Can Strict Criminal Liability for Responsible Corporate Officers be Justified by the Duty to Use Extraordinary Care?

Abstract

The responsible corporate officer (RCO) doctrine is, as a formal matter, an instance of strict criminal liability: the government need not prove the defendant’s mens rea in order to obtain a conviction, and the defendant may not escape conviction by proving lack of mens rea. Formal strict liability is sometimes consistent with retributive principles, especially when the strict liability pertains to the grading of an offense. But is strict liability consistent with retributive principles when it pertains, not to grading, but to whether the defendant has crossed the threshold from noncriminal to criminal conduct? In this essay, I review the two most plausible arguments supporting an affirmative answer in the context of the RCO doctrine. First, perhaps this doctrine reflects a rule-like form of negligence, akin to a rule that prohibits selling alcohol to a minor. Second, perhaps this doctrine expresses a duty to use extraordinary care to prevent a harm. Neither argument is persuasive. The first argument, although valid in some circumstances, fails to explain and justify the RCO doctrine. The second argument, a duty to use extraordinary care, is also inadequate. If “extraordinary care” simply means a flexibly applied negligence standard that considers the burdens and benefits of taking a precaution, it is problematic in premising criminal liability on ordinary negligence. If instead it refers to a higher duty or standard of care, it has many possible forms, such as requiring only a very slight deviation from a permissible or justifiable standard of conduct, placing a “thumb” on the scale of the Learned Hand test, identifying an epistemic standard more demanding than a reasonable person test, or recognizing a standard that is insensitive to individual capacities. However, some of these variations present a gratuitous or incoherent understanding of “negligence,” and none of them sufficiently explain and justify the RCO doctrine.

1 Introduction

1.1 The RCO Doctrine

Under the responsible corporate officer (RCO) doctrine, a corporate manager is criminally liable for unlawful acts committed by other corporate employees even if the manager did not personally commit an unlawful act and even if the manager was unaware of the violation, so long as the manager can be characterized as “standing in responsible relation to a public danger”¹ and so long as the manager “had, by reason of his position in the corporation, responsibility and authority either to prevent in the first instance, or promptly to correct, the violation complained of, and … failed to do so.”²

What is the scope of the RCO doctrine? Is it a doctrine of vicarious liability, or instead a doctrine imposing on the corporate officer an affirmative duty to prevent violations of the law by subordinates?³ I will assume, for purposes of this paper, the latter interpretation. The former interpretation, asserting that an individual, A, can be vicariously liable in criminal law for another individual, B’s, criminal conduct, is flatly inconsistent with retributive blame—unless A is himself personally culpable in some way, in which case the pure concept of vicarious liability is not the basis of criminal liability. Parents should not be vicariously liable for the crimes of their children, spouses should not be vicariously liable for each other’s crimes, and RCOs should not be vicariously liable for the crimes of those they supervise, unless, in each case, the “vicariously” liable party is personally culpable for the primary actor’s wrong.

We should distinguish the question of whether wrongdoers who are engaged in a common criminal enterprise are properly considered to be vicariously liable for each other’s crimes. In that context, it is at least arguable that, by joining such an enterprise, the wrongdoer has “changed his normative position” and is properly treated as at least somewhat culpable for foreseeable crimes that he facilitates.⁴ But this basis of criminal liability does not explain vicarious criminal liability in the RCO doctrine context. It is not wrongful conduct to have children,⁵ to marry, or to take on supervisory responsibility in a business.

The justifiability of the RCO doctrine is also distinct from the question of whether making the corporation criminally liable for the criminal acts of employees is defensible as a matter of policy or principle. The proper scope and limits of corporate criminal liability are important and difficult topics of their own. Rather, the question here is whether a corporate manager may be criminally punished because other corporate employees for whom he was responsible violated a criminal statute.

1.2 When Strict Criminal Liability May be Consistent with Retributive Desert

In prior work, I have suggested that strict criminal liability is sometimes consistent with principles of retributive desert. (For purposes of this paper, I assume that such desert is at least a necessary condition of just criminal punishment, even if consequentialist considerations may play a role in determining the nature and extent of punishment.) When strict liability pertains to an element of the crime that affects the grading or degree of punishment for the crime, as in the misdemeanor manslaughter rule, the felony murder rule, or rules that differentiate grand larceny from petty larceny, it is sometimes normatively justifiable.⁶ By contrast, when strict liability pertains to an element of the crime that affects whether the conduct is criminal at all, it is much more difficult to justify. Even in the latter context, however, when strict criminal liability concerns criminalization, it is occasionally defensible.

Strict liability with respect to criminalization is most plausible when the criminal law provision reflects a rule-like form of negligence. Consider two examples. First, suppose a state makes it a crime to sell alcohol to a person under the age of 21. And suppose the purpose is to ensure that those who purchase intoxicating beverages have the maturity not to create excessive risks to the health and safety of the drinker or those whom the drinker might injure. Further suppose that the law requires knowledge or recklessness as to whether the buyer is underage. Although this law is not formally strict, because it requires mens rea as to the “underage” element, it seems to be substantively strict, because the defendant will be convicted even if he reasonably believes that the person he sells to is mature enough to handle alcohol as well as most adults. Nevertheless, bright line rules of maturity are defensible, because they give fair notice to potential defendants and because, as applied over a predictable range of cases, they can be expected to better effectuate and express the underlying social concerns, reducing risks of harm caused by immature drinkers and imposing sanctions for those who create unjustifiable risks of harm. So it is not unjust to employ this rule in lieu of a vague standard of care, such as: “It is unlawful to sell alcohol to any person if the seller knows (or should know) that the buyer lacks the maturity to drink alcoholic beverages safely.”

Second, suppose a state makes it criminal to engage in any conduct that causes the death of another (or, more narrowly, to drive a car and thereby cause the death of another), in lieu of a more typical negligent homicide statute that criminalizes negligently causing death (or that criminalizes driving a car negligently and thereby causing death). In this case, it is indeed unjust to employ either of the first, rule-like formulations, because these formulations cannot be expected to better effectuate the underlying social policy of reducing, or imposing sanctions for, unjustifiably dangerous driving, relative to a standard that explicitly requires proof of negligence. The “causing death” rules are much more overbroad than the “selling to an underage person” rule, and much more likely to unjustly punish those who, through no fault of their own, cause death—such as surgeons who bring about death as a foreseeable side effect of a dangerous form of surgery, or drivers who cannot fairly be expected to avoid killing reckless pedestrians.

Another possible basis for strict criminal liability with respect to criminalization is that it invariably reflects the defendant’s breach of a duty of extraordinary care. I examine and critique this argument at length in Sect. 3 below.

2 The RCO Doctrine as a Rule-Like Form of Negligence

Can we interpret Dotterweich and Park as requiring negligence, but in rule-like form, as in the sale-to-minors example? On closer examination, this is an implausible interpretation. The decisions do not identify any specific rule-like criteria the satisfaction of which suffices for criminal liability. Indeed, Justice Frankfurter’s opinion for the Court in Dotterweich shows a cavalier indifference to whether the standard the Court enunciated requires genuine fault, and to whether it even provides coherent criteria for application in future cases: “To attempt a formula embracing the variety of conduct whereby persons may responsibly contribute in furthering a transaction forbidden by an Act of Congress … would be mischievous futility.”⁷

To be sure, in some imaginable cases, an application of the RCO doctrine could be justified on the basis that the RCO’s conduct violated a rule-like form of negligence. Suppose a corporate executive instructs employees at its egg production company not to wear protective clothing and not to clean and sanitize equipment, despite awareness that this greatly increases the risk of salmonella outbreak; or instructs its employees to conceal evidence that some eggs tested positive for salmonella; or orders that no more tests of its eggs be conducted, despite positive test results⁸; or falsifies records about which food safety measures were taken; or bribes a government inspector to release eggs for sale that had been tagged as failing to meet quality standards.⁹ Such facts are compelling evidence that the executive is at least grossly negligent, if not reckless or knowing, with respect to the actus reus of the crime, i.e., the fact that the eggs that are subsequently shipped in interstate commerce are contaminated.

However, the RCO doctrine has frequently been applied in much broader circumstances than this.¹⁰ For example, in Dotterweich itself, the Court upheld the conviction of a corporate president despite a lack of evidence that he knew that drugs had been misbranded or adulterated. And, in United States v. Starr,¹¹ the secretary-treasurer of a corporation had instructed a warehouse janitor to clean up a mice infestation; the janitor failed to do so, and may indeed have been trying to sabotage the company, but the court nevertheless upheld the conviction.

3 The RCO Doctrine as Based on Breach of a Duty to Use Extraordinary Care

Another possible justification for the RCO doctrine is that it properly expresses the duty of a corporate manager to use “extraordinary care.” This section highlights evidence that Dotterweich and Park might be so interpreted; explains that “extraordinary care” might mean either a flexible negligence standard or a higher duty or standard of care; reviews five possible interpretations of a higher duty of care; and notes that on any of these interpretations, the problem of insufficient culpability persists.

3.1 Evidence that the RCO Doctrine Expresses a Duty of Extraordinary Care

Park states that Congress has imposed on RCOs “the highest standard of foresight and vigilance.”¹² Norman Abrams’ careful analysis of Park ultimately concludes:

The most plausible interpretation is that the culpability standard thereby established is one of extraordinary care. Under this view, all that must be proved by the government is a deviation from that standard—something certainly less than common law negligence; it can be characterized as “very slight” or “slight” negligence to be distinguished from “ordinary” negligence.¹³

Similarly, the impossibility or “powerless to prevent the violation” defense recognized in Park and Dotterweich might support this “highest care” standard, for it suggests that, even if the precaution required to avoid the primary actor’s violation of the law is quite burdensome, liability will exist under the RCO doctrine unless the precaution could not “possibly” have made a difference.¹⁴

Moreover, the “responsible relationship” requirement superficially appears to require some genuine fault. Indeed, the Court in Park states:

The concept of a ‘responsible relationship’ to, or a ‘responsible share’ in, a violation of the Act indeed imports some measure of blameworthiness; but it is equally clear that the Government establishes a prima facie case when it introduces evidence sufficient to warrant a finding by the trier of the facts that the defendant had, by reason of his position in the corporation, responsibility and authority either to prevent in the first instance, or promptly to correct, the violation complained of, and that he failed to do so. The failure thus to fulfill the duty imposed by the interaction of the corporate agent’s authority and the statute furnishes a sufficient causal link. The considerations which prompted the imposition of this duty, and the scope of the duty, provide the measure of culpability.¹⁵

Yet the Court’s vague reference to a “measure of culpability” is disingenuous. For the government need not specifically prove the RCO’s negligence in order to secure a conviction under the RCO doctrine. And if the corporate officer were to provide evidence that she took reasonable steps to correct or prevent a violation, that evidence would presumably be inadmissible, unless the evidence is so compelling that it shows that she was “powerless” to prevent the violation.

In footnote 19, the Court further explains:

Assuming, arguendo, that it would be objectively impossible for a senior corporate agent to control fully day-to-day conditions in 874 retail outlets, it does not follow that such a corporate agent could not prevent or remedy promptly violations of elementary sanitary conditions in 16 regional warehouses.¹⁶

But this assertion, too, falls short of a requirement that the government must actually prove negligence by the RCO in failing to prevent or remedy violations of the law.

3.2 What Does “Extraordinary Care” Mean?

In several areas of the law, courts and legislatures have recognized a duty of extraordinary care or its equivalent, a test of slight negligence. In every area, however, they have confronted enormous difficulties explaining the test, especially in elucidating how the test differs from an ordinary negligence test. An ordinary negligence test is usually understood to require some type of balancing of the advantages and disadvantages of taking a precaution. Precisely how that balancing should be accomplished is a controversial matter, but it is widely accepted that, in paradigm negligence cases, the actor is required to make a reasonable tradeoff between the burden of a precaution and its benefits in risk reduction.¹⁷ Thus, even under an ordinary negligence test, more extensive precautions are required in order to avoid unusually great risks of harm. In this sense, an actor is often required to take “extraordinary care” even under the “ordinary care” standard. I shall call this sense of “extraordinary care” the “flexible negligence standard.”

But, alternatively, “extraordinary care” can refer to a higher standard of care. On this view, the actor must meet a more exacting standard than ordinary care. Perhaps the usual balance of the advantages and disadvantages of taking a precaution required under the flexible standard must be performed differently, or perhaps in some other respect a more exacting standard is demanded. The analogy here is to gross negligence, a widely recognized concept that requires proof that the actor was, in some sense, “more negligent” than an ordinarily negligent actor. If it is ordinary negligence to exceed the speed limit by 5 mph, then it is gross negligence to exceed the limit by 30 mph. Gross negligence requires a gross or substantial deviation from the standard of ordinary care. On this alternative view, then, “extraordinary care” is genuinely distinct from ordinary negligence because, as compared to such ordinary negligence, it requires a lesser deviation from the standard of ordinary care. I shall call this second sense of “extraordinary care” a “higher duty of care.”

If the justification for the RCO doctrine is that the conduct of the RCO violates a flexible negligence standard, that is an intelligible conception of negligence. The main difficulty with this justification is that ordinary negligence is arguably insufficiently culpable to warrant criminal liability.¹⁸ If instead the justification is based on the RCO failing to meet a higher standard of care, additional difficulties arise, as we will see in the next subsection. These include uncertainty about the meaning of a higher standard of care and the worry that this understanding of extraordinary care is either a gratuitous or an incoherent conception of negligence.

In tort cases, courts and legislatures have long recognized both conceptions of extraordinary care. However, the strong trend is toward rejecting the “higher duty (or standard) of care” conception and analyzing cases that were formerly treated under that conception as merely instances of the flexible negligence standard.

Consider first some examples of the first conception. In an early English case, Mackintosh v Mackintosh, the court explained:

[I]t must be observed that in all cases the amount of care which a prudent man will take must vary infinitely according to circumstances. No prudent man in carrying a lighted candle through a powder magazine would fail to take more care than if he was going through a damp cellar. The amount of care will be proportionate to the degree of risk run, and to the magnitude of the mischief that may be occasioned.¹⁹

Several jurisdictions have employed jury instructions for tort cases that clearly endorse this first conception.²⁰ Thus, Ohio’s former jury instruction stated: “The amount of care increases in proportion to the danger which reasonably should be foreseen. Ordinary care is a relative term. The test, though, is still ordinary care under the circumstances.”²¹ New York’s jury instruction provides: “Negligence requires both a reasonably foreseeable danger of injury to another and conduct that is unreasonable in proportion to that danger.”²²

On the other hand, in several contexts, tort law historically has endorsed the second conception. Thus, common carriers are often held to a higher than usual standard of care towards their passengers. For example, a California statute provides: “A carrier of persons for reward must use the utmost care and diligence for their safe carriage …”²³

Moreover, bailment law has long recognized a tripartite standard of care, requiring either slight, ordinary, or great care, depending on the nature of the bailment. “Great” care is equivalent to liability for “slight” negligence.²⁴ In maritime law, the rule until recently was that a plaintiff in a Jones Act case need only prove that the defendant was slightly negligent.²⁵ The traditional tort rule that a plaintiff’s contributory negligence is a complete bar to recovery was often qualified by the exception that a plaintiff could recover if the plaintiff’s negligence was only slight while the defendant’s negligence was gross in comparison.²⁶ Some jurisdictions require a “high” or “the highest” standard of care when the actor is engaged in an inherently dangerous activity, such as transmission of electricity.²⁷ Another legal context in which courts sometimes impose a liability standard more rigorous than ordinary negligence but short of strict liability is products liability.²⁸ Some jurisdictions require a higher degree of care, or impose liability for “slight” negligence, in other circumstances, such as high-risk amusement park rides.²⁹

Judicial explanations of the content of this higher duty are not very illuminating. For example, the Nebraska Supreme Court explained: “Any one of common sense knows that slight negligence actually means small or little negligence.”³⁰ In the next section, we will consider some other possible understandings of a higher duty of care.

3.3 Five Possible Interpretations of a Higher Duty of Care

Unfortunately, the “higher duty of care” conception of extraordinary care confronts almost insurmountable problems, either of irrelevance or incoherence, depending on how it is cashed out. I will examine five possible interpretations of that conception.

1. Consider first the view that “higher duty of care” refers to the extent to which the actor has deviated from a socially desirable or socially permissible standard of conduct. We can indeed make sense of gross negligence in this way, as an extreme or gross departure from the permissible standard of care. Driving 30 mph over the speed limit would be an example. Never inspecting company property for rodents or contaminated food, when law or custom require regular inspections, would be another. But how would we cash out the meaning of “higher duty of care” or “slight negligence” in such cases? Suppose it is ordinary negligence to exceed the speed limit (at least in typical circumstances). Then, is it slightly negligent to travel just below the speed limit? Yet if this conduct falls short of ordinary negligence, it seems that the conduct is not slightly negligent—rather, it is not negligent at all. Similarly, if ordinary care requires a company that distributes perishable food to inspect its facilities at least once a day, then a failure to inspect twice a day would seem to be, not slightly negligent, but not negligent at all.

Thus, the notion of a higher standard of care seems to be incoherent, since the standard of ordinary care itself identifies the borderline between unreasonable and reasonable conduct. Yes, it is conceptually possible to pick out a subcategory of reasonable conduct that is closer to that borderline, but one cannot intelligibly characterize that conduct as “slightly unreasonable” or “slightly negligent.”³¹

Of course, if a “higher standard of care” is just another way of expressing the flexible standard of negligence, then the higher standard is a gratuitous and irrelevant conception, because it is just another name for ordinary negligence, applied to those situations in which the need for precaution is especially great. If you are driving a celebrity through a crowd of delirious fans, you need to use much more caution than if you are driving on an apparently deserted road. If the RCO has been notified of rodent infestation in a particular warehouse, he has a duty to take additional precautions to reinspect that warehouse, precautions that can be much more extensive than would be required for warehouses in which there is no sign of such problems.

2. Second, suppose we cash out “higher duty of care” (or “slight negligence”) as a question of the degree to which the actor’s risky conduct (or omission) is unjustifiable. “Gross negligence” is extremely unjustifiable conduct, whose social costs are unusually high or whose social benefits are unusually low; while “ordinary negligence” is somewhat unjustifiable; and “slight negligence” is just barely unjustifiable. This interpretation does make some sense of the three categories, but it simply repeats the problems of the first interpretation. Even if conduct is just barely impermissible, a reasonable person should not engage in it.

3. Third, suppose we employ the Learned Hand conception of negligence to express these different degrees of negligence. Hand’s famous algebraic test provides that one is negligent if and only if the burden of taking a precaution (B) is less than the risks of harm that the precaution would prevent (which we can abbreviate as P × L).³² So if B is 100 while PL is 101, D is negligent; if PL is 99, D is not negligent. On this conception, once again, it is easy to understand ordinary negligence: B must be less than PL, even if it is less only by a peppercorn. And it is easy to understand gross negligence: B must be much less than PL. But again, it is difficult to make sense of slight negligence. If B > P × L, D is not slightly negligent; rather, he is not negligent at all.

To be sure, the Hand formula is controversial, and examples such as the one just given can provoke indignant reactions. Can the permissibility of risky conduct really depend on whether B exceeds or falls short of P × L by a peppercorn? Is such a cold cost–benefit calculus a morally acceptable way to distinguish reasonable from unreasonable conduct? There are a number of important questions here, questions that it is not the task of this paper to explore.³³ What is crucial to understand, however, is that any criterion of negligence that requires trading off one value or interest for another will create many of the same problems. The criterion might be the Hand test, understood to express economic efficiency; or the Hand test, understood more broadly; or a test that balances the values of freedom of action against the value of physical security; or a “reasonably prudent person” test in which these issues are less visible (because of the opacity and vagueness of the test).³⁴ But any such criterion will have close cases, such that adding just a bit to the burden or benefit side of the balance will tilt the balance in the opposite direction.³⁵ And, again, it is difficult to make sense of a higher standard of care or its equivalent, slight negligence, if balancing of this sort is the proper mode of judging negligent conduct.

Difficult, but not impossible. As I will explain in the next subsection, there is indeed a coherent version of a higher standard of care that departs from simple Hand formula balancing. However, this approach is still problematic, as we shall see.

4. Fourth, we might understand a higher standard of care as an epistemic standard, as requiring actors to take precautions that an omniscient observer would take, or that a person would take if she had unlimited time and resources to investigate the relevant facts.³⁶ We can often separate the question of what counts as an unreasonable or unjustifiable risk, or an insufficient precaution, from the question of what counts as unreasonable failure to realize that one has taken an unreasonable risk or insufficient precaution—just as we can separate the question of what counts as reasonable force in self-defense (e.g., what degree of force is proportionate in light of the threat), from the question of what counts as unreasonable failure to recognize the facts relevant to the first question.

This fourth, epistemic interpretation is promising, for it allows us to employ the flexible standard of care with respect to the first element of the analysis, whether it was unreasonable not to take a precaution, but to employ a higher standard of care with respect to the second element, whether the actor knew or should have known the facts that are relevant to whether the first element was satisfied. Moreover, criminal law culpability often places great weight on the existence or extent of the actor’s awareness of the legally salient facts.

Nevertheless, the problems with other interpretations of a higher standard of care do not disappear simply because we focus on beliefs about burdens and risks, rather than on the burdens and risks themselves. For the extent to which a reasonable actor should investigate the facts depends on what the facts might reveal. The secondary duty to investigate demands a higher level of care if the primary duty not to cause unjustifiable harm involves very serious harms, such as death, rather than more modest ones, such as minor property damage. And the same problems that a “higher duty of care” analysis poses in the context of the primary duty also arise with the secondary duty. The extent to which an RCO should conduct investigations of unsafe or dangerous conditions in a plant depends on the seriousness of the condition and on the burden of conducting such investigations (including the social costs of expensive investigations, which raise the cost of the company’s services or products). If a reasonable RCO would investigate with frequency Y or with intensity Z, then the RCO’s failure to investigate with greater frequency 3Y or with greater intensity 3Z is not “slight” negligence; it is not negligent at all.

5. Fifth and finally, we might interpret a “higher standard of care” as insensitive to individual capacities, even more insensitive than are the standard ordinary or reasonable person standards. The reasonable person test in criminal as well as tort law makes allowance for physical but not mental or intellectual disabilities (unless the mental disability reaches the extreme of insanity). The law asks what a reasonable blind person would do, but not what a reasonable stupid, inattentive, or clumsy person would do. The language of some “higher duty” jury instructions captures this idea by framing the negligence inquiry as “the highest possible degree of skill” or as “exceptional skill.” (This fifth interpretation overlaps with the fourth, insofar as cognitive failures stem from incapacity.)

But the RCO doctrine will only rarely instantiate this fifth interpretation, because the acts and omissions that the doctrine criminalizes will normally be executive decisions about precautionary protocols, inspections, and the like that the actor is perfectly capable of making. The actor’s complaint, rather, is that the precautions that could prevent the criminal violation in question are unrealistic or unduly burdensome. Thus, the doctrine is unlikely to apply to a case where rodent infestation could be prevented only by a person with superhuman vision who could detect rodents underground or inside walls. But it might apply in a case where the RCO hires an insufficient number of inspectors to check on dangerous or unhealthy conditions at a factory, even if the RCO had no reason to know, in advance, that the number was insufficient. Similarly, even if the RCO personally lacks the skill to identify signs of rodent infestation from droppings or other evidence, the RCO certainly has the capacity to hire individuals who do have that expertise.

3.4 Is “Thumb on the Scale” Balancing a Solution?

It is worth considering another approach to negligence and the balancing of values that offers the promise of making a higher standard of care intelligible and attractive. This approach is what I call the “thumb on the scale” version of a balancing formula.³⁷ On this version, an actor does not escape liability simply because B is slightly greater than P × L. Rather, a thumb is placed on the scale, so that B is negligent if but only if B < n × P × L, where n is greater than 1. For a heavier thumb, simply increase n.

For example, suppose the weight of the thumb is 2, so that D is negligent if and only if B < 2 × P × L. Compare this with the ordinary Learned Hand test:

Under the “thumb” approach, the result is more favorable to the plaintiff than under the Learned Hand test, in the two shaded rows. (Notice that, even if the risks and benefits are largely commensurable, as in cases where PL refers only to risks of damage to property, the approach is distinct from a simple Hand approach.)

This conception of negligence is quite defensible. Why should an actor always be permitted to impose the risks of his activities on others, simply because the actor’s burden to avoid the risk is slightly greater than the risks? To be sure, the actor might well have chosen the least-cost, Learned Hand test option, based on a simple comparison of B and PL, if the actor were the only person to suffer the risk as well as incur the burden of preventing the risk from being realized. But it does not follow that the actor should be free to dump those risks on others. The “thumb” approach is especially attractive when the individuals who suffer the risks do not benefit much, or at all, from the activity³⁸—for example, when they are especially likely to suffer environmental harms but benefit very little from the commercial activities that lead to those harms. The approach is less attractive when those who are placed at risk benefit more significantly, as when the victims are employees or customers of the company producing the hazard.

On first blush, it might seem that the argument for a thumb on the scale is a strict liability argument in negligence clothing. But this is not so. The argument is that the primary actor should have acted differently, and that is unquestionably a negligence argument, not a strict liability argument. The “thumb” version of the Hand test is a negligence standard in a straightforward sense: the primary actor should not have imposed the risk, even though B was slightly greater than PL. (However, if B had been substantially greater than PL, i.e., if B had been greater than n × P × L, then the actor would not have been negligent in this sense.)

To be sure, it is intelligible to speak of strict liability “duties.” But these are not duties owed by primary actors to act otherwise, to take a precaution that would have avoided a harm. Rather, insofar as they are properly characterized as “duties,” they are obligations to compensate victims of the activity, no matter how reasonably or prudently the activity is conducted.³⁹ Tort law recognizes such duties in a number of categories, including abnormally dangerous activities, product liability for manufacturing flaws, and liability for harms caused by wild animals.

The “thumb on the scale” approach undoubtedly provides more protection for potential victims of risky activities than would a simple Hand test. And it is a coherent conception of a “higher” duty of care, since it is defined by contrast with the (no-thumb) simple Hand test, or by contrast with the precaution that a person would (or prudentially should) take if the person were to internalize all the advantages and disadvantages of taking the precaution. However, the approach is much better suited to tort liability than to criminal liability. Specifically, it is unclear whether the approach can solve the overbreadth problem with criminal law’s RCO doctrine. If the fairness considerations that underlie the thumb approach justify only a thumb of no more than 2 or 3, then the approach might indeed limit the RCO doctrine, which on its face justifies criminal liability even if n is 5, or 10, or 1000. Thus, the approach might forbid requiring an actor to undertake enormously burdensome precautions, such as personally following up on an hourly basis with every employee who has any responsibility for complying with legal requirements.⁴⁰ On the other hand, this approach does not map very well onto the RCO doctrine, because that doctrine is not limited in its application to cases where the victims of the activity do not benefit from the activity.

3.5 The Problem of Insufficient Culpability

Even if these problems could be surmounted, the other enormous hurdle that an “extraordinary care/slight negligence” approach must clear is that such conduct is almost always insufficiently blameworthy to deserve criminal punishment. Even ordinary negligence may be insufficiently blameworthy (depending on how that concept is further specified). The approach of the Model Penal Code requiring “gross” negligence as the minimal degree of fault before the state can justifiably punish a defendant is, in my view, largely persuasive. However, defending this position is beyond the scope of this paper.⁴¹

To be sure, some forms of conduct and some omissions by RCOs could satisfy a gross negligence standard or a similarly demanding culpability standard. Moreover, some instances of negligent or grossly negligent RCO behavior might be an awkward fit with a typical culpability standard requiring that the actor knew about, or was reckless concerning, the specific acts or omissions by employees that violated criminal prohibitions. To that extent, a very narrow version of the RCO doctrine might be both defensible and useful. For example, suppose a corporate officer deliberately sets up an inadequate inspection system, in order to save costs or obtain a competitive advantage over a rival, knowing that over time the system will permit a significant increase in the number of incidents of violation of food, safety, or environmental laws. Suppose he also directs that employees conceal any such violations from supervisors and from the government. The officer might honestly be unaware of any particular violation, either at the time it occurs or thereafter. In this situation, a criminal law requirement that he be aware of specific violations is insufficiently demanding. The same might even be true of a requirement that he be reckless, i.e., aware of a substantial risk of specific violations.

If the RCO doctrine is limited in this fashion, or in a similarly drastic way, it could be a valuable supplement to other criminal law culpability standards. For example, we might limit the doctrine to where aggravating factors are present—e.g., the RCO conceals evidence, or orders that evidence of violations not be reported, or lies to government investigators or auditors. But this approach would be dramatically different from the way that many courts currently interpret the RCO doctrine.

The only precaution that can guarantee that the actor will not cause the harm specified by a strict liability statute is to avoid entirely the activity that might, despite the exercise of extraordinary care, cause the harm. If the statute prohibits the distribution of adulterated or misbranded food or drugs, the actor can avoid engaging in that line of business. If the statute prohibits driving that causes a death, the actor can avoid driving. But it is absurd to characterize failure to take such a precaution as necessarily negligent, even in the sense of failing to use extraordinary care. After all, it is socially desirable that individuals engage in productive and mutually beneficial activities even though every conceivable activity creates some risk of causing harm, and even though the widening ambit of the criminal law increasingly subjects more and more activities to regulation and potential punishment.⁴² And it is absurd to deny an “impossibility” defense (if that is the only protection from unjust punishment) on the ground that the actor had a viable “possibility” open to her—namely, not engaging in the activity. On this view, no statutory violation is impossible, because an actor could simply decline to act in any way that could conceivably come within the statute’s ambit. The aspiration of the criminal law is not, and should not be, to create a nation of inactive recluses.

4 Conclusion

The RCO doctrine is very difficult to justify if one believes that retributive desert is at least a necessary condition of criminal punishment. This paper has investigated two possible arguments in support of the doctrine. First, strict criminal liability is sometimes justifiable as a rule-like form of negligence. However, this argument fails to explain and justify the RCO doctrine.

Second, perhaps a duty to use extraordinary care to prevent harm underlies and vindicates the RCO doctrine. On closer examination, this argument also fails. If “extraordinary care” simply refers to a flexible negligence standard that considers the burdens and benefits of taking a precaution, it is problematic in premising criminal liability on ordinary negligence. If instead it refers to a “higher duty of care,” it has five possible forms. It might: (1) require a lesser degree of deviation from a permissible standard of conduct than an ordinary care standard requires; (2) identify conduct that is just barely unjustifiable; (3) place a “thumb” on the scale of the Learned Hand test in favor of victims; (4) identify an epistemic standard that is more demanding than the reasonable person test; or (5) recognize a standard that is insensitive to individual capacities to take care. However, some of these variations present a gratuitous or incoherent understanding of negligence, and none of them sufficiently explain and justify the RCO doctrine.