Corporate Moral Responsibility

At the time of its collapse, Arthur Andersen was one of the Big Five accounting firms. But as a result of its role in the Enron scandal, Andersen was forced to shutter its business.

Andersen’s conviction raises numerous questions of philosophical interest relating to the possibility ofjoint action and the proper assignment of responsibility. For example, at the mid­point in its deliberations, the Andersen jury was at an impasse. Each juror believed that at least one Andersen employee culpably ordered the document shredding but the jurors could not agree about which employee that was. Assuming that the jury had deadlocked over the identity of the culpable employee even while each member believed that some employee was culpable, could the jury as a body have concluded that an employee impermissibly ordered the document shredding? For example, could the jury have found that Andersen had ordered the shredding of documents if six jury members believed that only employee Jones had ordered the shredding and the other six members believed that only employee Smith had done so? This is a question about determining when individual group members’ acts amount to an act of the group and, if so, what the nature of the group act is. Questions of this kind will arise for any account that aims to conceptualize a “joint agent” (Bratman, 2014) or a “plural subject” (Gilbert, 2000), as some take the corporation to be (List and Pettit, 2011). But the Andersen case raises a second question: Suppose that the jury members succeeded ultimately in agreeing upon the identity of the perpetrator.

That question has two different meanings, and I begin our inquiry by drawing out the dis­tinction between them.

28.1.1 Ascribing Acts and Assigning Responsibility

It is clear that corporations can act only through individuals. Still, as I and others have argued at greater length (French, 1984; Hess, 2014; Sepinwall, 2016a), this constraint is consistent with corporate agency. Corporations can and do function as agents even though they require indi­viduals to carry out their acts. In this respect, they are like armies, which carry out their own operations even though they require individual agents to perform the maneuvers of which the operations consist. One way to see that corporations act is to consider that some acts can be attributed only to the corporation qua group agent. For example, corporations can merge, dissolve, enter into contracts, and so on. To be sure, all of these acts require that individuals do

some things — in particular, the underlying acts that together effect the merger, dissolution, con­clusion of a contract, etc. But the corporation’s, e.g., entry into a contract is its own act — none of the individual members of the corporation have entered into the contract (none of them, for example, is bound by the contract’s terms in his or her individual capacity), nor has the aggregate of them entered into the contract, which again would entail that every one of them is individually bound by the contract’s terms.¹ Moreover, the corporation has in a sense directed the individual actions that together constitute its act — it gives the individuals in question a reason to act as they did, and it outlines the chain of command that integrates their individual contributions (French, 1984).

When we ask whether a corporation is morally responsible for some act, we might have two very different things in mind. First, we might be asking: when may we ascribe an act of some corporate employee to the corporation itself? That is, when should we take an employee’s act to be her own — off the clock, or outside the scope of her employment? And when should we take her act to be an act of the corporation? After all, it will be appropriate to blame the corporation for a wrongful act of an employee only if that act is properly ascribable to the corporation. So, in any case where there is a question of the corporation’s blameworthiness, we must first ask a prior question about whether the wrong in question is properly ascribed to the corporation. Call this a question of act ascription. This is an ontological question about how we link up acts and agents, deciding which agents own which acts. But there is a second question: Once we determine that an employee’s act is an act of the corporation, and that the act is morally wrong, whom may we blame? In particular, is the corporation an appropriate target of censure? Call this a question about the proper assignment of blame. Having determined that the employee’s wrongful act is properly ascribed to the corporation does not automatically entail that the cor­poration should be held morally responsible for that act because it is not clear that corporations are the kinds of entities that can bear moral responsibility in their own right. Or, more formally, it is not clear that corporations are moral agents.

Before elaborating on the debate about corporate moral agency, it will be useful to note that the literature on corporate moral responsibility is rife with accounts that conflate the ascrip­tion and assignment questions.

To see this, consider that there is no inconsistency in recognizing that an individual has committed a wrong and yet there is some feature of her that forecloses our finding her mor­ally responsible. Most relevantly here, a person who is below the age of moral maturity, or who operates under cognitive or emotional deficits that undermine her moral agency, is not blame­worthy for her wrongs. She lacks one or more capacities necessary to ground our taking her to be morally responsible. This may be because of a permanent deficit afflicting her (e.g., she is cognitively challenged and so cannot reliably distinguish between right and wrong) or instead because she is temporarily compromised (e.g., she has been hypnotized or drugged or other­wise made incapable of exercising her moral agency).

There is no reason to think that the situation for corporate agents is any different. Corporations can act. They can, in particular, commit wrongs, understood here as the infliction of harm without justification or excuse (Gardner, 2005). That they can do so does not yet entail that they are blameworthy for those wrongs.³ They would turn out to be blameworthy only if they satis­fied some further conditions, and I will go on to deny that they do. But we should pause first to consider why some other theorists have been at such great pains to establish that corporations are indeed moral agents and thus appropriately liable to blame when they transgress.

There is much at stake in the question of corporate moral agency. In the most acute cases, a corporation unjustifiably imposes massive harm and yet there is no one individual who culpably contributed to that harm. If it turned out that the corporation was not a moral agent we would appear to have wrongdoing for which no one was answerable. One might think that pragmatic necessity therefore entails corporate moral agency. Otherwise, corporations would be able to perpetrate wrongs with impunity. I turn to accounts that seek to leverage pragmatic consider­ations in support of corporate moral agency before examining accounts that defend corporate moral agency more directly.

28.1.1.1 A Residue from Corporate Wrongs

Recently, a handful of theorists have explored discursive dilemmas as part of an effort to argue that a corporation (or other institutional group) can be morally responsible for an act even if no member of the group would be morally responsible for that act (Pettit, 2007; Copp, 2006; List and Pettit, 2011).⁴ Discursive dilemmas arise where a group decision procedure yields a decision for the group to which none of its members would assent. To see how this works, con­sider a committee decision procedure that asks each member of the committee to vote on a Y/ N question by voting on a series of sub-questions, and the votes on the sub-questions are then aggregated to determine the outcome for the committee itself.

Philip Pettit argues that the “permanent possibility” that the group decision will diverge from that of each of the group’s members “indicates that... the members will have to create a group agent that comes apart in a manner from the way that they are individually disposed” (p. 183). The positing of a group agent is then a rational response to the possibility of discur­sive dilemmas.⁶ But this group agent must do more than render coherent individuals’ acts or decisions on behalf of the group. Where the group arrives at a decision that none of its members endorses, the group must bear responsibility for that decision. As Pettit explains it, there stands to be “a deficit in the accounting books, and the only possible way to guard against this may be to allow for corporate responsibility of the group in the name of which they act” (p. 194). The idea of a “deficit” or “residue” motivates much of the literature defending corporate moral respon­sibility (List and Pettit, 2011; Dempsey, 2013). The worry is that corporate harms for which we cannot distribute full responsibility to the corporation’s individual members will leave us with a remainder — an apparently intolerable state of affairs demanding that the corporation incur the responsibility that remains along with the blame it warrants (cf. Garrett, 2015).

In response to these accounts, one might well wonder what the remainder consists in and why it must be assigned anywhere. If no individual member is to blame, it is not clear why we should think there has been a collective wrong. To be sure, in cases where one experiences a setback to one’s interests, it might well be more satisfying to assign moral fault somewhere. But the mere satisfaction that the assignment would produce does not entail that the fault must be assigned somewhere, less still that it must rest with the group (Miller, 2006, p. 187). Corporate harm that results from the aggregation of individual contributions, all of which are innocent in their own right, might involve nothing other than sheer bad luck. If (contrary to fact) the explo­sion of British Petroleum’s (BP’s) Deepwater Horizon leading to the massive oil spill in the Gulf of Mexico had resulted only from a system malfunction that could not have been prevented, along with one or more blameless human errors, then we would have to accept that the spill, as devastating as it was, was just a massively unfortunate accident (National Commission, 2011). We could, of course, still hold BP accountable in other ways — notably, we could require that it pay to clean up the Gulf. But we would have to consign ourselves to the fact that there was no one and nothing to blame (cf. Wolf, 1985).

In some cases where no individual acted in a blameworthy fashion, then, we might conclude that the corporation is blameless too. In other cases — including cases of discursive dilemmas — we might instead allow that members of the corporation are blameworthy, qua members, even though each of them is individually blameless all-things-considered. This, I take it, would be the appropriate way to respond to individuals who participate in a group decision, each deciding in favor of the morally wrong option but for reasons that are innocent in themselves (cf. Copp, 2006, 214—220). Consider, for example, the board of a private prison corporation as it faces a multi-factor decision about whether it should lobby the government to increase mandatory minimum sentences for non-violent crimes. The board ends up voting in favor of lobbying — a decision that, I assume, would be morally wrong (see Hughes, 2018; Isaacs, 2018) — but each individual member has a justification or excuse for his or her contributions to the wrong. For example, one of them, who would have voted against lobbying, had to miss the meeting at which the decision was made to stay home with a sick child. Another always follows the lead of the missing member and, given that he was not present and she had not done the work to know which way to decide, she simply voted with the majority, knowing that her vote would not make a difference to the result. Yet a third member genuinely but mistakenly believes that current prison sentences provide insufficient deterrence, etc. In cases like this, David Copp has argued that “no member is all-in blameworthy” (Copp, 2007, 373). So, on Copp’s account, we may not assign moral responsibility to any of the individual members.

I do not believe that we are compelled to Copp’s conclusion that there is no individual moral responsibility in cases where each group member has a justification or excuse for his contribution to the group act. Copp can view each of the individuals in his hypotheticals as blameless only because he presupposes that conflicting reasons and duties can fully cancel one another out. Yet those who subscribe to the notion of a moral residue (e.g., Williams; Marcus) will have no trouble making sense of the idea that the members who were subject to these conflicting obligations bear guilt for the obligation the conflict forced them to forsake. Further, even if one rejects the idea of a moral residue one would still not need to conclude that there is a responsibility deficit that only corporate moral responsibility can fill. One could allow the group’s responsibility to redound equally to its members, as an account of shared responsibility would mandate, or one could “apportion[] responsibility among the members of a group... as and to the degree to which each contributed to the group act” (Ludwig, 2007, 411). We should prefer corporate moral responsibility to these alternatives only if the corporation genuinely can bear responsibility in its own right.

28.1.1.2 Corporate Moral Agency and Blame

What we need in addition to a theory that answers the question of ascription is an account that establishes that the corporation is the kind of actor that is an appropriate target of our judgments of moral responsibility and the blaming responses those judgments prompt. One can seek to argue for or against this conception of the corporation in two different ways. First, one can identify the necessary and sufficient conditions for moral agency and then argue that the corporation does or does not possess them. Alternatively, one can identify the rationale for our blaming practices and then argue that blaming the corporation does or does not satisfy that rationale (Tollefsen, 2015). The first strategy is unpromising given the deep disagreement over just what the necessary capacities for moral agency are, let alone whether the corporation possesses them (see Sepinwall, 2016a). The second strategy, by contrast, is, I believe, decisive. To refute the idea that the corporation is an appropriate object of blame, it will suffice to show that blaming the corporation cannot achieve the end that blaming in general seeks. To be sure, the second strategy cannot avoid pursuing some of the considerations explored in the first — once we determine the end of blame, we will need to assess whether the corporation has the capacities that that end presupposes. But the inquiry will be more focused for it is possible that the conditions for full moral agency are not entirely co-extensive with those for being an appropriate object of blame. Of relevance here, it might be the case that some capacity in addition to those necessary for moral agency is required in order for one to be an appropriate object of blame. Indeed, that is just the line I will pursue, arguing that the capacity to experi­ence guilt is necessary for being an apt target of blame even if it is not necessary for moral agency.

To begin: how can it be that someone is a moral agent and yet not liable to blame for their wrongdoing? The disconnect between the concepts of “moral agent” and “blameworthy” arises because, in asking what it takes to be a moral agent, we are often asking a prospective question — namely, is this individual or entity capable of judging right and wrong and acting in line with that judgment? But in asking whether someone is blameworthy, we are always asking a retro­spective question — namely, would it be appropriate to blame this entity or individual for what he has done?⁷ And it is possible for someone to reliably render moral judgments and act on those judgments and yet to not be blameworthy even when she decides to do something she correctly assesses to be wrong all-things-considered.

To see what I have in mind, consider Mr. Spock: he is cool-headed and entirely capable of abiding by his will (cf. Charland, 1998; Warren, 1973). If he intentionally harms someone else without justification or excuse, we can imagine that he has done so after a suitable process of deliberation, assigning all of the relevant moral considerations their due weight, nonetheless deciding to pursue the morally wrong course of action, while in complete control of his will. Why then might we find this Mr. Spock, who has dispassionately chosen to commit a wrong, an inappropriate object of blame?

To be clear, the question is not whether it would be appropriate all-things-considered to blame Mr. Spock. Sometimes, we rightly withhold expressions of blame, or we refrain from blaming at all (i.e., forming whatever attitudes or experiencing whatever emotions blame involves — more on this below), because of circumstances surrounding the wrongdoing. If, for example, the wrong was minor (Spock dismissed a reasonable proposal with undue harshness) and Spock was having a particularly bad go of it (leading a crew that threatens insubordin­ation, and facing the prospect that his shuttlecraft won’t have enough fuel to return to the Enterprise), then the victim of the crime might decide not to blame Spock. What would that involve?

On one possibility, the victim would choose not to express blame — she would hold her tongue, rather than reproaching Spock. But I want to consider a more dramatic possibility: she might forswear blame altogether. To do so, the victim would not need to disavow the judgment that Spock had acted wrongly in injuring her. For blame does not consist merely in a judgment of wrongdoing. If it did, two co-felons’ gleeful celebrations of their successful crime — which rest on a judgment that they have done something wrong — would be an instance in which each blames the other. So the belief that they have done wrong cannot be all that blame consists in (e.g., Wallace, 2011, p. 348; Coates and Tognazzini, 2012). Instead, blame has an essential emo­tional component. It is this component that the victim would forswear were she to decide not to blame Spock.⁸

In forswearing the emotional component of blame, the victim would decide not to resent Spock. As such, she would forego the phenomenological correlates of anger (the state of excited negative emotion), along with whatever else resentment entailed (e.g., perhaps a desire for ret­ribution). Further, in forswearing resentment, the victim would, I submit, desist from blaming altogether. But why should that be? One answer to that question can be found in accounts of blame that conceive of it as essentially constituted by the reactive attitudes. Peter Strawson offers the seminal account (1962 and 1993). For Strawson, our assignments of moral responsibility are constituted by the emotional responses we have to the attitudes and intentions of others as these are displayed in their actions (1993, pp. 49, 56—57). When it comes to wrongdoing, these reactive attitudes consist of a perspective-based triad of emotions — guilt, for our own wrong­doing; resentment, for wrongdoing against us; and indignation, for wrongdoing aimed at others (Strawson, 1993). Importantly, for Strawson, these reactive attitudes are not the products of a finding of blameworthiness. Instead, they constitute that finding. As Gary Watson puts it, on a Strawsonian account, “to regard oneself or another as responsible just is the proneness to react to them in these kinds of ways under certain conditions” (1987, 256—257). As such, for Strawson, there can be no blame without the moral emotions.⁹

I want to suggest a different way of foregrounding the essential role of the reactive attitudes in blame. While reactive attitude theorists typically focus on the emotional experiences of the blamer, I shall now argue that a capacity for emotion on the part of the blamee is necessary for blame to have uptake. It is because Spock lacks a capacity for emotion that he cannot be blameworthy.

To see this, consider the practical ends that blame seeks. Theorists offer different accounts of the function of blame. For example, we blame to register an impairment in our relationship with the offender (Scanlon, 2008), to record a demerit in the offender’s moral ledger (Zimmerman, 1988), to communicate that the offender’s act was wrong (Duff, 2003); to “protest” (Hieronymi, 2004) the offender’s act, and perhaps also to punish (exposure to blame being painful in its own right) (Sepinwall, 2016b). At least the last two of these, I contend, require emotion if the end they seek is to be achieved (cf. Wolf, 2011).

Start with protest. As Pamela Hieronymi describes it, when we warrantedly judge that someone has wronged us, “our first response is, and ought to be, anger and resentment.... In resentment the victim protests the trespass, affirming both its wrongfulness and the moral significance of both herself and the offender.” Hieronymi argues that that resentment is critical to blame by focusing on the change wrought by forgiveness. When we forgive, she describes, we do not renounce the judgment that the offender has wronged us. Instead, forgiveness involves giving up a different judgment — namely, the judgment that the moral significance of the wrong and one’s worth stand in need of affirmation. That is the judgment that grounds one’s resent­ment. When the offender atones, he makes it the case that the blamer no longer has reason to hold that judgment; as a result, “resentment loses its footing” (24) (cf. Rosen, 2015). So one change forgiveness brings about is to move the blamer from a state of anger to a state where anger makes no sense.

But anger is crucial not just for the blamer’s stance but for the blamee’s as well. For blame to function as protest, the blamee’s emotions must be engaged too. Hieronymi describes the protest as a “claim” or a call to attention. Importantly, it is not just the content of the victim’s claim — namely, that she has been wronged — that does the calling; the anger itself plays an essen­tial role here. “Anger... seeks to set things right by demanding the recognition of a wrong.” Angry blame does so by making manifest the pain the offender has caused. It allows the offender to see the effect of his transgression in all its force. The empathic offender — the one who can appreciate the blamer’s pain — cannot help but feel pain himself, knowing that he is the source of the anger that he is, through blame, intended to absorb (see Sepinwall, 2016b).¹⁰ In short, anger brings the wronging to light.

One who is inured to anger, like Spock, will not experience the protest as protest. In this respect, Spock is like a young child who knows she has broken a rule in, say, deliberately knocking over her brother’s block structure but who does not and cannot react to her brother’s rebuke with any kind of moral emotion at all. She can see that her brother is upset, and she can see that she is the cause of that upset — more specifically, that her rule-breaking act is the cause of his upset. She may fear sanction or retaliation. But she, like Spock, cannot experience the dis­tinctive pain of guilt.¹¹ For Hieronymi, blame is meant to induce that pain; the anger in blame is meant to convey the full force of the wrongdoing. Blame will not then have its intended uptake if its target is insensitive to the protest blame conveys.

Anger is necessary for blame to function as punishment too. Butler’s Sermons are frequently quoted as one wellspring for the view that resentment (which again I take to be constitutive of blaming) aims at punishment: He describes resentment as a “weapon put into our hands by nature, against injury, injustice, and cruelty” (116; see, e.g., Hieronymi), which aims for “the misery” of the offender (127). I have articulated blame’s punitive dimensions at greater length elsewhere (Sepinwall, 2016b). For now I want to focus on this “misery” and why it is that we should understand inflicting it to be among blame’s key functions.

Hieronymi understands anger as indispensable to calling the offender to repair his wrong. But on a punitive account, blame already goes some way toward rectifying the status imbalance that the wronging wrought between the offender and his victim. It does so at least in part by inflicting pain. To borrow from Jean Hampton’s general understanding of punishment, the puni­tive dimension of blame functions to debase the offender, by subjecting him to the pain of being an object of reproach (1992). Blame, that is, already effects the end of restoring the status of the victim that protest seeks by treating the offender as one who deserves pain and then inflicting that pain to right the moral balance.¹²

Now, if all of this is right, then blame crucially relies on the blamee’s having the emo­tional bandwidth to receive the blame as protest and punishment. But that is just what those like Spock lack. More specifically, there are two ways in which one might understand Spock’s infirmity. On one understanding, what Spock lacks is the ability to feel guilt, or to appreciate first-hand the blamer’s experience of anger. As such, the missing piece is a capacity for affect. On a second understanding, what Spock lacks is the capacity for guilt tout court. The thought here would be that dispassionate guilt — guilt with no phenomenological correlates — is not guilt at all. I am inclined to this second view, according to which emotions necessarily involve their felt aspects (see Hurley). But I will not aim to defend that view here, nor need I rely on it. It suffices for my purposes to establish that blame’s essential ends require affect. For it is on this basis, we shall now see, that blaming corporations falters.

28.1.2 Corporations as Irrational Targets of Blame

If blame aims to inflict pain on the offender — as a matter both of having him appreciate the full force of his wrongdoing and having him incur punishment — then the target of blame must have the capacity to experience pain. If the offender does not have this capacity then blame will fail to achieve its end. It is clear that the corporation has no capacity for pain. This is just the reality underpinning the by-now cliched lament that the “corporation has... no body to kick” (Coffee, citing Baron von Thurlow, 1981). If I am right that a capacity to experience pain is necessary for blame to achieve its ends, and if the corporation lacks a capacity for pain, then it follows that it makes no sense to blame corporations.¹³ To be sure, we might still have reason to offer statements condemning morally impermissible corporate conduct and even to experience and express outrage when we do so. We would offer these statements, perhaps in an enraged state, to let others know that we take it to be the case that no one should act as the corpor­ation has, and to affirm the victims’ sense that they have been treated in a way that no moral agent should treat another. But we could not rationally be protesting the corporation in the sense that we were calling it to appreciate its wrongdoing; nor could we be seeking to punish it through our expression of blame. For again the corporation would be insensible to our call or our punishment.

But perhaps corporate blame need not rely on the corporation’s own capacities. For what if we could preserve the protest and punitive dimensions of blame by directing the pain underpin­ning these functions to the corporation’s individual members? Deborah Tollefsen has something like this strategy in mind when she argues that we can secure a capacity for corporate remorse by relying on the capacities for guilt of the corporation’s members (Tollefsen, 2003).

Contra Tollefsen, I am not convinced that members’ “vicarious” feelings (Tollefsen, 2008, 12) can secure the ends of blame.¹⁴ If blame aims at making its target feel bad, then those who blame the corporation aim (however misguidedly) to make it feel bad. The fact that someone else will feel bad won’t suffice to make corporate blame sensible. The corporation will not itself have absorbed the blamer’s anger, and so his blame will not have had its intended uptake. So recruiting members to feel on the corporation’s behalf will not overcome the corporation’s inability to feel in its own right.¹⁵

I take the foregoing arguments to ground at least a deep skepticism about the appropriate­ness of blaming corporations, though perhaps not a decisive argument against doing so.Yet even if it turns out that the corporation can bear blame, it behooves us to consider whether targeting the corporation is the only, let alone the best, way of blaming when the corporation commits a wrong.

28.2