Guilt by (Anticompetitive) Association: Criminal Enforcement as a Response to Labor Monopsony

Marissa Piccolo¹Marissa Piccolo is a J.D. Candidate at the University of Chicago Law School, Class of 2022.

* * *

This piece is part of the University of Chicago Law Review Symposium 2022: Law and Labor Market Power.

At the close of 2020 and start of 2021, for the first time in its history, the Department of Justice (DOJ) brought criminal antitrust charges against monopsonistic employers for entering into wage-fixing and no-poach agreements. Both employers pled not guilty and were acquitted in April 2022. In light of these developments, this Essay evaluates criminal law enforcement as a response to the problem of labor monopsonies. From a due process perspective, no-poach agreements are not conclusively per se illegal under the Sherman Act, and prosecuting activities governed by the alternative standard—the rule of reason—raises fair notice concerns. The DOJ, however, has thus far been disciplined and only prosecuted horizontal agreements that are presumptively anticompetitive—even if they are in the novel context of labor markets.

The principle of fair notice still has implications, however, for determining whether criminal law enforcement creates the optimal level of deterrence against labor monopsonies or unduly chills productive economic activity. All in all, there are critical differences between labor monopsonies and other anticompetitive arrangements that render criminal prosecution less likely to overdeter. Labor monopsonies harm workers’ consumer welfare, and there is less reason to defer to firms when the employee should be the most relevant decision maker.

I.  Background

A. The Origins of Criminal Antitrust Enforcement

Section 1 of the Sherman Antitrust Act of 1890 prohibits arrangements “in restraint of trade or commerce.” Violations were originally punishable through civil penalties or as criminal misdemeanors, but civil and criminal enforcement actions were rare. From 1890 to 1904, the DOJ brought only seventeen civil and six criminal cases.²Richard A. Posner, A Statistical Study of Antitrust Enforcement, 13 J.L. & Econ. 365, 385 (1970).

A few key moments marked a turn towards criminal antitrust enforcement. One was the leadership of Assistant Attorney General Thurmon Arnold from 1938 to 1943. Arnold—a former Army lieutenant and New Deal Democrat—was of the view that “civil [enforcement is] little more than a form of unemployment relief for lawyers since it carries no penalties,” whereas “criminal prosecution is the only effective instrument under existing statutes.” Accordingly, his Antitrust Division “brought approximately 340 Section 1 cases, 231 of which were criminal prosecutions,” targeting “old-fashioned price-fixing conspiracies” but also some “industries [like the insurance business] generally thought [to be] exempt from antitrust laws.”

 Another was the rise of the OPEC oil cartel and political pushback following spikes in gas prices in the 1970s. In response, Congress “made Sherman Act offenses felonies, raised the maximum corporate fine twentyfold (from $50,000 to $1 million) and increased maximum jail sentences for individuals from one to three years.”³Donald I. Baker, The Use of Criminal Law Remedies to Deter and Punish Cartels by Bid-Rigging, 69 Geo. Wash. L. Rev. 693, 695 (2001).

Throughout, sentencing courts and DOJ officials struggled to define which violations of Section 1 should be subject to civil liability and which should instead be subject to criminal sanctions. A consensus emerged that criminal sanctions should be reserved for Section 1 violations that are per se illegal, including “hard-core violations” like price-fixing and market allocation that are so nakedly anticompetitive they are presumed unlawful. By contrast, anticompetitive behavior with redeeming procompetitive benefits became governed by the rule of reason.

Largely in response to perceptions that Arnold had been overzealous, the DOJ implemented a policy to only criminally prosecute “willful violations” of the antitrust laws, which would occur: (1) “if the rules of law alleged to have been violated are clear and established—describing per se offenses” such as price-fixing, or (2) “if the acts of the defendants show intentional violations.”

B. Present Day: First Criminal Prosecutions of Wage-Fixing and No-Poach Agreements

1. The DOJ Stakes Out Its Position.

While the debate over the merits of civil versus criminal antitrust enforcement continued, a “litigation gap” persisted between product- and labor-market antitrust enforcement. Following the 2008 recession, however, the Obama administration became increasingly interested in using antitrust to address wage stagnation. A “turning point” was a 2010 civil suit against “Google, Apple, and other top Silicon Valley tech firms for agreeing not to solicit each other’s software engineers,” which eventually resulted in an over $400 million dollar payout to victims.

With an eye toward future criminal enforcement, the Obama Administration’s DOJ and Federal Trade Commission issued a joint guidance document in October 2016 alerting human-resources professionals that “the DOJ intends to proceed criminally against naked wage-fixing or no-poaching agreements” because they “are per se illegal.” This decision to limit criminal liability to only per se violations aligned with longstanding DOJ policy. Specifically, the guidance document distinguished between per se violations and agreements “reasonably necessary to a larger legitimate collaboration between the employers,” seemingly referring to agreements amongst franchisors and franchisees thought to be entitled to the rule of reason.

In a 2019 civil suit, the DOJ fleshed out its position, stating that “franchise no-poach agreements that are ancillary” are entitled to rule-of-reason review, whereas “the per se rule applies if the no-poach agreement is a naked horizontal restraint.” In that case, the plaintiffs argued that these agreements resembled traditional cartel collusion. Conversely, the DOJ Antitrust Division intervened on behalf of the defendants to maintain that “the restrictions were vertical restraints” imposed by the franchisor and, even if they could be characterized as “horizontal no-hire agreement between franchisees . . . [they] would still be considered ancillary if reasonably necessary to the legitimate franchise collaboration.”

This prompted the question: what “collaborations” beyond franchise agreements are entitled to rule-of-reason analysis because they are “vertical” or “ancillary to certain legitimate agreements?” Despite—but more likely because of—this uncertainty, the DOJ has thus far been disciplined in only prosecuting activity that is presumptively anticompetitive and therefore per se illegal—even if courts have not definitively held so yet.

2. The First Indictments: Jindal and DaVita.

Wage-fixing cases amongst nonfranchising entities present the strongest candidates for per se treatment. Accordingly, the DOJ brought its very first criminal case against a labor monopsony engaged in such conduct in United States v. Jindal (E.D. Tex. 2020). Neeraj Jindal owned a physical therapist staffing company and was alleged to have conspired with the owner of a competing company to “provide[ ] and receive[ ] non-public rates paid to” employees, “communicate[ ] about rate decreases; discuss[ ] and agree[ ] to decrease rates paid to” employees and “implement[ ] rate decreases in accordance with the agreement reached.” The DOJ identified smoking-gun text messages among competing business owners stating “I think we’re going to lower [employee] rates” and “I’ll do it with u”—and even one including a thumbs-up emoji. In rejecting defendants’ motion to dismiss, the court held that the conduct was “perilously close” to price-fixing. In reality, “perilously close” is an undersell: the only difference between price-fixing and wage-fixing is that the latter exists in the context of labor monopsonies rather than market monopolies. The defendant won an acquittal at trial, however, suggesting that despite strong evidence of wrongdoing, juries may not “find anticompetitive conduct in labour markets serious enough to merit a criminal conviction.”

Prosecutions for entering into no-poach agreements are more complex. It is contested whether all no-poach agreements are presumptively anticompetitive and therefore should be per se illegal. Unlike wage-fixing, no-poach agreements are not as “perilously close” to price-fixing, which is traditionally considered a clear cartel activity. Further, no-poach agreements are common in the franchisor-franchisee context because they arguably facilitate training and uniformity across branches. After all, in 2019 the DOJ staked out its position that “the typical franchise relationship itself is a legitimate business collaboration” and likely valid under the rule of reason.

Accordingly, the district court presiding over the first criminal prosecution of a no-poach agreement was skeptical of the DOJ’s position. In United States v. DaVita (D. Colo. 2021), the DOJ brought criminal charges against a healthcare company, DaVita Inc.; its CEO, Kent Thiry; and its competitor Surgical Care Affiliates (SCA). Like in Jindal, there were smoking-gun emails, such as “I thought there was a gentlemen’s agreement between us [SCA] and DaVita about poaching talent.” In opposing the defendants’ motion to dismiss, the government repeated its 2019 position that no-poach franchise agreements “raise unique ancillarity issues” and are “exempt from the per se rule” but argued that the collusive behavior at issue was different and amounted to a naked agreement among competitors not to compete.⁴Opposition to Defendants’ Motion to Dismiss at 14, United States v. DaVita Inc., 2022 WL 266759 (D. Colo. Jan. 28, 2022) (No. 21-00229).

 Judge R. Brooke Jackson, in denying the defendants’ motion to dismiss, did not adopt the clear line rule advocated for by the DOJ exempting franchise no-poach agreements and relegating others to per se treatment. She instead characterized per se treatment of no-poach agreements as generally unwarranted. Taking a “middle ground approach, Judge Jackson held that “if naked non-solicitation agreements or no-hire agreements allocate the market, they are per se unreasonable.” But because “the indictment does allege that the non-solicitation agreement allocated the market,” Judge Jackson allowed the case to go forward. But eventually, DaVita also ended in acquittal in April 2022, meaning the DOJ has lost the first two criminal cases it has brought.

II.  A Lack of Fair Notice? 

There is a colorable due process concern that defendants did not have fair notice that their monopsonistic conduct was unlawful. Although the concept of fair notice might bear on whether these criminal prosecutions are optimal from a deterrence perspective—as discussed in the next Part—a due-process challenge along these lines would be unavailing. The text of the Sherman Act is vague, only prohibiting arrangements “in restraint of trade.” United States v. Lanier (1997) tells us, however, that we must consider whether the statute “as construed” renders “it reasonably clear at the relevant time that the defendant’s conduct was criminal.” Therefore, defendants charged with per se violations cannot tenably raise a void-for-vagueness argument. On the other hand, those engaged in conduct subject to the rule of reason arguably can given the rule of reason’s intensively fact-specific and nebulous inquiry. As Professor Daniel Sokol put it, “the lack of developed antitrust case law as to the rule of reason amplifies the potential vagueness of the application of criminal sanctions.” After all, it is DOJ policy to not bring charges against activity entitled to the rule of reason.

Accordingly, the activities prosecuted thus far qualify as per se violations of Section 1. In Jindal, the DOJ analogized defendants’ wage-fixing to price-fixing—perhaps the seminal horizontal agreement in restraint of trade. Over a century of U.S. case law has firmly established the per se illegality of price-fixing. Notably, the fact that wages—the price of labor—are at issue means that the application of Section 1 in Jindal is novel, but that fact alone does not render the statute unacceptable from a Fifth Amendment perspective. As the court noted, “the lack of criminal judicial decisions only indicates Defendants’ unlucky status as the first two individuals that the Government has prosecuted for this type of conduct.”

The same holds true for the first prosecution of a no-poach agreement in DaVita. Despite the court’s initial skepticism, the pertinent agreement is one among competitors to not recruit one another’s employees. No-poach agreements thus closely resemble nonsolicitation agreements to carve up a market and not pursue another’s customers, which have historically been regarded as per se illegal.

Although parties to some no-poaching agreements have a stronger claim that the rule of reason governs—and therefore a stronger void-for-vagueness claim if prosecuted—the DOJ has avoided prosecuting them. The closest call was the latest prosecution in United States v. Patel (D. Conn. 2021), where five senior executives of outsource engineering suppliers entered into no-poach agreements. Because each of the suppliers had a contract with the same major aerospace company in Connecticut, there was a plausible argument that a “vertical relationship between the aerospace company and each of the engineering suppliers” existed. This would have rendered the no-poach agreements reasonably ancillary to a legitimate business endeavor, such as maintaining the “allocat[ion of] employees working on projects for the company.” Even with this gloss, the known facts reveal that the scope of the no-poach agreements was broad and covered all employees, not just ones involved with the aerospace company. For example, an email from the CEO of one supplier to another stated: “Our general aim is NOT to recruit from the local ‘competition’ because no one wins; salaries rise, the workforce get [sic] unstable, and our margins all get hurt.” The Sherman Act patently bars this protectionism.

If the DOJ was to pursue charges for conduct squarely entitled to rule-of-reason review, there would be greater due process concerns. Of course, it is unlikely that many loss-averse prosecutors would do so given the substantial obstacles to a successful prosecution of such conduct. It is difficult, if not impossible, to prove that conduct does not have any procompetitive benefits beyond a reasonable doubt. As Professor Lawrence Sloan notes, “the rule of reason in operation is based on doubt, due to shifting burdens of proof for plaintiff and defendant.”

The more serious concern attends prosecuting and trying activity under a per se theory that should be entitled to rule of reason. The DOJ has sought to avoid this, distinguishing between per se violations and agreements “reasonably necessary to a separate, legitimate business transaction or collaboration between the companies.” But that line is not so clear. As DaVita shows, even activity that is ultimately entitled to per se treatment requires a preliminary inquiry resembling the rule-of-reason analysis. Therefore, even if the lack of notice is insufficient to raise constitutional issues, there might still be a literal lack of notice that might affect how parties behave and chill economically productive behavior. The DOJ attempted to avoid this by releasing its 2016 Guidance, but there is no doubt the recent criminal prosecutions were more attention-grabbing for employers.

III.  Finding Optimal Deterrence

Proceeding on the assumption that labor monopsonies are harmful and lead to reduced wages, more rigorous antitrust enforcement is needed. But is criminal enforcement well suited to the task of achieving the optimal level of deterrence?

One theory against criminal liability is that civil enforcement will do the trick if properly pursued. Criminalization raises the cost of engaging in illegal activity. On some occasions, it brings society “closer to [the] optimal [level of] deterrence because it increases the severity of penalties.” This is only true, however, when the preexisting penalties were inadequate, not when they have simply laid dormant in the way that labor markets have been neglected by antitrust law.

The argument is as follows: so long as civil fines are imposed and properly calibrated, they will deter harmful monopsonistic activity. Plus, society gets the benefit of having the funds placed “on the benefit side of the social ledger.” “[A] term of imprisonment, on the other hand, yields no comparable social revenue” and has the disruptive effect of incarceration for the individual and their community. Of course, this is only true if “the optimal fine [does not] exceed the offender’s ability to pay.”

That being said, there are tools unique to criminal enforcement that aid deterrence and account for specific issues in antitrust. A major difficulty for enforcement is the information asymmetry between the government and firms. Agreements to restrain trade are secret, and parallel conduct provides no more than a hunch that there is collusive behavior occurring below the surface. As Professors David Besanko and Daniel Spulber note, “asymmetric information can be a significant factor in the decision to tolerate some degree of collusion even though price fixing is illegal per se.”⁵David Besnako & Daniel F. Spulber, Antitrust Enforcement Under Asymmetric Information, 99 Econ. J. 408, 408 (1989). But unlike in civil enforcement, criminal tools, such as grand jury investigations, allow prosecutors to more quickly and more efficiently identify when collusive activity is occurring and to either pursue penalties or drop the case.

Criminal enforcement, however, poses a greater risk of both under and overinclusiveness than civil enforcement does. First, its severity can lead to uneven application. There are concerns about perceptions of harshness and collateral consequences, which can pull in opposite directions. For example, the DOJ Manual on Principles of Federal Prosecution of Business Organizations states prosecutors must consider “collateral consequences, including whether there is disproportionate harm to shareholders, pension holders, employees, and others not proven personally culpable.” Prosecuting larger companies might seem less harsh, as they have sufficient resources to mount a strong defense, but there also might be more collateral consequences. As Professor Eric Posner noted, the wage-fixing between the CEOs in the 2010 Silicon Valley case was “blatant,” and “if the conspirators had fixed prices rather than wages, they would probably have gone to jail.” Further, when faced with high profile defendants, like Steve Jobs, the DOJ might be less willing to pursue criminal charges for fear of the headlines it would create and the possible public backlash that could ensue. This might lead the DOJ to go after lower-profile employers instead of monster monopsonies like Amazon, which are arguably more harmful to workers. In sum, there are considerations unique to the criminal context that might prevent prosecutors from going after those who should be most deterred.

On the other hand, there is the parallel risk, common to white-collar prosecutions, of being overinclusive and chilling economically productive activity. The fair-notice issues previously mentioned are not due process violations, but ambiguity over whether activity that is per se illegal—and therefore fair game under DOJ guidelines—can overdeter firms from entering into otherwise procompetitive arrangements where a wage-fixing or no-poach agreement may be ancillary. The risk of incarceration and stigma of a criminal fine makes firms more sensitive to potential criminal penalties.

But there are critical differences between labor monopsonies and other monopolies that mean that criminal prosecution is less likely to overdeter economically productive activity. In the context of consumer goods, there has been a marked willingness to excuse some anticompetitive behavior if it results in lower prices. Known as the consumer-welfare standard, this approach has come to dominate antitrust enforcement over the last few decades and counsels against deterrence through criminal enforcement except in extreme cases.

But the underlying logic for such hesitation does not apply in the context of labor monopsonies. In that context, the consumers in the market—that is, the employees—are themselves being devalued. Even if one applies a consumer-welfare lens, low prices are not good for the worker-consumer. A lower price is a lower wage, meaning that worker-consumers can now purchase fewer goods in the marketplace. Labor monopsonies are also worse for all consumers because the monopsonistic employer “reduces labor costs by hiring fewer people,” which in turn means “output declines.” That decline in output means that “consumers normally pay higher prices for goods and services sold by labor monopsonists.”

Moreover, workers—unlike consumer goods—are economic decision-makers. Under a consumer-welfare approach, antitrust law may defer to firms that take advantage of economies of scale—perhaps by establishing agreements between manufacturers and distributors or suppliers—even though those agreements might resemble anticompetitive arrangements. But in the labor monopsony context, there is a relevant decision maker with valuable information about economic productivity outside that arrangement: the employee. Therefore, society should not be as concerned about chilling the employer’s decision-making. We should value employee decision-making based on market forces, such as employee decisions about what firm they might like to move to. After all, employees themselves have valuable information about where they will be more economically productive as they are sensitive to wage increases.

* * *

In sum, overdeterrence through criminal liability raises fewer concerns in the labor monopsony context than in the traditional monopoly context. This raises the specter of further debates about how criminal liability deters wrongdoing at an individual and organizational level as well as debates about when fines versus incarceration are appropriate. Acquittals in the face of strong evidence suggest jurors had concerns about fairness. For now, this Essay has endeavored to evaluate the budding use of criminal antitrust enforcement to address the problem of labor monopsonies. As long as the DOJ remains disciplined and does not prosecute activity entitled to rule-of-reason review, the criminal strategy affords the constitutionally required notice and fills a gap in the DOJ’s civil enforcement policy.

* * *

Marissa Piccolo is a J.D. Candidate at the University of Chicago Law School, Class of 2022.