Streetwise Professor

May 1, 2024

One Tool Lina Bashes Away At Non-Competes With Her Market Power Hammer

Filed under: Economics,Regulation — cpirrong @ 12:46 pm

Well, the FTC done gone ahead and done it, implementing a broad prohibition on “no compete” clauses in contracts. As with most recent FTC rules and court cases, it brings to mind Coase’s observation to the effect that some people see a market practice they don’t understand, their immediate conclusion is that it is some sort of nefarious exercise of market power.

The FTC’s main bugaboo is that non-competes create lopsided bargaining power in favor of firms. Such bargaining power per se should not give rise to antitrust concerns. And indeed, lopsided bargaining power may be an efficiency enhancing feature that makes both firms (the supposed beneficiaries) and employees (the alleged victims) better off.

One needs to remember that there are at least two stages here. Ex ante, when the employer and employee agree on terms to a contract, which may or may not include the non-compete clause (and the employee can reject an offer that includes one), and ex post, when the employee has found a potentially superior outside opportunity and would like to get out of the restriction. The bargaining power under non-competes is unbalanced in the ex post stage.

A threshold issue is why would the employer and employee agree ex ante to an arrangement that is inefficient, i.e. reduces their joint wealth? If the benefit of the restriction to the employer is smaller than the cost to the employee, there would be a bargain without the non-compete that would make both parties better off. Crucially, this is true regardless of the distribution of bargaining power. Thus, it is highly likely that the mutual agreement to employment contracts including non-competes indicates that they are efficiency enhancing.

This then raises the question: what is the source of efficiency at the ex ante stage that induces the parties to enter into an arrangement that ties the employee’s hands and limits his/her bargaining power ex post?

Property rights economics provides an answer. In a zero transactions cost world, who owns assets is irrelevant. (The Coase Theorem.) However, if certain things are non-contractible (because, for instance, they cannot be verified by 3rd parties like courts), transactions cost are positive and the distribution of asset ownership matters, and can impact efficiency. It matters precisely because it can affect bargaining power ex post.

The owner of assets has bargaining power over non-owners with whom he contracts: he can withhold the employee’s access to the asset, thus depriving him/her from any benefits from it.

This can affect efficiency because it affects incentives to invest ex ante. The party with the bargaining power (the asset owner) extracts most of the rents from the relationship between the two parties. That affects that party’s incentive to make (non-contractible) investments that increase the amount of ex post rents to be split between employer and employee because he gets the lion’s share of these gains.

Placing the bargaining power in the hands of the party who is best able to make rent-increasing investments therefore leads to better investment decisions, and to greater surplus to be divided between the parties.

Designing contracts that allocate ex post bargaining power unevenly can therefore improve incentives ex ante, and crucially, make both sides better off–including the putatively disadvantaged party. Disadvantaged ex post, anyways.

Non-compete clauses, by the FTC’s own admission–hell, lament–allocate bargaining power. Property rights economics suggests that is likely a feature, not a bug. Crucially, it can explain why these clauses come into existence in the first place.

Firms can make non-contractible investments in their employees’ general human capital. Moreover, apropos my recent post on learning by doing, firms can increase knowledge and information by increasing their output or scale. If employees can walk across the street and sell that knowledge and information to others the firm’s incentive to invest in employees’ human capital or engage in actions that lead to employee learning by doing is reduced. However, with non-competes firms have the bargaining power that allows them to extract those ex post rents, which preserves their incentives to make non-contractible investments in generalized employee human capital, including human capital generated as the result of experience/learning by doing..

The real conundrum with non-competes is the one common to all arrangements relating to the production and dissemination of information. There are static losses if information already in existence is sold at a price above its marginal cost–effectively zero–or not sold/transferred at all. However, if information is transferred at a zero price, but is costly to produce, too little is produced. This is effectively a dynamic effect.

The same tension/trade-off exists with patents and copyrights. The ex post trade in costless information is restricted in order provide incentives to create it in the first place. Non-compete clauses that allow firms to extract most of the rents associated with knowledge production in their employees provides similar incentives.

I would also note again (as I have in earlier posts) that non-competes don’t create bargaining power ex ante, and restricting them will therefore not affect the distribution of gains at the contract formation stage. However, the magnitude of the gains to be split will be smaller in the absence of non-competes if they have the incentive effects discussed above.

Non-competes are a form of vertical restriction. Such restrictions often–and I would say typically–have efficiency rationales. Indeed, these rationales often relate to attempting to restrict free riding on information. Think of Lester Telser’s argument about resale price maintenance, or research on things like exclusive dealing and exclusive territories. Non-competes clearly relate to property rights in information, and therefore eliminating them will have baleful consequences.

But the Lina Khan FTC only has a market power hammer in its tool box, and is therefore bashing away at what is likely an efficiency enhancing contract feature that has nothing at all to do with the exercise of market power qua market power.

Print Friendly, PDF & Email


  1. I assume it’s not really an intellectual issue but merely someone trying to buy votes in an election.

    Comment by dearieme — May 2, 2024 @ 5:49 am

  2. To my lay mind, non-compete clauses are open to abuse, especially as the potential employee has no accurate way of gauging its potential cost (a company of any size would have a far better idea). Though even if I took it on faith that it was all quite true; No one ever built a statue to a critic

    I’d be interested to hear what you believe are the most detrimental, or value destructive, regulations, laws, taxes etc to the US economy, or developed economies more generally. I feel like a lot of your articles touch on this but don’t address it directly.

    Comment by lundenwic — May 2, 2024 @ 7:14 am

  3. And yet how does that explain non-compete clauses in the fast food industry? What is the human capital there, the skill of flipping burgers?

    Comment by [email protected] — May 3, 2024 @ 9:35 am

  4. Who, historically, made the greatest violation of the non compete clause?
    I suggest: Martin Luther

    Jocularity aside there are obvious imbalances in knowledge. Medieval apprentices to lucrative guilds were often sponsored by parents. (With a contract that the boys wouldn’t be fed salmon more than three times a week.)

    All young people entering the jobs market are initially ignorant and a liability, not an asset. Only when they have learned from the employer how to do the job are they an asset. If that asset is transferred to a competitor the apprentice master loses.

    Over the years we’ve tried various methods. Slavery, indentured labour, apprenticeships, technical schools, student loans, gardening leave…

    None of these are perfect. But restricting the choice doesn’t seem likely to solve the problem.

    Comment by philip — May 3, 2024 @ 3:58 pm

  5. Are we going to see a generation that: (i) Signed up for student loans, got the education, and then had the federal government pay off the loans for them, and (ii) Signed job contracts with non-compete clauses, enjoyed the extra income that such a clause brings, and then have the clause torn up by the Feds?

    Would that be the same generation where some of the wealthier ones benefit from the regressive policies on electric cars, getting capital and running costs subsidised by their less affluent peers?

    What next – will they be forgiven their mortgage loans? Offered free divorces?

    Comment by dearieme — May 4, 2024 @ 4:44 am

  6. This generation is paying tuition rates that have grown more than inflation each of the past 30 odd years. This generation is also facing home values that are at higher value-to-income ratio than anytime before. They are victims of terrible policy failures and I feel sympathy for them.

    Please, most regressive policies are those put in place that keep on benefiting the aging baby boomers, in the U.S. and most spectacularly in the U.K.

    Comment by [email protected] — May 5, 2024 @ 1:31 pm

  7. Thank you for the economics lesson. This was one of those where I saw the policy decision and my own confirmation bias told me it was probably wrong (I don’t trust this administration). But, in practice on trading floors I saw non-competes; mostly among brokerage groups who didn’t want brokers to leave to start their own gig. Sometimes they existed in local trading groups and I am quite sure they probably exist now in the electronic trading firms.

    An easy answer is pay your employee more to keep them. However, your post raises other ways to “compensate” employees. Hackman-Oldham theory would tell you to give more autonomy to people to motivate them. Job design can be changed. It’s not just about the money.

    Comment by Jeff Carter (@pointsnfigures1) — May 5, 2024 @ 5:31 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress