1 Introduction

Respondeat superior (‘let the master answer’) is a legal doctrine conferring liability from one party onto another because the latter stands in some relationship of authority over the former. For example, if my child breaks a vase in your house while we are over for dinner, I seem appropriately on the hook for the breaking of the vase. I am the one who is ultimately responsible. Not just any relation of authority will do. And the doctrine will not cover just anything done by that agent under one’s authority. What the doctrine is meant to capture, though, is the culpability of someone in authority as expressed or demonstrated through the culpable behavior of someone in their remit.

Applications of respondeat superior license vicarious liability, a form of strict liability where the agent is held liable for the actions of another and without requiring proof of fault by the principal. Traditionally, this was a doctrine within tort law, where one can find it in a number of contexts. Since 1909, however, respondeat superior has also been applied within the criminal law, specifically in the context of corporate criminal liability.Footnote 1 Prosecutors use respondeat superior to hold employers and firms criminally responsible for what is done by their employees, specifically what is done within the scope of their employment and with an intent to benefit the firm. And this has become the dominant way of holding corporations criminally liable.

It’s not hard to understand its popularity in this context. We want to hold corporations criminally responsible.Footnote 2 But many forms of criminal responsibility would require proof of mens rea concerning the crime committed, and it is controversial how we could demonstrate mens rea for a firm or if firms themselves are even capable of instantiating a culpable state of mind.Footnote 3 And so, this is a way to hold firms responsible without having to take anything for granted about their mentality. We can infer their culpability straightaway on the basis of the demonstrable crimes of the employees.

Despite this appeal, many have rallied against using respondeat superior in particular to secure corporate criminal liability, and it has been challenged from the start. Authors have long suggested that the doctrine is overinclusive, because it would hold firms responsible for the conduct of rogue employees (e.g., Bharara 2007). Others have brought out how it may be underinclusive, because it will not clearly implicate the firm when it is not clear which individual employee satisfies the elements of the crime (e.g., Vu 2004). Or it may be both (e.g., Abril & Olazabal 2006; Diamantis 2016; Thomas 2021). Even apart from these challenges, we may worry that it punishes too harshly (Weissmann 2007) or that it has focused on satisfying aims inappropriate for the criminal law (Luskin 2020).

Apart from these concerns, I want to focus on whether respondeat superior really is appropriate to appeal to it as a doctrine in the criminal law, in particular on a picture of criminal liability that requires desert. The doctrine may be argued as being a powerful motivator for firms to police their members, providing a clear deterrence-based justification for its use. However, I want to put that argument to one side, instead focusing here on the question of whether firms or employers can deserve to be held liable on the basis of the criminal conduct of their employees. So, I am assuming that appealing to the doctrine is only appropriate if it genuinely defeasibly licenses judgments of criminal liability given a theory of when agents are criminally responsible.

Given this, in section (I), I present the view offered by Antony Duff on which criminal responsibility is understood in terms of moral responsibility, in particular in terms of the concept of answerability. Then, using this framework in section (II), I take up the question of the article: Is respondeat superior justified? I argue that it is not, because we cannot infer from the presence of a principal/agent relation that the principle is answerable for the conduct of the agent. Firms typically lack the authority to offer the reasons why employees act as they do, and this undermines the ambition of calling them to answer for this conduct in the form of criminal proceedings.

In section (III), I consider and respond to one significant challenge to using answerability to cash out criminal liability. Then, in section (IV), I consider a number of challenges to the view of why respondeat superior is unjustified. Perhaps we are being misled by focusing on firms, or answerability does not require the ability in a case to offer reasons why the action was done, or perhaps vicarious liability generally does not require answerability. Responding to these challenges leaves us with a better understanding of why respondeat superior is unjustified, and, luckily, I think we don’t need it. In section (V), I conclude by highlighting ideas for how best to hold firms criminally responsible without appealing to respondeat superior.

2 The Basis of Criminal Responsibility

We talk in terms of responsibility across both the moral and legal domains. So, one would hope that there is a coherent notion of responsibility at work between them, or that we could understand the relation between them or the differences between them in terms of the differences between these domains.Footnote 4 At the very least, there is clear overlap—we are criminally responsible for many things for which we take ourselves to be morally responsible, and we may think there is a moral obligation (at least prima facie) to obey the law.

One particularly strong reason to draw together moral and criminal responsibility comes down to our understanding of crime. On a popular way of conceiving of what a crime is (or what should be a crime), crimes themselves are a certain kind of moral infraction. Specifically, they are public wrongs. We may say that public wrongs are a matter of wrongs done to the public itself, or that they are wrongs that are of special concern to the public (Duff 2014; Marshall & Duff 1998; Duff & Marshall 2019; Kennedy 2021), or that they are only those wrongs of public concern that the state ought to punish (Lee 2015). Regardless of precisely how we understand the view, it leads us to draw an explicit connection between moral and criminal responsibility. Crimes just are a kind of wronging, and so responsibility for crimes reduces to moral responsibility for a kind of wrong done.

This alone suggests quite a lot about the justification of respondeat superior. Insofar as it is only appropriate to hold someone morally responsible insofar as the sanctions they are opened up to are deserved for their wrongdoing, tying criminal responsibility to moral responsibility will only justify holding someone criminally responsible if the punishment is deserved given the crime committed. This already runs afoul of more instrumental conceptions of the criminal law, but it opens a path towards judging respondeat superior, as we can ask after whether the master really does deserve the punishment for the wrongdoing of those under their charge.

Saying only this much leaves us still a bit lost concerning how to pursue our question. However, given how much has been written on what it is to be morally responsible, there is the promise of leveraging a prominent account to consider whether employers satisfy it when employees commit crimes. Duff (2007, 2009, 2018, 2019) himself has advocated in a number of places for understanding criminal responsibility in terms of what those writing on moral responsibility have characterized as answerability.

Critical to what it is to be responsible in the sense of answerability is that it is appropriate to demand of an agent answerable for some deed why it was done. Answerable agents can be called to answer for their conduct, where the expectation is that they can offer an explanation for why they performed the particular action in terms of the reasons available to them. More specifically, we may think that a contrastive element is required, where agents can be expected to answer why the action was done rather than another, or why the agent acted in accordance with some of their reasons and not others (Shoemaker 2015:ch.2).

This is a powerful understanding of moral responsibility,Footnote 5 as we may take our practices of moral responsibility to almost entirely consist in holding ourselves to be answerable for our conduct and soliciting others to answer for what they have done when necessary. And Duff maintains that this sense of responsibility is especially well-suited for understanding criminal responsibility. In particular, he sees this process of answerability as played out in the law through criminal proceedings. The prosecution sets forth the case for why the agent satisfies the criteria for having committed the crime. Then, the defendant has the opportunity to argue that they were either justified or should be excused for what they did. The defendant is called to answer for what has been done, and they respond by either admitting guilt or providing a sufficient explanation. This captures how the criminal law centrally involves our calling wrongdoers to account (Duff 2010).Footnote 6

Once we have this conception of culpability in the criminal law laid out, it becomes easier to assess whether respondeat superior is a valid doctrine. So, it is to this issue that I now turn. We use respondeat superior to take us from the responsibility for wrongdoing of an employee to the responsibility of the employer. Given that we have tied criminal responsibility for some conduct generally to a conception of responsibility as answerability, whether respondeat superior is a valid doctrine rests on whether employers generally are answerable for what employees do. This would allow us to infer corporate responsibility from employee responsibility. So, we can assess the validity of respondeat superior by assessing whether it really is true that employers are answerable when employees are.

3 Is Respondeat Superior Justified?

As we have seen, for respondeat superior to be a valid doctrine within the criminal law, it must be appropriate to ascribe criminal responsibility to an authoritative party on the basis of the criminal responsibility of the individual(s) under their authority. So, understanding criminal responsibility in terms of answerability, the question becomes: Are authoritative individuals like firms answerable when their employees engage in criminal wrongdoing? And here, I think the answer here is ‘no.’Footnote 7

Consider a case in which an employee blackmails a government agent, threatening to expose the agent’s affair unless he awards her firm a coveted government contract. Suppose it is clear that the employee herself is responsible for her crime. This is an action she performed for various reasons, and we can easily expect her to be in a position to answer why she chose to blackmail rather than bid fairly on the contract. Even if, for whatever reason, she in fact could see no other way forward but to blackmail the government agent, it is still appropriate to hold her criminally responsible. (Claiming that the action was necessary would constitute her answer as a potential justification.) She has put herself up as answerable for her conduct as an employee generally, and she is charged with navigating the firm’s business within the bounds of the law.

In this case, respondeat superior would have us hold the firm criminally responsible as well. The employee committed the crime in the scope of her employment and arguably with the intent to benefit the firm. However, I think respondeat superior gets the wrong verdict in this case, because I do not think that the firm is answerable for the employee’s crime.

Consider whether the firm is able to offer the needed reasons demanded by answerability for this crime. In the case, the employee has reasons for why she engaged in blackmail. And she may be aware of the reasons for her to not have blackmailed the government agent. Even if we are ready to have this employee’s conduct redound onto the firm, it is hard to see how the firm could answer for the crime in terms of her reasons. The flat-footed way to put the point would be to say that the firm has its own reasons; it can’t appropriate the employee’s reasons for action. Her reasons for acting are particular to her own normative standpoint, and this is a standpoint not shared by the firm.

As distinct group agents, firms embody their own normative perspective, their own rational point of view of what is to be done in light of their reasons to act (Rovane 1998; Hess 2010). In other words, firms have their own reasons for doing things. It is a challenge to say precisely what those reasons are, where they come from, how those reasons are recognized and acted on. But there is work to try to meet this challenge (Silver 2022), and those convinced that firms are agents with distinct values can be sure that the reasons firms negotiate are distinct from the set of reasons that bear on the conduct of individual employees.Footnote 8 This is a significant view to have to take on board, and we will discuss it more in section (IV). However, this is an assumption that we must make if we want to say that corporations can genuinely deserve punishment and be distinctly criminally responsible. (If we deny that firms are distinct agents acting from their own vantage point and with their own reasons, then they will not possibly be among the kinds of things that can deserve punishment.)

Taking this on board, then, we can also recognize that firms will not only have their own reasons, but they have particular ways of speaking to their reasons. Pettit (2017) discusses how firms can and often do have spokespeople that offer the firm’s reasons for action. Pettit accepts that firms can be called to justify what they do, and he accepts that they should do so in terms of their reasons. What is special about spokespeople, for Pettit, is that they are empowered within the firm to offer these reasons. It’s not that spokespeople are in some special epistemic position to recognize the reasons why the firm behaves as it does. Rather, spokespeople have been given the authority to commit the firm to the particular justifications offered.

Though the firm itself may have authority over certain activities of the employees, spokespeople cannot generally answer why some employee commits some crime. Since the firm is distinct from the employee, spokespeople lack the authority to determine this fact,Footnote 9 even if they are in a good position evidentially to suggest a likely reason. This point is critical. The claim is not that the firm is necessarily in a bad epistemic position to be reasonably certain as to why the employee’s act was committed. The firm may well be sufficiently positioned to recognize the objective features of the employee’s environment; it may even be responsible for engineering those incentives. So, we can take for granted that the firm is able to say what considerations should have borne on the commission of the crime in question. Nevertheless, the firm cannot settle why the employee’s action was performed, that it was done for one reason rather than another. The firm lacks the authority to answer why the crime was committed, and so it cannot be answerable.

If the firm is not answerable, then the firm is not criminally responsible in this instance, and respondeat superior as a doctrine will have been shown to be invalid. But we might wonder about specifically what is going wrong here? Is the claim that the elements of the crime are not properly attributable to the firm after all, or is it something else that explains the firm’s inculpability? Answering this question is actually a bit tricky, and it depends on our precise understanding of the actus reus and mens rea criteria, as well as how we understand the difference and relevance between intent and motive.

One natural way to try to establish this result (that firms are not answerable) would be to fall back on the idea that the employee’s conduct here isn’t properly attributable to the firm. If the employee’s conduct is not attributable to the firm, then the firm cannot satisfy the actus reus element of the crime. The firm didn’t do anything for which it may be answerable, and so the firm is not criminally responsible. In some cases, I think this response is not out of line. For cases where the employee has been expressly banned from engaging in this illegal conduct and where the employee goes out of her way to hide this conduct, it does seem like the firm has done all it can to disassociate themselves from and disavow the conduct.

Worried about just this sort of uncontrollable, low-level employee conduct, some have suggested refining respondeat superior as a principle to only concern those higher-up officials in the firm capable of directly expressing the firm’s values with their conduct, and so which cannot as easily be disavowed.Footnote 10 This is an attempt to pin liability on the firm only when the conduct is truly attributable to the firm.Footnote 11 Though a reasonable suggestion, it can prove hard to demonstrate exactly which employees count as relevant in this context. But the problems with this suggestion go much deeper.

First, this seems perhaps connected to the thought that what the firm does just is what certain top employees do. But this idea itself faces problems,Footnote 12 chief among them is that respondeat superior is still a doctrine of vicarious liability. We cannot re-work the doctrine so that we only find the firm liable for the firm’s actions.

The bigger problem with this approach, though, is that while we may be happy to screen-off the conduct of certain extremely uncompliant employees, we generally do want to say that much of what employees do for the firm is attributable to the firm. The firm’s hiring the employee is in part causally responsible for what crime the employee commits in virtue of their employment. Further, we want to say that employers and employees stand in a principal-agent relation, where employees act as the agents of employers. On this model, it does seem as if the actions of employees are meant to stand for the actions of the firm; employees act on behalf of the firm. So, I am willing to grant this idea that the conduct of an employee can often be attributed to the firm.

There may, however, still be a way of denying that the firm genuinely satisfies the actus reus component of the crime. While the behavior of the employee may be attributable to the firm, it is less clear that the behavior is attributable qua action. What is often taken to distinguish an action from a mere bit of behavior (e.g., my raising my arm from my arm’s rising) is that actions are up for being explained in terms of reasons, that they are appropriately met with the question of why they were performed (Anscombe 1957). Having already argued that we cannot appropriately ask the firm why certain crimes of the employees were committed, this suggests that the act of the employee cannot be attributed to the firm after all, even if her behavior can be.

This is likely the best way to deny that firms are criminally responsible in these cases. However, other ideas are worth considering. We may instead deny that the firm genuinely satisfies the mens rea element of the crime. After all, if the firm is not in a position to answer authoritatively as to why the crime was committed, then we may think that the firm must lack the right kind of knowledge state. Despite having some initial appeal, I think that this is not a profitable line to pursue. For one thing, those sympathetic to corporate criminal responsibility and corporate agency are often happy to aggregate the mental states of employees and attribute them collectively to the firm (e.g., Sarch 2019). If the employee knew what she was doing as she blackmailed the government agent, then we may want to say that the firm also counts as knowing this in some sense, and so the firm will satisfy the knowledge element of the crime.

The highest level of mens rea involves not only knowledge but purpose/intent, and we have already stipulated as necessary for respondeat superior that the agent of the firm performed the action with the intent to benefit the firm. So, we might maintain that the firm cannot be shown to have this desiderative element, or that it is not something that the firm itself is in a position to speak to. Nevertheless, I am not convinced that the firm could not be shown to inherit these purposive states as well in some sense. After all, the firm may have engineered the situation to give the employee the incentive to have this intention, which seems to suggest that the intention is appropriately attributed to the firm.

Moreover, the thing that I am claiming firms are not in a position to do—to authoritatively speak to the motivating reason that led to the crime—is likely not necessary for satisfying mens rea. When one is providing a motivating reason for one’s action, what is more clearly at issue is the motive with which one acts. And it is widely maintained that motive is irrelevant, at least to the establishment of mens rea. Appreciating this, however, might point us towards a different way of denying corporate culpability. If the firm can satisfy the actus reus and mens rea elements of the crime, and so the crime can arguably be attributed to the firm, we could still treat their inability to answer as to why the crime was performed as providing an exemption. Whereas motive may be irrelevant to mens rea, some have argued that motive is relevant for establishing a defense for the crime committed (e.g., Binder 2002).Footnote 13

The defendant is called to answer for the crime committed, and the defense may consist in outlining why the crime was necessary to avoid some other wrong, and so motivated. Despite technically satisfying the mens rea element of the crime, if the firm cannot articulate a motive for the crime, then the firm is not in a position to offer a defense. That is just to say what I have been saying, that the firm is not in a position to answer for the crime, and so must be exempt from this prosecution that constitutes a call to answer.

The role of motive in the criminal trial is controversial, and we need not settle the matter here. We also need not settle whether the criminal act of the employee ultimately cannot be attributed to the firm, or whether it can be, but the firm is to be exempted for its inability to answer for the crime. Either way, this achieves the result of allowing us to deny that corporations are criminally responsible via respondeat superior, though these would have been the cases in which we took the doctrine to paradigmatically apply. We are thus unjustified in appealing to respondeat superior as a doctrine.

4 Is Answerability Required for Liability? The Case of Strict Liability

In making this particular case against respondeat superior, we needed a fair number of assumptions about the nature of criminal liability. For example, we took for granted a view about crime itself, that it is a matter of public wronging. That view is controversial,Footnote 14 though luckily our project does not require it.Footnote 15 Even if crimes were not understood in terms of public wrongs (or moral wrongs of any kind), it could still make sense to understand criminal responsibility in terms of answerability, as answerability as a form of responsibility is found in moral and non-moral contexts.Footnote 16

A more significant assumption to investigate, however, is the claim that criminal responsibility requires answerability. If it does not, then a firm’s failing to be answerable will not prove that they are not criminally responsible in cases where respondeat superior is applicable. So, the argument of the last section needs liability to require answerability. Unfortunately, Shoemaker (2013: 161–165) has challenged exactly this part of Duff’s theory. He points to cases where we may really want to say that individuals are criminally responsible, but where they do not seem morally responsible, and that this is borne out by the fact that these individuals would not be answerable as answerability is often understood. This suggests answerability is critical for moral responsibility, but not criminal responsibility.

As an example, consider a case where a shopkeeper inadvertently sells some tainted meat. The shopkeeper can still be criminally liable for selling tainted meat as a matter of strict liability,Footnote 17 but we may not think that the shopkeeper is properly answerable for the crime. Now, one response to this case is to bite the bullet and suggest that in fact the shopkeeper should not be held criminally responsible because she is not answerable. That is, we could simply deny a regime of strict liability. Readers happy to take this route are invited to move on to the next section worry free. However, even Duff himself does not want to accept this result. If we want to show that answerability is genuinely necessary for criminal culpability, then we have to offer a story for how strictly liable agents are answerable after all.

Duff (2007: 257) suggests that the shopkeeper is answerable and could answer sufficiently for this crime by bringing to bear evidence of due diligence, but Shoemaker is unconvinced by this suggestion. As we saw, being answerable involves being able to answer in terms of one’s reasons, specifically the reasons why one performed the act rather than another. And this is something that Shoemaker thinks the shopkeeper cannot do. Duff needs to say that the shopkeeper is answerable for the crime as a matter of strict liability, but how can the shopkeeper be answerable in this sense when she didn’t even know that the meat was tainted? She cannot answer the question necessary for answerability: “Why did you Φ in light of the reason(s) not to Φ?” ([emphasis in original], Shoemaker op. cit.: 164). As Shoemaker goes on to explain,

…even if I can give you an answer for why I sold meat voluntarily, I cannot give you an answer for why I sold meat voluntarily in light of the reason not to do so (that is, it was tainted), because from my position at the time I had access to no such reason; that reason certain shone no light on me. But in saying that I had no access to reasons not to sell the meat, I have only “answered” the question by saying that there just is no answer of the sort being sought (ibid.: 164–5).

Critical for why an agent may not be answerable for some deed is if the reasons for why not to do it are not accessible to her. The claim is thus that in cases of inadvertent behavior like the shopkeeper’s, she does not properly have access to the reason for why not to perform the action she does, and so she cannot offer those reasons when prompted.

Recently, Duff (2019) has responded to these concerns of Shoemaker. There, he seems to agree that these reasons for why not to perform the action in question are important, but he disagrees over their being inaccessible to the agent. He says,

…there is reason not to act in that way—to sell this food will be to sell unsafe food…; and that reason is accessible to me, in the sense that I can understand it with hindsight, and could have understood it at the time…So I would say, as against Shoemaker, that I had reason not to sell the food…; and that I can now be called to answer for my action of selling this unfit food…My answer might involve giving my reasons for the intentional action that turned out to be an action of selling unfit food…it might also involve explaining why I did not take further precautions that might have revealed the reason not to act thus…(5–6)

As we see, Duff’s response involves accepting that answerability does involve being the appropriate target of the question about why you acted as you did in light of the reasons not to. What he denies is that the reasons not to in cases of non-culpable inadvertence are inaccessible to us. There were reasons for the shopkeeper not to sell the meat. And the shopkeeper is apt to understand them and will recognize them in hindsight. This is what he takes to secure the answerability of the shopkeeper. She is answerable even if the answers given will involve not why the shopkeeper sold the meat despite the clear reason not to, but rather why the shopkeeper sold the meat and why she engaged in other conduct that did not reveal any reason not to.

Frankly, it can be hard to see how this is a successful response to Shoemaker’s objection. Yes, there was a reason for the shopkeeper to not sell the meat. But as the shopkeeper went to sell the meat, it is surely a stretch to say that this reason was accessible to her, just because it would have revealed itself upon intensive investigations.Footnote 18 Stretch or no, the real question is: Does this reason count as accessible in the sense relevant for answerability if the agent merely in principle could have discovered it, or does the reason need to be epistemically accessible to the agent as she decided to act for her to be answerable for that action? In answer to this question, I think that both authors are right in a certain way.

Shoemaker is right that the reasons relevant for answerability are the reasons that the agent in fact acted in light of. If the reason not to sell the meat only existed in an objective sense and was not apparent to the agent, then she is not able to answer as to why she did not heed it. That she cannot answer, however, is not to say that she should not be held as answerable. It may be that she cannot answer for this or that action; however, she may nevertheless be held to a standard on which it is appropriate to demand an answer of her, even if none is forthcoming.

We can respond on Duff’s behalf by noticing that the agent is still the appropriate target of the demands of answerability. As a shopkeeper legally permitted to sell meat, the agent assumes responsibility for making sure that the meat is not tainted.Footnote 19 And this involves being willing to take responsibility should anything go wrong.Footnote 20 Whether or not the shopkeeper has reasons to give, she takes upon herself the duty to search out reasons of the relevant kind and offer them when asked. If she has no answer, as in the case of inadvertence, then her reply will not be a matter of undermining the appropriateness of the question, as Shoemaker suggests. Instead, what she says will be an explanation of why there is no answer to offer, which amounts to presenting an excuse for the wrongdoing, as Duff suggests.

It is a much larger conversation to nail down the right way to conceive of answerability and to puzzle out apparent differences in how it is used to explicate moral/criminal responsibility. Hopefully, our purposes do not require seeing through such a task. I have conveyed a plausible and prominent way of cashing out criminal responsibility in terms of answerability, and I have defended this conception against the charge that it cannot work specifically for cases of strict liability. Taking this for granted, then, I want to move on to consider challenges to how I made the case for claiming that those in positions of authority are generally not answerable for the actions of those under their authority, where principals are not answerable for the actions of their agents.

5 Answering Further Concerns

I am suggesting that respondeat superior is not a valid doctrine in general, because it cannot deliver the conclusion that a superior is responsible for the actions of those done under their charge. To clarify this point, let’s look through a few possible concerns.

As a first concern to address, we may worry about how much of the case was made by thinking about corporations and corporate liability in particular. I suggested that corporations as agents can be responsible—they have reasons to behave as they do and can answer in terms of them—but they are nevertheless not responsible (in the sense relevant for vicarious liability) for the conduct of their employees. But, on a fairly common way of thinking about corporations, this is all quite anthropomorphic. References to corporations are, after all, just references to a certain plurality of individuals (e.g., Hasnas 2010) or else a legal fiction (e.g., Jensen & Meckling 1976). So, there’s no sense to be made of talking about their reasons for action. Understanding firms as abstract entities of some kind, we may also question whether any conduct of employees can really be attributed to them.

Now, I disagree with this characterization of corporations. Following a long line of proponents of corporate moral agency,Footnote 21 I think that corporations can act and be responsible. They are not mere fictions. But this is not the place to see through this debate. Instead, we should quickly note two points that hold even if I am wrong: First, if nothing can be attributable to firms themselves non-metaphorically, and they cannot really have reasons, then the case is even stronger for saying that we are unjustified in using respondeat superior to conclude their criminal responsibility.Footnote 22 It may be that we cannot use the corporate case as I have to argue against the validity of respondeat superior as a doctrine, because there is no actual superior whose responsibility we are failing to validly infer. Still, the damage done is much the same. The primary use of respondeat superior in the criminal law is applied to corporations, and we will have undermined this use.

Even apart from corporations, though, the challenge to respondeat superior as a general doctrine stands. We may wonder, for example, whether we should infer parental responsibility from the responsibility for crimes committed by children. We may take the conduct of children to be appropriately attributable to the parents, and so respondeat superior would arguably be similarly appropriate. However, not only do we not have a tradition of applying the doctrine in this context, but the justification for this tradition could be the very same as that given above. Even when the conduct of children is attributable to the parents, those parents lack the ability to answer for the crimes of children in terms of the reasons why the crime was committed. Parents may often have authority over their children, but they do not seem to have the general authority to set the reasons of their children by themselves answering why the crime was committed. We do not need to resolve any metaphysical debate about corporations to see this point, and it equally suggests that respondeat superior is invalid.

As a second concern for the above proposal, one might think that I am being drawn into the same line of thought as Shoemaker in his objection to Duff as discussed above. Have I not just said, as Shoemaker did, that the agent in this case is not answerable because they cannot answer in terms of the relevant reasons? Going back to the blackmail case, Duff may recognize that the firm cannot offer the reasons of the employee, but he can nevertheless say that the firm answers in terms of what they have done for due diligence. Moreover, using what I said above, perhaps the firm could take responsibility for the conduct of their agents, taking themselves to be generally answerable for this conduct even if in this specific case they cannot answer in terms of the employee’s reasons. The worry is that if having an answer is not necessary for being answerable, then perhaps firms are answerable after all even if they cannot answer in these cases.

This point about due diligence is significant and will be discussed more below. For now, however, note first that it is not as if firms generally take themselves to be answerable for what their employees do, only being unable to answer and needing to take responsibility for the bad conduct of employees. The firm generally lacks access to or authority over the reasons for their employees’ conduct, so the suggestion would need to be that firms take responsibility for all employee conduct.Footnote 23,Footnote 24 And this is taking responsibility for quite a lot. The shopkeeper, in contrast, can generally answer for what she does, though she may be willing to take responsibility for some of her inadvertent behavior, likely having a reasonable expectation that she will be able to control what she does and avoid inadvertently selling tainted meat. For firms, though, it does not seem like they should have a similar expectation. As I will discuss more below (and in note 27), firms likely do not and should not have the kind of control over their employees necessary for it to be reasonable for them to take responsibility for employee conduct.

As a third possible challenge, we might doubt that the firm needs access to their employees’ reasons to be vicariously criminally liable. After all, we do think that some cases of vicarious liability go through, but it seems obvious that, in all cases of vicarious liability, the liable agent will not be able to offer the reasons why the crime was committed. When I am vicariously liable, it is stipulated that the crime was not committed by me, so it is no surprise that I cannot answer for it. Unless we are ready to reject vicarious liability as a form of liability, though, we seem pressured into admitting that answerability is too high of a standard to meet.

If there is simply no way to ever be able to answer for the crime of another, then so much the worse for vicarious liability generally in the criminal law. However, I do think that vicarious liability can be maintained while continuing to require answerability. Critical is recognizing how it is possible to be in a position to answer for the actions of another. What we need are situations in which one agent has the authority to give the reasons why another agent acts. Though uncommon, there are cases where one agent has the power to apparently set the reasons of another. This occurs in contexts in which one agent dominates another, determining the option set for the acting agent and their incentives. When this happens—when one agent sufficiently controls another—it does seem plausible to think of that agent as answerable for what that agent does.

Consider another setting for vicarious liability, one that strikes me as more appropriate: In law pertaining to military conduct, there is much discussion of ‘command responsibility.’ This is taken to be a principle analogous to respondeat superior used to hold those in command responsible for the conduct of subordinates (Solis 2010: ch.10).Footnote 25 As a context for vicarious liability, I think this stands a much better chance of successfully conferring liability onto superiors, and the reason is simple: Subordinates are often acting directly on the orders of superiors. They must obey, and they can be imprisoned for failing to. So, the reason the subordinate does this or that comes from the superior officer.

When we think about how superior officers may be guilty of criminal wrongdoing because of the conduct of their subordinates, it may be that they are guilty in some stronger sense. Perhaps they explicitly directed the subordinate to act illegally, and this may constitute an inchoate crime like solicitation, conspiracy, or incitement. But suppose they did not direct the illegal conduct. Still, the superior has such explicit command of the conduct of the subordinate, regularly enjoining them to do this or that, that it is more plausible to say that they are answerable for what the subordinate does. Not only is the subordinate’s conduct more clearly attributable to them, but superior officers are in a position to answer for the reasons motivating their subordinates to act insofar as they are understood to be the fount of those reasons.

Now, this is not absolute. It may be that military superiors sometimes lack the control necessary to genuinely be vicariously liable. But at least the presumption is that superiors will be answerable in this way, because the military is structured to dominate the agency of servicemembers. It is taken for granted that subordinates act at the behest of their superiors, because that is the standard set for their conduct. So, a form of vicarious liability does seem appropriate. Conduct is done for the reasons given by superiors. In cases where subordinates act illegally but outside of their directives (though perhaps not explicitly against their standing orders), then the superiors are answerable—not in the sense that they can give answers in this case, but because they generally assume responsibility for the conduct of those under their charge.

If this kind of control is needed for an agent to be genuinely responsible for the actions of someone acting under their authority, then it is clear that merely having some kind of authority over someone will not guarantee responsibility for their conduct. Moreover, given how high this standard is, I think the presumption should be that an inference to the superior’s responsibility seems invalid in most of the cases in which respondeat superior is typically applied. In workplace, non-military cases, many employees have wide discretion concerning how they spend their time and execute their duties in the firm’s interest. And this wide discretion undermines the idea that the firm can generally expect to be answerable for employee conduct. This looser authority relation makes it harder to understand even an employee’s immediate superior as answerable for their conduct,Footnote 26 and it makes it even more implausible to understand the firm itself as answerable.Footnote 27

It may certainly be that employees can have their reasons influenced by the reasons that bear on the firm’s conduct—they may purposefully act in the firm’s interest. They may even commit crimes in part in the interest of the firm. If they were encouraged to do this by managers or the incentives of the firm, then perhaps there is again a more direct story to tell for how either managers or the firm itself may be guilty of an inchoate crime. Absent this direct evocation, however, it is hard to see how the authority of the firm is such as to guarantee firm responsibility for employee wrongdoing. So, as responsibility for an individual in authority cannot be generally inferred from the fact that a crime is committed by an individual under their authority, the doctrine of respondeat superior is invalid.

6 Moving Beyond Respondeat Superior

I hope to have shown that respondeat superior is invalid as a doctrine; it cannot be appropriately used to secure the liability of one agent with authority over a second agent for the wrongful conduct of the second agent. And it cannot be used to do this because it is not necessarily or even presumptively true that the agent with authority is in a position to be able to avow the reasons for why the wrongful action was performed. Though these agents are in some kind of position of authority, they do not standardly possess sufficient control over their charges to set the reasons for which they act. Perhaps they can possess such control. Perhaps domains in which this level of control is expected can allow for vicarious liability, and it may be that vicarious liability can be pursued in other domains where this level of control can be demonstrated. As a general rule, though, we cannot accept respondeat superior.

Luckily, a rejection of respondeat superior does not have to mean a rejection of any kind of criminal culpability for agents who preside over wrongdoing. When thinking about significant agents like firms, it is important to see how abandoning respondeat superior need not lead to abandoning all routes to culpability when their employees act wrongly.

It may be appropriate to hold firms responsible as accomplices to the criminal act, guilty of aiding and abetting (Moohr 2007). Even failing this, I think we can still hold onto the idea often suggested that firms can be criminally responsible insofar as they fail in their duty of due diligence. The focus on due diligence is nothing new (e.g., Fisse 1983; Busy 1991), and it allows us to hold the firm responsible for criminal negligence, where they failed to meet a reasonable standard of due diligence to discourage illegal conduct. This suggestion of appealing to negligence is already fairly well established (Fisse & Braithwaite 1988: 486; Weissmann 2007; Weissmann & Newman 2007; Sheley 2019).Footnote 28,Footnote 29

This charge of negligence is importantly different than pursuing vicarious liability in terms of who bears the burden of demonstrating a failure of due diligence. For vicarious liability, the prosecution’s case is made merely by showing that the employee committed the crime. The crime of the employee establishes the crime of the firm, so then it is on the firm’s defense to offer a justification via establishing due diligence. For negligence, however, not only must the prosecution show that the employee committed the crime, but they must establish the failure of due diligence, since the crime consists in this failure. The defense then must answer by justifying or excusing the inadequate state of their monitoring/reporting practices.Footnote 30 This is a higher standard, it but seems appropriate given the nature of the crime. The firm has failed to do enough to adequately control its agents, not inadvertently committed crimes.

There are challenges with this approach. It requires thinking much more about how to spell out and justify the duty the firm is failing to obey.Footnote 31 What is necessary for due diligence, and how is this affected by changing technology?Footnote 32 We may also worry about the efficacy of this kind of approach—Is it right to rest so heavily on compliance?Footnote 33 Nevertheless, that there is a heathy tradition of scholars proving out other ways of asserting corporate criminal liability will hopefully make readers more comfortable recognizing that respondeat superior is invalid and should be abandoned.