Employee Training and the Value Proposition of Non-Competes

The best part of the current debate over non-compete agreements is not that we're nearing a consensus: it's that we're having the debate.

With the White House's Call to Action, and the research that led to this unprecedented move, commentators have begun to explore why firms have non-competes in the first place. The answer usually is that many view non-competes as protecting trade secrets, customer relationships, or training investments.

In the main, those all sound rational, depending of course on the firm's line of work and the employee's ability to inflict damage after leaving. My personal views on this largely have not changed in the nearly 20 years I have been practicing in this field. I summarize them as follows:

1. Non-competes should not be per se invalid and should be reserved for those at the highest levels of the company.
2. Businesses use them indiscriminately without considering their benefits and drawbacks.
3. A garden-leave approach, which provides tangible benefits to employees during the restricted period, is economically efficient and the preferred template for businesses to use.
4. Courts should not enforce non-competes when there is a termination without cause.
5. The judicial reformation or "blue-pencil" doctrine is inappropriate and creates poor incentives. It should be banned.
6. Less onerous restraints, such as non-disclosure agreements, most often provide sufficient protection.

What is notable about my list, though, is that it doesn't really account for the employers' incentive to train workers. Do non-competes actually facilitate training by encouraging firms to devote scarce resources to developing employees' skills? Would firms continue to train employees if non-competes were not allowed?

There are a couple of different viewpoints on this.

In a Wall Street Journal op-ed, Jason Furman and Alan B. Kruger discussed "monopsony" power and the use of non-competes. They note that "[i]f monopsony power creates barriers to workers switching jobs, it can slow labor turnover, reducing dynamism and innovation." They then cite the now-commonly approved benchmark that nearly 20% of workers have signed a non-compete agreement. (There are about 1,500 non-competes suits per year, according to Evan Starr.) Noting that nowhere near that many workers have access to trade secrets, Furman and Kruger conclude that "[t]here is no reason why employers would require fast-food workers and retail salespeople to sign a noncompete clause - other than to restrict competition and weaken worker bargaining power."

The counter to this monopsony argument focuses on the employer's inability to recapture training costs without a non-compete agreement in place. The analysis does not focus on trade secret access.

Hoover Institute fellow, David Henderson, summarizes his thoughts in a rebuttal to the Furman/Kruger rationale. Simply put, Henderson notes that an employee who receives training may be able to obtain a higher wage elsewhere than that which the employer is willing to offer. While carefully stating that he is not endorsing non-competes, Henderson clarifies the "training" rationale for why a low-wage worker may be subject to a non-compete.

However, I find this theoretical rationale misses the mark. First, American businesses long have assumed the responsibility for educating their employees. It seems illogical that they would fail to do so merely because of the possibility of future competition. Second, formal training is no panacea. Most effective training occurs on the job, through the collective experiences of day-to-day work and a general exposure to the marketplace. Courts long have concluded that general skills and knowledge are not protected interests.

Quantifying training, too, is nearly impossible and cannot be reduced to a mathematical equation that will conclusively demonstrate whether non-competes incent or deter training. If quantification were possible, then training repayment agreements - or quantifying damages based on such a hypothetical contract between market competitors - would seem to be a better solution than broad non-compete enforcement. Could we even envision a new regime where a repayment model supplants an enforcement model? I doubt it, but it makes for an interesting discussion.

The bigger problem, though, may be reciprocal use of non-competes and the development of industry standards, using some sort of an anti-trust "relevant market" analysis. A firm, for instance, may not be inclined to use a non-compete but does so only in response to its direct competitors' pre-existing inclination to do so. In that sense, the reluctant firm is only trying to avoid the free-rider problem: if competitors see that the reluctant firm has no non-compete regime with its workforce, those competitors will be less willing to invest in appropriate employee training, thereby lowering production costs. This explanation for why firms use non-competes in a blanket fashion actually seems to prove what Furman and Kruger are saying: monopsony power stems from artificial means to restrict competition, that being an industry-wide practice of using non-competes in the first instance.

If anything, therefore, the attempt to justify non-competes on the basis of protecting training costs seems to explain why so many workers without the ability to harm their employers are subjected to them.