A number of non-cartel antitrust infringements remain crimes under US law, even if they are not prosecuted in practice. This article, available here, deals with the implications of recent claims for increased antitrust enforcement for the application of such provisions. A natural extension of enforcement would be to advocate the use of criminal sanctions for various antitrust violations outside of collusion which are “on the books” but have not been used in over a generation. The article argues that a return to the criminalisation of non-collusion related antitrust abuses is problematic not only as a matter of optimal deterrence, but also unconstitutional as a matter of law.
Section one describes how antitrust criminalisation is a form of achieving deterrence.
Antitrust enforcement builds on models of optimal deterrence. Under an optimal deterrence antitrust framework, a firm or individual will be deterred where the expected costs of illegal activity, taking into account the probability of detection and magnitude of the penalties, exceed the expected benefits of such activity. Criminalisation is thought to bring antitrust policy closer to optimal deterrence because it increases the severity of penalties.
The basis for holding both individuals and firms accountable is that, by doing so, antitrust is better able to address the different organisational and individual incentives and motivations for infringing competition law. Criminalisation creates risks for firms in terms of stock market return-based penalties, reputational penalties, debarment, and government fines. The risk of criminal sanctions and their repercussions create incentives for firms to monitor their agents to ensure antitrust compliance. However, due to agent-principal issues, the incentives of a firm may be different from those of its agents. Hence, criminal penalties for individuals may be appropriate.
However, it has long been thought that criminal sanctions are only appropriate for hard-core conduct. In areas of traditional antitrust civil enforcement (for example, bundling, exclusive dealing, tying, resale price maintenance (RPM), non-price restraints, et cetera), the case for incarceration is weaker. In non-collusion cases, criminalisation might deter the very sort of risk taking that should be rewarded in a market economy. Further, imposing monetary fines as regards behaviour that might have some offsetting procompetitive justifications avoids creating significant costs to society and serves to compensate victims of anticompetitive conduct.
Section two introduces the US criminal antitrust regime and places it in historical context.
Under the Sherman Act as enacted in 1890, all antitrust criminal offences were misdemeanours, which meant that they could result in up to one year of incarceration. However, criminal enforcement of antitrust was originally quite rare even for cartels. Members of cartels were incarcerated once in 1921, but not again until 1959. When there was significant criminal enforcement for non-cartel matters, the criminal penalties imposed were fines rather than jail time, and not very severe fines at that.
During the first fifty years of antitrust enforcement under the Sherman Act, of 252 criminal prosecutions only twenty-four resulted in jail sentences, and only eleven of those involved businesspersons (the rest were trade-union leaders). Moreover, ten of the eleven cases involved acts of violence, threats, and other forms of intimidation. From 1940 to 1955, only eleven prison sentences were imposed, and in almost every case the sentences were suspended. It was not until 1959 that a prison sentence was imposed for price-fixing conduct in which no acts of violence or union misconduct were involved. In that case, the court imposed a ninety-day prison term on four individuals and fined each $5,000.
In 1974, the Antitrust Procedures and Penalties Act (APPA) transformed antitrust criminal penalties into felonies with a maximum term of imprisonment of up to three years. These new rules were met with increased enforcement vigour on the part of the Department of Justice (DOJ). In 1977, the DOJ promulgated guidelines that pushed for higher jail terms for antitrust offences, particularly for cartels. In 1981, a court finally imposed the maximum prison term of three years as a penalty for collusion.
Since then, there has been an incremental increase in both fines and criminal enforcement against collusive practices. According to empirical work by Joseph Gallo and his co-authors, incarceration was very rare in the pre-1963 antitrust era, in which prison was imposed in only 24 of 463 criminal convictions (i.e. in 5.1% of criminal convictions). Similarly, during the 1963 to 1973 period, 4.8% of criminal convictions resulted in incarceration. The APPA changed the nature of enforcement. Since its passage through 1984 (the end of the Reagan administration’s first term), 20% of criminal convictions resulted in incarceration.
All of these convictions and jail sentences were for collusion, with the exception of a couple of monopolisation convictions in 1969 (six months in jail) and 1973 (thirty days in jail). None of these jail sentences for monopolisation occurred after antitrust offences became felonies. In effect, the record of enforcement of criminal cases outside of explicit collusion has been limited since 1945. More than a generation has passed since there have been any criminal indictments, let alone jail time, for a non-collusion case. The DOJ has brought a single criminal non-collusion antitrust case since the introduction of felony penalties. In 1978, the DOJ indicted a firm for criminal RPM, which resulted in a nolo contendere plea and a fine.
Section three explores the implication of antitrust’s goals for criminalisation.
In its earlier years, US courts identified multiple goals inherent in competition law, while encouraging behaviour that was not consumer welfare enhancing. Such goals include ‘to perpetuate and preserve, for its own sake and in spite of possible cost, an organization of industry in small units which can effectively compete with each other’(United States v. Aluminum Co. of America); and the protection of viable, small, locally owned businesses, even if occasional higher costs and prices might result from the maintenance of fragmented industries and markets (Brown Shoe Co. v. United States), among others. During a period between 1940 and 1970, courts were concerned about competitors (even inefficient ones) rather than consumers. Courts ignored merger efficiencies and treated them as unlawful; found various marketing practices for vertical price and non-price restraints to be per se illegal; significantly limited unilateral refusals to deal; limited intellectual property rights; were overinclusive in their approach to ancillary restraints; and took a view that efficient price discrimination that lowers prices for consumers should be per se illegal.
However, the sole recognised goal of antitrust for a generation has been to promote consumer welfare. This meant focusing on the economic effects of business conduct, adopting a rule of reason for most antitrust enforcement, and assessing mergers on the basis of economically-minded considerations. A related change was an increased focus on the weighing of pro- and anticompetitive harms in civil procedures, instead of criminalising potentially pro-competitive business conduct. Once the goal of antitrust became solely to promote consumer welfare, the moral and legal basis for punishing business practices changed. The idea that pursuing efficiencies was somehow immoral was laid to rest. Similarly, a change from per se illegality to a rule of reason made criminal enforcement problematic ,as it set a much lower standard for incarceration for individuals than other criminal statutes.
Criminal enforcement of non-collusive antitrust activity that is per se illegal creates two potential constitutional law problems, which are the topic of the rest of the article. As a matter of constitutional law, these two doctrines limit possible overreach by those inclined to use existing law to “get tough” on antitrust violations.
Section four deals with the desuetude challenge to the criminalisation of non-collusive conduct.
Desuetude occurs where a law loses its authority due to a lack of usage. When this lack of usage has been long enough, a “negative custom” replaces the law. Non-use for a lengthy period of time suggests either that the law is obsolete or was never legitimate in the first place. Statutory desuetude, in particular, occurs when some combination of the sustained non-application of a law, contrary practice over a significant duration of time, official disregard, and the tacit consent of public and political actors leads to the implicit repeal of that law. What exactly constitutes long use (or disuse) is not totally clear – with countries adopting rules that go from a decade to over a hundred years.
The types of cases that often are brought up in the context of desuetude are ones that address issues of morality – which goes to explain why desuetude is more common in the context of criminal rather than civil law. Non-enforcement of a statute suggests that the risk of prosecution is zero, or close to it. Thus, the sudden enforcement of obsolete statutes creates due process concerns because the statutes can be used in a discriminatory manner and because it can be said that the law did not provide notice of the criminal nature of the conduct.
The case for applying the desuetude doctrine to non-collusive practices is strong. There has been no criminal indictment for non-collusion offences since the 1970s, and no criminal indictment under the Robinson-Patman Act (a statute to protect small retailers against larger and more efficient retailers) since the late 1950s. During this period, the moral perception of antitrust criminality by antitrust agencies, the antitrust bar, and academics changed. The only discussion of criminal penalties during the current antitrust era has been reserved for collusion. The absence of notice of the criminal nature of non-collusive conduct is also problematic. The DOJ has shown prosecutorial discretion not merely by not bringing such challenges, but also by making clear that criminal enforcement is reserved for cartels
Section five addresses the vagueness challenge the criminalisation of non-collusive conduct.
Void for vagueness is a constitutional doctrine that applies where the (criminal) law in question is too vague to provide notice of the type of conduct that is to be deemed illegal. In Lanier, the U.S. Supreme Court offered guidance as to when a criminal statute may be unconstitutionally vague. The purpose of the fair warning requirements is to ensure that the “statute, either standing alone or as construed, ma[k]e it reasonably clear at the relevant time that the defendant’s conduct was criminal.”
The most important federal antitrust void for vagueness Supreme Court 1913 decision is Nash v. United States, where a defendant claimed that the term “restraint of trade” in the Sherman Act was not clear. Despite the Sherman Act not laying out the elements that must be proven in a successful criminal action, the Supreme Court rejected the void for vagueness claim. Since Nash, no party has successfully brought a void for vagueness federal antitrust case. Supreme Court decisions make clear that the nature of the inquiry matters for a void for vagueness claim, as does the fact that the economics of the conduct in question clearly suggest that the practice is anticompetitive. These cases suggest that antitrust is not vague when the conduct being undertaken is clearly illegal (e.g. per se illegality) and where this illegality is well understood. As a result, it is clear that the criminalisation of price fixing, which the Supreme Court continues to find per se illegal, is not void for vagueness.
Despite this, one could argue that the void for vagueness doctrine applies with regard to the Sherman Act for non-collusion criminal enforcement. A first reason for this is that such conduct is subject to a rule of reason approach that balances pro- and anticompetitive effects. A rule of reason burden-shifting approach in antitrust looks very different from the approach taken under criminal law with regard to burdens of proof, which requires evidence beyond a reasonable doubt. By definition, the rule of reason is anything but beyond a reasonable doubt. Rather, the rule of reason is based on doubt, due to the shifting burdens of proof that the plaintiff and defendant must meet. As such, going beyond a core criminalisation focus on per se conduct would raise grave (and probably winning) due process challenges under the void for vagueness doctrine. As the Supreme Court noted in US Gypsum (1978), “[w]ith certain exceptions for conduct regarded as per se illegal because of its unquestionably anticompetitive effects … the behavior proscribed by the Act is often difficult to distinguish from the gray zone of socially acceptable and economically justifiable business conduct.’
Vagueness is also connected to the desuetude point discussed above, since “[a] penal enactment which is linguistically clear, but has been notoriously ignored by both its administrators and the community for an unduly extended period, imparts no more fair notice of its proscriptions than a statute which is phrased in vague terms. ’
Section six concludes.
The author argues that antitrust must promote consumer welfare, and that law enforcement must bring cases that optimally deter anticompetitive conduct while avoiding bringing cases that inhibit the sort of business risk taking that promotes consumer welfare. In short, antitrust nowadays has a workable institutional structure, a singular goal that allows for a regular application of law as to both process and substance.
For a generation, criminal enforcement has been limited to those cases that involve “naked” collusion— so-called “hard-core” cartels where there is direct evidence of illegal activity. Antitrust has rowed back from imposing criminal penalties for other types of conduct. Where there are concerns regarding exclusion or predation, often these cases are ones in which the balance between anticompetitive effects and procompetitive justifications are complex and require a rule of reason analysis. To criminalise behaviour that has the potential to be efficient under a rule of reason analysis returns antitrust to an era of business inhospitability, would chill business risk taking, and could raise vagueness and desuetude issues.
I am not aware of anyone arguing that Google executives should go to jail for exclusionary practices, so I wonder how much this paper takes on a strawman or might be seeking to pre-empt future arguments to this effect.
At the same time, if you want to understand the evolution of criminal antitrust enforcement in the US, this is not a bad place to start. The paper may focus a bit too much on US constitutional doctrines to be of great interest to readers outside the US, but the discussion of the trade-offs involved in criminalising certain types of courts extends to criminalisation efforts across the world.
The one point where I find it hard to share the author’s assumptions concerns the ease with which price-fixing conduct can be criminalised. International experience has demonstrated how difficult this actually is, and the author’s discussion of the (lack of) role of intent in US criminalisation efforts provides a window into why the application of criminal penalties to cartelists is more common in the US than elsewhere. These differences in approach across the world are apparent in the paper reviewed below, which focuses on the most feared individual sanction a European cartelist can suffer: being deported to the US.