Labor Noncompete Agreements: Tool for Economic Efficiency or Means to Extract Value from Workers?

• Published date: 01 December 2021
• Author: David J. Balan
• Date: 01 December 2021
• DOI: 10.1177/0003603X211045443
• Subject Matter: Article

Article

Labor Noncompete Agreements:

Tool for Economic Efficiency or

Means to Extract Value from

Workers?

David J. Balan*

Abstract

A number of theoretical arguments have been offered in favor of noncompete provisions in labor

agreements. While there has been considerable empirical research on the effects of those provisions,

there has been little direct evaluation of the arguments themselves. In this article, I lay out and evaluate

three commonly heard arguments, namely, (1) the voluntary nature of labor agreements justifies a

strong inference that the terms of those agreements, including noncompete provisions, are beneficial

for both workers and firms and that they are economically efficient, 2(A) noncompetes facilitate

efficient knowledge transfer from firms to workers, and 2(B) noncompetes encourage efficient firm-

sponsored investment in worker training. These arguments, though not entirely without merit, mostly

do not survive close scrutiny, and in fact such scrutiny reveals strong arguments that point in the

opposite direction. In addition, noncompetes may cause important additional harms that are not

measured in conventional economic research.

Keywords

noncompetes, labor noncompetes, postemployment restrictive covenants, PERCs

I. Introduction

Noncompete provisions in labor agreements have become widespread in the United States.

1

In recent

years, empirical researchers have studied the effects of noncompetes on wo rker mobility, hiring,

entrepreneurship, investment, innovation, wages, and other economic outcomes. This research agenda

is quite new, and determining the true, causative effect of noncompetes on those outcomes is

* Federal Trade Commission, Washington, DC, USA

Corresponding Author:

David J. Balan, Federal Trade Commission, Washington, DC 20580, USA.

Email: dbalan@ftc.gov

1. See Evan P. Starr et al, Noncompete Agreements in the US Labor Force, 64 J.L. & ECON. 53–84 (2021); ALEXANDER J. S.

COLVIN & HEIDI SHIERHOLZ, NONCOMPETE AGREEMENTS (Econ. Policy Inst. 2019).

The Antitrust Bulletin

2021, Vol. 66(4) 593–608

ªThe Author(s) 2021

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0003603X211045443

journals.sagepub.com/home/abx

challenging. But recognizing these limitations, the evidence as it exists today, while somewhat mixed,

generally shows noncompetes to be economically harmful and not beneficial.

This empirical evidence must be interpreted in light of the strength of the theoretical arguments for

or against noncompetes. If there were strong theoretical arguments in their favor, the empirical

evidence accumulated to date may not be sufficient to convincingly demonstrate that noncompetes

are harmful on balance. But if the theoretical arguments in favor of noncompetes are weak, or if there

are strong theoretical arguments against them, then the theory and the empirical evidence would both

point in the same direction, strongly indicating that noncompetes are likely to be harmful and even

more strongly indicating that they are unlikely to be highly beneficial.

2

The purpose of this article is to provide a critical evaluation of those theoretical arguments. There

are three major arguments that are commonly offered in favor of noncompetes, namely, (1) the

voluntary nature of labor agreements justifies a strong inference that the terms of those agreements,

including noncompete provisions, are mutually beneficial to workers and firms and that they are

economically efficient in the sense of increasing total economic surplus, 2(A) noncompetes facilitate

efficient knowledge transfer from firms to workers, and 2(B) noncompetes encourage efficient firm-

sponsored investment in worker training.

There is a substantial body of literature that makes or develops these theoretical arguments.

3

The

structure of this article is to enumerate, explain, and respond to these arguments one by one.

4

To summarize my conclusions, these arguments sound plausible and have some limited merit, but

all three largely fail upon close scrutiny, and in fact such scrutiny reveals strong arguments to the

contrary. This theoretical conclusion, combined with the empirical evidence (discussed below) that

mostly finds noncompetes to be harmful, together constitute strong reason to believe that noncompetes

are in fact harmful, and even stronger reason to believe that they are not highly beneficial such that

restricting or banning them would risk major economic damage.

5

Moreover, noncompetes may cause harms that are not generally within the purview of economics

and that are not normally studied in economic research. A worker who is bound by a noncompete has a

large barrier to leaving a firm (on top of other barriers that likely exist), rendering them less able to

avoid or resist mistreatment at their firm, including true exploitation or abuse by a predatory employer

2. In Bayesian terms, if the theoretical arguments in favor of noncompetes are strong, then the priors would be strong that

noncompetes are highly beneficial, and it would take a large amount of contrary evidence to overturn those priors. But if

those arguments are weak, and/or if there are strong arguments that noncompetes are harmful, then the priors would be that

noncompetes are harmful (or at least not highly beneficial), and it would take a large amount of contrary evidence to overturn

those priors.

3. Perhaps the clearest exposition of Argument #1 is at David D. Friedman, Non-Competition Agreements: Some Alternative

Explanations, daviddfriedman.com, April 2, 1991, http://www.daviddfriedman.com/Academic/non-comp/Non-

Competition.html. See also Maureen B. Callahan, Post-Employment Restraint Agreements: A Reassessment, 52 U. CHI.

L. REV. 703–28 (1985). Articles that advance Argument #2(A) include Jonathan M. Barnett & Ted Sichelman, The Case for

Noncompetes, 87 U. CHI. L. REV. 953–1049 (2020); and Brandon S. Long, Protecting Employer Investment in Training:

Noncompetes vs. Repayment Agreements, 54 DUKE L.J. 1295–320 (2005). Articles that advance Argument #2(B) include

Paul H. Rubin & Peter Shedd, Human Capital and Covenants Not to Compete,10J.L

EGAL STUD. 93–110 (1981); and Long,

supra note 3.

4. Points similar to some of those made in this article can be found in Eric A. Posner, The Antitrust Challenge to Covenants Not

to Compete in Employment Contracts, 83 ANTITRUST L.J. 165–200 (2020) and in the survey articles referenced in note 6.

See also NON-COMPETE CONTRACTS: ECONOMIC EFFECTS AND POLICY IMPLICATIONS (U.S. DEPT OF THE

TREASURY, OFFICE OF ECON. POLICY 2016); NON-COMPETE AGREEMENTS: ANALYSIS OF THE USAGE,

POTENTIAL ISSUES, AND STATE RESPONSES (The White House 2016).

5. Returning to the Bayesian framing from note 2, the conclusion of this article is that the correct priors are that noncompetes are

likely to be harmful and are very unlikely to be highly beneficial. Given these priors, the existing empirical evidence provides

little grounds for updating. In other words, there is a consonance rather than a tension between the empirical evidence and the

theory.

594 The Antitrust Bulletin 66(4)