be bound by a non-compete. More often the cases address managers, senior-level executives, sales people, and the like. For the non-inventor employees, California really is a terribly difficult place to enforce non-competes, as compared with all the rest of the country (other than North Dakota and Oklahoma). But with inventors you have the whole issue of intellectual property. California, like the rest of the states, has adopted the Uniform Trade Secrets Act. Yes, inventors can leave and go where they want, but they can't take the intellectual property their employer paid them to develop with them when they go. In fact, the chart the authors attach indicates that non-competes themselves may be enforceable in California to protect trade secrets. Moreover, appropriately drafted assignment-of-inventions agreements, which inventors are routinely required to sign as a condition of employment, are enforceable in California (and other states throughout the country).
Q: So is there any merit to the suggestion the article makes that California is an "open labor market" where you "never stop hiring"?
Schnell: In some contexts, that's right—especially if you're a salesperson or a manager—although the idea that you can sign a non-compete in Massachusetts that picks Massachusetts law and requires litigation in Massachusetts and then just scurry to the West Coast and be home free is terribly naïve. If a case is started first against the employee in Massachusetts to enforce the non-compete, the employee may well lose, even if the employee wouldn't lose in California. Experienced non-compete lawyers are always worried about the race to the courthouse.
But there is another problem for inventors. Employers often will not hire a high-level technical person to work on developing something new—inventing something—if the work would involve anything close to what the new employee did for the old employer. Setting aside the contractual and statutory prohibition on disclosing trade secrets which the new job might involve, the risk is too great that the inventor invents something very valuable and the old employer can claim rights in the invention because it derived from work the inventor did before he or she left. So while inventors can often move employers, the downside is that they may not be able to advance their research at the new shop. That can be very unattractive to an inventor.
Q: The article says that research shows that non-competes are justified as a way to induce investment: "Firms will not invest in R & D unless they can keep their people,"—but that such an analysis is wrong. Do you agree?
Schnell: This won't come as a surprise, but I don't. If you read the research used to justify the conclusion, it relates to managers, not inventors. Maybe it is correct in that context, but I have my doubts. But with respect to inventors, can you imagine a Silicon Valley "angel investor" putting money into a start-up where the inventors would be free to walk to a competitor with the invention once it is successful? Of course not.
Q: So from a policy perspective, what's the take-away here?
Schnell: The article says non-competes lead to a brain drain, but the statistics don't prove that, and may prove the opposite. In two decades of litigating about non-competes I have handled lots of cases, and I can't remember a case where someone declined to take a job because doing so required that he or she sign a non-compete. Like the example of Matt Marx at the beginning of the piece, most employees just sign a non-compete, assuming it is commonplace. There are much more important factors in the employee's decision-making, such as compensation, benefits, and the opportunity for upward mobility. From the employer's perspective, often non-competes are needed because competitors have them and you want a level playing field. In addition, the idea that an employee should not be able to take her contacts, expertise, knowledge of a market, plans for new products, and the like and basically sell them to a competitor through the hiring process,—assuming the restriction is reasonable in duration and extent—has been accepted in all major commercial states except California. Not the only place where California sees things differently.