Locked in or held up? The economics of employee non-compete clauses

This article was originally published in TL4's ThoughtLeaders4Competition Magazine. 

Non-compete clauses (“NCAs”) that restrict employees from working for a competitor or starting a competing business after leaving their employer have come under increasing scrutiny by policymakers and competition authorities. The U.K. Department for Business and Trade (“DBT”) published a working paper on NCAs in November 2025, which raised questions about whether these clauses restrict competition.¹ The U.S. Federal Trade Commission (“FTC”) introduced a rule banning NCAs in 2024,² but, despite this rule subsequently being revoked,³ the FTC has stated that anticompetitive NCAs are still a focus.⁴ A ban on NCAs for certain workers, especially below a salary threshold, has also been proposed in Australia.⁵ 

By contrast, the broad view from the European Commission (“EC”) is that NCAs are “considered less restrictive ways of protecting employers’ investments in training or non-patent IP as, unlike no-poach agreements, they are transparent vis-a-vis employees”.⁶ 

The economics of NCAs are complex and fact‑specific. On the one hand, relaxing the restrictions that NCAs create allows locked-in workers to move to where they are most productive and where they prefer to work. This better allocation of a factor of production strengthens competition in product markets, encourages new start-ups,⁷ and ultimately fosters innovation and economic growth.⁸ 

On the other hand, employers may have valid reasons for using NCAs. When a worker leaves, the impact goes beyond simply the frictions of finding a replacement employee: the returns to employee training are reduced and employers are disadvantaged when their investment in training benefits a competitor. This can lead to a “hold-up” problem⁹ of employers foregoing investment in employees, consequently reducing innovation¹⁰ and undermining productivity.¹¹ In other words, an employer may have made an employee-specific investment, which it may not be able to recoup if the employee leaves and joins another firm that is able to “free-ride” on a rival employer’s investment in training.

Separating lock-in from hold-up: Human capital theory

Given the potential benefits and disbenefits of NCAs, it is important to understand situations in which these clauses may hinder competition and innovation, and those in which they may mitigate a hold-up and free-riding problem. A key distinguishing factor is the type of human capital – a worker’s knowledge, skills, and experience – developed in a given role. Workers can perform very different roles. For instance, consider a stylised example of a general accountant and an IT system technician working on a firm’s proprietary system. The accountant’s skills may typically be applicable when working across a range of firms: if the accountant moves to a new employer, that new employer may free-ride on the investment the accountant’s previous employer made in the accountant’s skills. On the other hand, the IT technician’s skills may (depending on the facts) be more closely tied to the specific systems and software that their employer uses. If the technician switches jobs, the investment their original employer made will not be as beneficial to the new employer, and so there will be less free-riding. In this stylised example, absent an NCA, the accountant’s original employer may be less likely to invest in the accountant’s skills, whereas the IT technician’s employer would be more likely to invest, regardless of whether an NCA is in place. In short, the risk of hold-up varies depending on the type of human capital developed on the job.

Generalising from this example, economists have distinguished between three types of human capital:

1. Firm-specific human capital^12:– skills valuable only within a single firm, such as familiarity with internal processes or proprietary tools. For example, an employee trained on a company’s custom software gains expertise that is largely unusable elsewhere. As these skills cannot be easily transferred, employees cannot leverage them to move to another firm. Therefore, the risk of an employer holding up investment in these skills is generally low.
2. Industry-specific human capital¹³ – skills valuable across firms in the same sector, such as proficiency with standard industry software or knowledge of sector regulations. For example, engineers trained on widely used Computer-Aided Design (CAD) software can apply these skills at other firms. The hold‑up problem exists, but its significance depends on labour market specifics, in particular on how easily a worker can move to another firm within the industry that will value the training received.¹⁴
3. General human capital¹⁵ – skills transferable across industries and firms, such as people management, project management, or public speaking. The hold-up problem may be most pronounced here, but it may be difficult to distinguish skills gained through employer training from those acquired independently. This makes it harder to assess in practice whether NCAs are justified.

This taxonomy provides a framework to assess whether a given employer training investment could justify an NCA. NCAs are more likely to resolve hold-up problems where industry specific or general human capital is created, and less likely to do so where firm‑specific human capital is involved. Consequently, NCAs may be more likely to be anticompetitive where firm‑specific human capital is created, though, even for industry‑specific and general human capital, NCAs could have anticompetitive effects. 

There is a further question of whether the specifics of an NCA are proportionate, and whether any less anticompetitive clause or employer conduct could plausibly deliver the same benefits. Even if an employer makes material investments in industry‑specific skills of an employee, there will be limits to the duration or exact conditions of an NCA that such investments can justify.

Policy and practice: Where to go from here

A general ban on NCAs may have adverse effects, as it could lead some employers to invest less when rivals can free‑ride. The extent of firm‑specific, industry‑specific, and general human capital creation differs across industries, and even across firms. Consequently, one‑size‑fits‑all interventions may yield benefits in some markets and harm in others. 

For authorities concerned with potentially anticompetitive NCAs in specific markets or for certain types of employees, a starting point could be to review the nature of employer-sponsored training provided by firms within an industry. Training materials could be examined to identify the type of human capital involved, and whether an NCA is likely to resolve a hold-up problem. Comparing the format and content of programmes across similar companies can also be helpful: the less overlap there is, the more likely the training generates firm‑specific human capital. However, training does not need to be formal or incur easily identifiable monetary costs to employers. For example, employees may be assigned tasks they are not yet fully prepared to perform consistently, while working alongside a peer or mentor in order to gain experience. Such assignments are also an investment for employers, as they entail the opportunity cost of time that could otherwise be used more productively (or savings on the labour costs of the employee being trained).

The harm from limiting the scope of NCA may also be mitigated depending on the viability of less restrictive alternatives. One approach to addressing investment hold-up concerns for formal training is the use of claw-back provisions for employees who resign within a certain period after being up-skilled.¹⁶ For example, if a firm funds training, it could require employees to remain with the firm for a set period afterward to recoup the cost, or to pay deferred compensation or bonuses once the service period is completed, subject to the enforceability of such contracts. However, as noted above, not all training is formal.

Conclusion 

A blanket rule or simple heuristics based on employee salary or firm size are unlikely to be appropriate to address potential anticompetitive effects of NCAs,¹⁷ while preserving the investment benefits these clauses can support. Rather, a more appropriate approach would be to focus on specific markets or types of employees where restrictive NCAs are commonplace. In these cases, assessing the type of human capital created is more appropriate to prioritise where interventions that limit or restrict NCAs should be made. This would both protect employer investment from being held up and minimise unnecessary restrictions on worker mobility.

1. Working paper on options for reform of non-compete clauses in employment contracts | GOV.UK
2. FTC Announces Rule Banning Non-competes | Federal Trade Commission
3. Federal Trade Commission Files to Accede to Vacatur of Non-Compete Clause Rule | Federal Trade Commission
4. Moving Forward: Protecting Workers from Anticompetitive Non-compete Agreements, noting the statement on page 11 that “the FTC should focus its enforcement resources on those non-compete agreements that do not advance a procompetitive purpose or else are not narrowly tailored to advance a procompetitive purpose.”
5. Major employers push back against Australian non-compete ban – Global Competition Review
6. Competition policy brief – Antitrust in Labour Markets, p5.
7. FTC Announces Rule Banning Non-competes | Federal Trade Commission, noting the statement that “[t]he FTC estimates that the final rule banning non-competes will lead to new business formation growing by 2.7% per year, resulting in more than 8,500 additional new businesses created each year.”
8. Working paper on options for reform of non-compete clauses in employment contracts |  GOV.UK, paragraph 8.
9. More generally, the hold-up problem arises when one party invests in a relationship, but the other party can take advantage of that investment afterwards (Moore, J., & Hart, O. (1988). Incomplete Contracts and Renegotiation. Econometrica, 56(4), 755-785).
10. FTC-Noncompete-Comment-Letter_FINAL_04.17.23.pdf, page 3.
11. Competition and market power in UK labour markets, paragraph 6.39. The Competition and Markets Authority (“CMA”) found on average workers with a non-compete agreement are slightly more likely than those without to receive formal on‑the‑job training, amounting to five or more training days per year on average.
12. Becker, G. S. (1962). Investment in Human Capital: A Theoretical Analysis. Journal of Political Economy, 70(5), 9–49.
13. Neal, D. (1995). Industry‑Specific Human Capital: Evidence from Displaced Workers. Journal of Labor Economics, 13(4), 653–677
14. In sectors with higher unemployment, a trained worker cannot easily find another employer willing to compensate for the training received from their current firm. Since switching jobs is more costly in such sectors, the current employer has greater monopsony power and can capture a larger share of the returns on its training investment (Acemoglu, D., & Pischke, J.-S. (1999). The structure of wages and investment in general training, Journal of Political Economy, 107(3), 539–572). As a result, the hold-up problem is likely less severe. Similar logic applies in more concentrated labour markets: if there is only one firm in a (local) labour market, or only one firm that hires a given occupation, industry‑specific human capital effectively becomes firm‑specific human capital.
15. Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education. New York, NY: Columbia University Press. The classic definition of general human capital is skills that are transferable to other firms. While the term is often used more broadly to include industry‑specific human capital, we restrict it in this article to skills transferable across sectors.
16. For example, this is suggested in the CMA’s response to the DBT consultation: CMA response to working paper on options for reform of non-compete clauses in employment contracts | GOV.UK, paragraph 14.
17. For example, this is proposed in Australia and in the DBT consultation.