Forthcoming in the Oxford Journal of Legal Studies

CORPORATE CRIMINAL RESPONSIBILITY AS

TEAM-MEMBER RESPONSIBILITY Ian B. Lee*

Abstract

This article puts forward a theory of corporate criminal responsibility as the shared responsibility of the members of a team for wrongdoing committed by one of their number in the pursuit of their common goals.

The theory of team-member responsibility advanced in this article differs from theories — such as those of Peter French and Phillip Pettit — under which corporate or group responsibility is viewed as the responsibility of the corporation or group as an autonomous moral person. Instead, this article defends a conception of a collectivity as a kind of relationship among individuals; under this conception, a collectivity is something more than the sum of its members, but not something having an autonomous existence apart from its members. To develop this idea, the article employs the concept of a team.

Where individuals participate in a team, doing their part towards a collective goal, they are entitled to share in the credit for the positive achievements resulting from their combined actions, and they also share in the discredit for negative achievements of the group. This is so for each member, regardless of whether the member’s individual contribution was a but-for cause of the group’s achievement.

Further, in the specific case of corporate teams, team-member responsibility can help to explain why there should be liability of the corporate entity for crimes committed by members of the corporation (a group which typically includes its employees, officers, directors and, in some circumstances, its shareholders) in the pursuit of the corporation’s goals.

Key words: corporations, criminal law, corporate criminal responsibility, corporate

criminal liability, collective responsibility.

* Faculty of Law, University of Toronto. A version of this article was presented at seminars at the Universities of Luxembourg, Sydney and Toronto, and at the Australian National University, Emory University and Queen’s University (Canada), and I thank the seminar participants for their comments. In particular, I am grateful to Bruce Chapman, Yasmin Dawood , Gail Henderson, Heidi Libesman, Paul B. Miller, Mayo Moran, Denise Reaume, Simon Stern, Hamish Stewart and Kevin Walton for discussions of this topic or comments on a draft. The author’s e-mail address is ian.lee@utoronto.ca.

2

CONTENTS

1. Introduction ................................................................................................................................................................... 2

2. State of play: existing theories ............................................................................................................................... 4

2.1. Neoclassical perspective .................................................................................................................................. 4

2.2. Group moral agency ........................................................................................................................................... 6

2.3. Assessment ............................................................................................................................................................ 9

3. Corporate theory and corporate criminal liability...................................................................................... 12

3.1. Relevance of corporate theory .................................................................................................................... 12

3.2. The corporation as a team ............................................................................................................................ 13

4. Team-member responsibility .............................................................................................................................. 19

4.1. Collectivities as teams .................................................................................................................................... 19

4.2. Nature of team-member responsibility .................................................................................................. 20

4.3. Cross-temporal teams .................................................................................................................................... 22

5. From team-member responsibility to corporate criminal liability...................................................... 24

5.1. Norms-signalling and the criminal law ................................................................................................... 25

5.2. The purpose of punishing a corporation ................................................................................................ 26

5.3. Regulatory offences ......................................................................................................................................... 28

6. Conclusion ................................................................................................................................................................... 29

1. INTRODUCTION

In this article, I develop a theory of corporate criminal responsibility as the shared

responsibility of the members of a team for wrongdoing committed by one of their number in

the pursuit of their common goals. I argue, in addition, that this understanding of corporate

responsibility for wrongdoing can help to illuminate the nature of and rationale for the practice

3

of imposing criminal liability upon corporate entities for crimes committed by their directors,

officers and employees in the pursuit of the corporation’s goals.1

The theory advanced here could fairly be described as a theory of collective responsibility.

However, it is to be distinguished from theories — such as those of Peter French and Phillip

Pettit — under which corporate or group responsibility is viewed as the responsibility of the

corporation or group as an autonomous moral person. Like French and Pettit, I shall suggest

that the corporation is a collective entity and not, for instance, a mere collection of atomistic

individuals. There is a difference between a collectivity and a crowd, and corporation is of the

former type. However, I am wary of anthropomorphic characterizations of the nature of

collective entities, and of a conception of the collectivity under which its identity is detached

from that of its constituent members.2

Instead, I view a collectivity as a kind of relationship among individuals;3 this makes a

collectivity something more than the sum of its members, but not something having an

autonomous existence apart from its members. To develop this idea, I employ the concept of a

team.

The article makes two basic claims. First, where individuals participate in a team, doing their

part towards a collective goal, they are entitled to share in the credit for the positive

achievements resulting from their combined actions, and they also share in discredit for

negative achievements of the group. This is so for each member, I shall argue, regardless of

whether the member’s individual contribution was a but-for cause of the group’s achievement.4

Second, I suggest that in the specific case of corporate teams, team-member responsibility can

help to explain why there should be liability of the corporate entity for crimes committed by

members of the corporation (a group which I shall suggest typically includes its employees,

officers, directors and, in some circumstances, its shareholders) in the pursuit of the

corporation’s goals. The article has the following structure. In Part 2, I describe the current

1 In this article, I shall try to distinguish between “criminal responsibility,” by which I mean moral blameworthiness in respect of criminal wrongdoing, and “criminal liability,” by which I mean the incurrence of punitive sanctions as a result of such wrongdoing. The article develops a theory of corporate criminal responsibility as a means of illuminating the rationale for corporate criminal liability. 2 Another attempt to stake out a position between methodological individualism and anthropomorphism is found in GR Sullivan, ‘The Attribution of Culpability to Limited Companies’ (1996) 55 Cambridge LJ 515, 538-39. 3 One of the virtues of the “nexus of contracts” metaphor (influential among economic analysts of corporate law) is that it captures the idea that the corporation consists of a relationship (or a set of relationships) among individuals. See Part 3. 4 "The argument has affinities with Christopher Kutz's argument that "intentional participation in a collective act warrants individual accountability for the consequences of that act." See Chrisopher Kutz, Complicity: Ethics and Law for a Collective Age (Cambridge University Press 2000) 11.

4

status of the theoretical literature about corporate criminal liability. In Part 3, I argue for the

relevance of corporate theory to the identification of the nature and purpose of the institution of

corporate criminal liability. The concept of team-member responsibility is elaborated in Part 4,

and its implications for the institution of corporate criminal liability are explored in Part 5.

2. STATE OF PLAY: EXISTING THEORIES

A theory of corporate criminal liability is a theory of the nature and purpose of the application

of criminal punishment to corporate entities.5 What does it mean to punish a corporation? What

purpose does corporate punishment serve? In particular, what does corporate punishment

achieve that other forms of public action, such as taxation, regulation, and the imposition of civil

liability, do not? To date, two theoretical perspectives have had prominence: (1) a neoclassical

perspective based on Gary Becker’s economic theory of the criminal law and the contractarian

conception of the corporation; and (2) a group moral agency perspective.

2.1. NEOCLASSICAL PERSPECTIVE

The application of the neoclassical law and economics framework to the issue of corporate

criminal liability yields the following prescription: while corporate liability is a useful cost-

internalization mechanism, it is far from clear that this liability should be criminal rather than,

for instance, civil or administrative.6

The function of laws that create liability, including both tort law and criminal law, is to attach a

price to behaviours that produce net negative externalities. For instance, assuming 100 percent

detection, a law imposing liability upon an actor in an amount equal to the social cost of the

activity (net of any social benefit) aligns the actor’s incentives with the interests of society. In

the real world, of course, some socially costly activities go undetected. Given that rational actors

will discount the nominal liability amount by the probability of non-detection, an upward

adjustment to the former is needed so as to ensure that actors will internalize the full net

external cost of their act.7

A corporation is not an individual. What is referred to in casual conversation as the “actions of a

corporation” are in reality actions taken by individuals in the context of an incorporated

5 Punishment is a definitional element of the institution of criminal liability, which distinguishes it from civil liability. See Karen Yeung, Securing Compliance: A Principled Approach (Hart 2004) 250. 6 John T Byam, ‘The Inefficiency of Corporate Criminal Liability’ (1982) 73 J Crim L & Criminol 582; Daniel R Fischel and Alan O Sykes, ‘Corporate Crime’ (1996) 25 J Leg Stud 319; Vikramaditya S Khanna, ‘Corporate Criminal Liability: What Purpose Does It Serve?’ (1996) 109 Harv L Rev 1477; John R Lott, Jr, ‘Corporate Criminal Liability,’ Encyclopedia of Law and Economics (2000) 494-95. 7Gary S Becker, ‘Crime and Punishment: An Economic Approach’ (1968) 76 J Pol Econ 169.

5

business. For instance, such actions might include a policy decision taken by the board of

directors to deploy the corporation’s resources in a particular manner, or an operational

decision taken by a mid-level manager. The function of imposing liability on a corporation is to

influence such actions; the underlying assumption is that the structure of rewards and sanctions

within the corporation is such that the pain imposed on a corporation is felt by the responsible

individuals.8

Many economic theorists of the firm believe that market forces lead rational individuals to

create corporate arrangements in which managers and the board of directors have strong

incentives to maximize corporate profits.9 From this belief it follows that society has reason to

be concerned that managers and boards may cause corporations to pursue profits regardless of

the cost for society.10 But it also follows that the behaviour of managers and boards of directors

will be highly responsive to corporate liability; monetary sanctions affecting the corporation’s

bottom line will cause corporate managers and boards of directors to internalize the social costs

of their decisions.11

It does not follow, however, that corporate liability should be criminal. A conventional economic

view is that the difference between criminal and civil liability resides in the availability of

distinctive sanctions under the criminal law, especially imprisonment.12 Imprisonment is useful

when the external costs of an activity (multiplied to take into account the probability of non-

detection) exceed the actor’s resources. In short, the prospect of becoming the defendant in a

civil suit is likely to have little impact on the choices of a pauper. The risk of losing his or her

liberty might have more.

Imprisonment is not an available sanction when the defendant is an “artificial person.” Instead,

a corporation convicted of an offence is most likely to be ordered to pay a fine. Given, however,

that the sanctions will be pecuniary,13 there is no reason for liability to be criminal rather than

civil.14

8 Fischel & Sykes (n 6), at p. 321. 9 See, e.g., Easterbrook & Fischel, The Economic Structure of Corporate Law (Harvard UP 1991). 10 See, e.g., Lawrence E Mitchell, Corporate Irresponsibilty: America’s Newest Export (Yale UP 2001); Robert Reich, Supercapitalism (Knopf 2007). 11 Easterbrook & Fischel (n 9) 39. 12 E.g., Byam (n 6) 599. 13 Stigma is sometimes cited as a non-pecuniary sanction that can be applied against corporate offenders. Aside from the obvious response that stigma imposes a deadweight loss (like imprisonment, but unlike a fine or damages), it may be questioned whether stigma is in fact a non-pecuniary sanction when it is applied against a corporation. Stigma is a reputational sanction; it affects the corporate bottom line through the reduced willingness of potential contracting counterparties to transact. The desire to avoid its impact on profitability is the source of the motivation of corporate actors to avoid doing things that will incur stigma. In other words, the way in which stigma operates on the behaviour of managers is

6

In fact, when the defendant is a corporation, civil liability seems preferable from an economic

perspective. Criminal prosecution contains procedural and substantive safeguards, especially

the reasonable doubt standard, the apparent purpose of which is to protect individuals from

unjustified incarceration. Where only money and not liberty is at stake, there is no good reason

to burden the cost-internalization process with these safeguards.15 In other words, the

institution of corporate criminal liability compares unfavourably with civil liability as a means

of behaviour modification.16

2.2. GROUP MORAL AGENCY

A very different approach to corporate criminal liability is to regard it as a response to the

corporation’s blameworthiness.17 This approach rests upon a philosophical argument,

associated with Peter French and Philip Pettit, that corporations are autonomous moral

agents.18 The latter argument consists, essentially, of three moves. First, group agents (such as

corporations) are distinguished from mere aggregates of individuals. Second, it is argued that

exactly the same as the way in which a monetary sanction operates on such behaviour. Consider that, in the case of individuals, where the level of damages necessary to achieve cost-internalization would exceed the defendant’s resources, the fact that criminal convictions carry stigma allows fines to be set at a somewhat lower amount in the knowledge that the defendant also receives a non-monetary penalty in the form of stigma. By contrast, when the accused is a corporation, there is a limit to the total amount of economic harm that can be visited upon it without driving it out of business. Stigma can be part of the package of harm, but when it is present the upper bound of the range of possible monetary sanctions is simply reduced accordingly. 14 Fischel and Sykes note that punitive damages can be used in the civil system to compensate for the probability of non-detection. Fischel and Sykes (n 6) 331. Other reasons suggested in the economic literature as to why criminal liability might be useful include the availability of probation and the possible utility of penalizing attempts. See, e.g., Wallace P. Mullin and Christopher M. Snyder, ‘Corporate Crime’ in Nuno Garoupa (ed), Encyclopedia of Law and Economics, (2d edn 2009) vol 3, 238. However, it is not clear why the kind of governmental control involved in a probation order could not be imposed in administrative proceedings, and how the taxation of or imposition of administrative or civil liability for risk-creating activities is functionally different from criminal attempt liability. 15 Fischel and Sykes (n 6) 332. 16 Dan Kahan has suggested a path to a neoclassical endorsement of corporate criminal liability: if one assumes that people have preferences for the law to express certain values, as it does when it punishes corporations rather than taxing them, then these preferences enter into the social welfare function and provide a reason for corporate criminal liability beyond its efficacy as a pricing mechanism. Dan M Kahan, ‘Social Meaning and the Economic Analysis of Crime’ (1998) 27 J Leg Stud 609. Kahan’s solution is intellectually unsatisfying because, in effect, it “assumes a can opener.” However, there is something to the idea that economic analysis can accommodate the notion that the criminal law does more than price behaviour. I shall return to this idea in Part 5. 17 See JA Quaid, ‘The Assessment of Corporate Criminal Liability on the Basis of Corporate Identity: An Analysis’ (1998) 43 McGill LJ 67; Brent Fisse, ‘Attribution of Criminal Liability to Corporations: A Statutory Model’ (1991) 13 Sydney L Rev. 277, 279; Brent Fisse & John Braithwaite, ‘Allocation of Responsibility for Corporate Crime: Individualism, Collectivism and Accountability’ (1988) 11 Sydney L Rev 468, 504-507. 18 Philip Pettit, ‘Responsibility Incorporated’ (2007) 117 Ethics 171, 172; Peter A French, ‘The Corporation as a Moral Person’ (1979) 16 Am Phil Q 207, 211 [hereinafter French, ‘CMP’]; Peter A. French, Collective and Corporate Responsibility (Colum UP 1984) [hereinafter French, C&CR]. See also Christopher Harding, Criminal Enterprise: Individuals, Organisations and Criminal Responsibility (Willan 2007).

7

group agents are capable of intentionality. Third, it is argued that group agents can fairly be held

morally responsible (i.e., blamed) for certain acts, decisions or consequences.

Group agents. What differentiates a collectivity or group agent19 from a mere aggregate of

individuals? For French, the key features are the existence of internal decision-making

procedures, the enforcement of standards of conduct on the part of individuals within the

group, and the immutability of the identity of the group despite a change in the identity of the

individuals playing any given roles within the group.20 Pettit’s conception is broadly similar.21

Intentionality. The second step is to argue that group agents act “intentionally” and that

collective intentionality is not reducible to individual intentionality. For French, the crucial

element is what he calls the “Corporation’s Internal Decision Structure” (its “CID Structure”) —

in essence, the set of offices and procedures that constitute the rule of recognition for corporate

decisions.22 The CID Structure enables an observer (at least in theory) to determine when a

decision is “of the corporation.” French further supposes that every corporation has

“established corporate policies,”23 which one surmises are to be gleaned from a global

examination of the decisions that have emanated from the corporation’s CID Structure, or

perhaps by examining some subset of them (e.g., policy directives so labeled by the

corporation). French asserts that when a particular corporate decision is consistent with the

corporation’s “established policies,” it may fairly be said that the corporation “intended” the

decision.24

Pettit adopts a simpler approach to intentionality: a group agent is capable of forming beliefs

and “action-suited desires,” and then acting in such a way that its desires are realized given its

beliefs.25 To show that group agency is not reducible to individual agency, and is in that sense

autonomous, Pettit draws upon the discursive dilemmas familiar to social choice theorists.26 A

well-known result in social choice theory is that, when the decisions of a group are submitted to

majority voting, they are liable to be internally inconsistent. In forming a group, individuals face

a choice between tolerating this inconsistency, on the one hand, and establishing a constitution

whereby the group acts consistently but its acts are detached from the judgments of its

19 Pettit uses the term “group agent.” French uses “conglomerate” to refer to the equivalent concept. 20 French, C&CR 13-14. 21 See Pettit (n 17) 172 (group agents “retain their corporate identity through continuous changes of membership,” they “endorse goals” and “follow procedures that enable them to pursue those goals”). 22 French, C&CR 41. Also French, ‘CMP’ 213. 23 French, C&CR 43-46. 24 French, C&CR 44. Also French, ‘CMP’ 213. 25 Pettit (n 17) 178. 26 Pettit (n 17) 181-82.

8

constituent members, on the other hand. When the latter choice is adopted, the individuals have

constituted the group as an autonomous agent.27

Responsibility of the group agent. Intentionality plays a crucial role in some philosophical

justifications of individual criminal liability. For instance, in one account, intentionality is

significant because of the perceived need for a non-instrumental justification of the application

of physical coercion by the state: the significance of intentionality is that, if the act that attracts

liability is intentional, then the punishment results, in a sense, from the defendant’s own will

rather than being inflicted upon him.28 In another philosophical account, intentional trespasses

against the person or property of another are criminally culpable wrongs, because culpability

consists of one individual treating another as a means to his ends, in violation of their equal

freedom.29

It is far from clear that intentionality should have the same importance in a theory of corporate

liability.30 There is no need to characterize the punishment of corporation as the product of “its

own will.” Moreover, the culpability of a corporation cannot obviously be grounded in a

principle prohibiting one individual from treating another as a means to his ends. A case has not

been made (and it is not obvious) that any such duty subsists between corporations (or more

generally intentional agents that are not human beings) and individual human beings.

In a theory of corporate transgression, concepts other than intentionality must assume greater

importance. French and Pettit pursue different strategies for supplementing intentionality.

French’s approach is to expand the basis of corporate accountability beyond intentional acts, by

developing two weaker principles, which he labels the “extended principle of accountability”

(EPA)31 and the “principle of responsive adjustment” (PRA).32 Under EPA, a person is morally

accountable for unintended effects the occurrence of which he “was willing to have occur as the

result … of his actions.”33 For instance, French argues that McDonnell-Douglas Corp. was morally

responsible in respect of the crash of one of the aircraft it manufactured because it was “willing

to manufacture and market” the aircraft despite “its knowledge” that the aircraft’s design

27 Pettit (n 17) 183-84. 28 A Brudner, The Unity of the Common Law (UC Press 1995); Punishment and Freedom (OUP 2009). 29 A Ripstein, ‘Beyond the Harm Principle’ (2006) 34 Phil & Pub Aff 216. 30 For a theory of corporate criminal liability that proceeds on the opposite premise, see Harding (n 17) 115 (stating that “the commission of [a] crime involves a more objectionable attitude and state of mind [compared to a tort],” and discussing the implications of this difference for corporate agency). 31 French, C&CR 134. 32 French, C&CR 156 and subsequent pages. 33 French, C&CR 134.

9

exposed passengers to a “higher probability” of life-threatening failure than competing

aircraft.34

Under PRA, a person becomes morally accountable for an “untoward event” caused by it, if it

does not subsequently take measures to prevent the event’s recurrence.35 The refusal to “adjust

one’s harm-causing ways … associat[es] oneself, morally speaking, with the earlier untoward

event.”36 For example, after an Air New Zealand passenger jet crashed into the side of mountain,

a commission of inquiry identified several organizational defects that contributed to causing the

disaster. Though it could not be said that the airline had intended the disaster or even had

willingly assumed the possibility of its occurrence,37 French concluded that Air New Zealand

was morally responsible in respect of the disaster because, afterwards, it did not act to correct

the defects identified.38

Pettit’s strategy is, once again, simpler. He argues that group agents do not only act

intentionally; they are also capable of forming “evaluative beliefs,” which is to say beliefs about

propositions of value.39 Group agents can make value judgments, in other words. A group agent

does so whenever it endorses a proposition of value presented for its consideration by its

members.40 Because a group agent is “capable of making a judgment on what is good and bad

and right and wrong” and of acting in consequence, it is “as fit as any individual human being to

be held responsible for what it does.”41

2.3. ASSESSMENT

In my view, neither of the prevailing theoretical approaches --- the neoclassical perspective and

the group moral agency perspective --- is fully satisfactory.

Group moral agency. The difficulty with the group moral agency theory is that the

blameworthiness of a group agent depends upon the existence of moral duties owed by group

agents towards individuals, and the existence of such duties has not been shown.

34 French, C&CR 138. 35 French, C&CR 156. 36 Id. 37 French, C&CR 154. 38 French, C&CR 161. 39 Pettit (n 17) 186-87. 40 Pettit (n 17) 187. 41 Pettit (n 17) 192. According to Pettit, the capacity to make value judgments distinguishes humans and group agents from robots and non-human animals --- even though the latter are capable of acting intentionally in the sense of forming desires and beliefs and acting in pursuance of them, they cannot make value judgments (Pages 178, 185.) Similarly, although French appears to concede that non-human animals can act intentionally (French, C&CR 166), they are, unlike corporations, incapable of “responsive adjustment” and are therefore not moral persons (id.).

10

Consider what it is, in French’s argument, that makes an act on the part of a group agent

wrongful and therefore something capable of engaging the group agent’s moral responsibility.

The answer, for French, is supposed to be supplied by the EPA and PRA principles. Yet, these

principles themselves depend upon a theory of the moral duties owed by group agents towards

individual agents, and French does not provide such a theory.

Thus, when French describes the responsibility incurred by the McDonnell Douglas Corp. under

EPA, he does not and cannot treat it as sufficient that the corporation’s design choice exposed

passengers to a risk of harm. Instead, the critical fact is that the design of the MD aircraft carried

a greater risk than that of competing aircraft manufactured by the company’s rivals.42 French’s

account assumes, in other words, the existence of an implicit principle of moral behaviour that

the manufacturers of aircraft should not produce and sell aircraft that expose passengers to a

greater risk of death than competing aircraft. The existence of such a principle strikes me as

debatable.43 More fundamentally, the question arises, and is not answered by French, as to why

a corporation owes moral duties to individuals, such as a duty not to harm them in particular

ways, at all.

Similarly, in French’s account of the responsibility of Air New Zealand, the relevant choices of

the airline were not described in neutral terms as elements of the causal chain leading to the

disaster, but --- borrowing the commission of inquiry’s label --- as organizational “defects”

which the airline had an obligation to remedy once it was made aware of them.44

What distinguishes a defect from an innocent causative act is not explained by French. Yet, there

must be a distinction. For not every harm caused by someone can give rise to an obligation of

that person to prevent its recurrence. Businesspeople knowingly and repeatedly harm

competitors in the marketplace.45 I harm a law student by giving him a poor grade on an

assignment, and I repeat this harm, but I do not wrong him, if I give him a poor grade on the

next assignment, too. What makes a harm wrongful is not that it is intentional or that it is

knowingly repeated, but that it violates a duty owed by one person to another. In the absence of

a duty, there is no wrong in causing or in knowingly repeating a harm.

At first glance, Pettit’s account does not suffer from this gap. He combines an undemanding

conception of intentionality (such that even robots possess it46) with the notion that

corporations (unlike robots) encounter and must dispose of competing “propositions of value.” 42 See note 34 above. 43 The Concorde was known to be less safe than its subsonic competitors. Was it immoral to operate flights on the Concorde? 44 See note 38 above. 45 Ripstein (n 29) 217. 46 Pettit (n 17) 178.

11

The capacity to make value judgments supplies a basis on which the resulting act can be

wrongful, in the sense of deserving blame.

However, like French, Pettit constructs corporate responsibility --- the responsibility of the

corporation as an entity --- so that it is detached from any responsibility of its members in

respect of the act.47 By hypothesis, it is the entity, and not necessarily any individual, that acted,

had the capacity to consider value judgments, and so on. The problem with this approach is that

it is not clear why such an entity is a member of our moral community. A corporation, in Pettit’s

understanding, differs from a robot only in that it makes value judgments. I am not certain that

it is fair to assume that anyone or anything that makes value judgments is a member of our

moral community, such that moral rights and duties subsist between us and them the violation

of which would be the kind of “bad value judgment” that rightly attracts blame.

The corporate moral agency approach seems to be most useful if we are prepared to assume

that certain acts (such as endangering life) are morally wrong regardless of the type of

intentional actor, but on this assumption we would also pass judgment on robots and on non-

human animals intentionally committing the same acts. If we instead assume that wrongdoing is

relational, and conceptualize it as a violation of duties owed by one member of a moral

community to the other members, then the group agency approach seems incomplete insofar as

it takes for granted that moral duties subsist between group agents and human beings. The

approach suffers, it would seem, from anthropomorphism.

Neoclassical perspective. At the same time, I do not wish to treat the neoclassical argument

against corporate criminal liability as conclusive against the desirability of this institution. In

part, this is because of reservations I have expressed elsewhere48 about the neoclassical

framework as an exclusive source of insight into the nature of the corporation. In addition,

criminal liability may have preference-shifting or norms-signalling functions,49 which the

neoclassical analysis ignores. In Part 5, I shall suggest that these functions provide a rationale

for corporate criminal liability.

47 See also Harding (n 17) 82 (on this model of responsibility, there is a “metaphysical shift … [implying] a reality for the corporate actor, which is distinct from, or ‘over and above’, that of the human members of the corporate organization. … [R]esponsibility … vest[s] in the collective or organizational agent, and not in any associated individuals.”) 48 Ian B Lee, ‘Implications of Sen’s Concept of Commitment for the Economic Understanding of the Corporation’ (2008) 21 Can J L & Jur 97 [hereinafter Lee, ‘Implications’]; Ian B Lee, ‘Citizenship and the Corporation’ (2009) 34 L & Soc Inq 129. 49 See., e.g., Kenneth Dau-Schmidt, ‘An Economic Analysis of the Criminal Law as a Preference-Shaping Policy’ [1990] Duke LJ 1. Insofar as corporate liability is concerned, I reach the opposite conclusion from Dau-Schmidt: he argues that, given that corporations do not have preferences to be shaped, criminal liability should be imposed only on individuals. I shall suggest that despite the fact that corporations do not have preferences, corporate criminal liability may be of use in shaping individuals’ preferences.

12

3. CORPORATE THEORY AND CORPORATE CRIMINAL LIABILITY

3.1. RELEVANCE OF CORPORATE THEORY

The divergent conclusions of the neoclassical and group moral agency perspectives do not rest

solely upon different understandings of the purpose of the criminal law. They also rest on

different ontologies of the corporation. One perspective treats the corporation as a framework

for interaction among agents, all of whom are individuals. The other considers the corporation

to be an agent in its own right.

More generally, arguments about the nature and purpose of corporate criminal liability typically

contain embedded theoretical understandings of the corporation.50 For instance, the claim that

corporate criminal liability unfairly punishes innocent shareholders51 appears to be premised

on what corporate theorists refer to as the “aggregate theory” of corporate personality, namely

that the corporation is the alter ego of its shareholders. It is on that basis that punishment of the

former is equated with punishment of the latter.52

A slightly different account of the alleged unfairness to shareholders might draw on the notion

that the shareholders have ceded control over “their” business to professional managers over

whom they have little practical oversight, with the result that corporate criminal liability results

in shareholders’ paying the price for the managers’ wrongdoing. The corresponding theory of

the corporation --- that it is a business under the control of managers whose discretion is more

or less unconstrained --- is that reflected in the work of Berle and Means.53 Today, similar

(though more optimistic) theories of the corporation are advanced by Stephen Bainbridge and

by Margaret Blair and Lynn Stout.54

50 They also, of course, assume a particular understanding of the criminal law. See, regarding the implications of such assumptions of the debate about corporate criminal liability, Nicola Lacey, ‘“Philosophical Foundations of the Common Law”: Social, Not Metaphysical,’ in J Horder (ed), Oxford Essays in Jurisprudence (2000), esp. at p. 27. 51 J Hasnas, ‘The Centenary of a Mistake: One Hundred Years of Corporate Criminal Liability’ (2009) 46 Am Crim L Rev 1329, 1357 (“corporate criminal liability … visits punishment on the innocent shareholder who had no control over the conduct of the [offending corporate] employee” and therefore is incompatible with a “liberal system of criminal justice”). 52 Without the aggregate theory, the effect of corporate punishment on shareholders would instead be akin to “collateral damage,” as when the incarceration of a family’s sole wage-earner causes harm to his or her dependents. For further discussion, see Part 5. 53 AA Berle, Jr and G Means, The Modern Corporation and Private Property (Macmillan 1932). 54 Stephen M Bainbridge, ‘Director Primacy: The Means and Ends of Corporate Governance’ (2003) 97 NW U L Rev 547; Margaret M Blair & Lynn A Stout, ‘A Team Production Theory of Corporate Law’ (1999) 85 Va L Rev 247, 250-251; Margaret M Blair & Lynn A Stout, ‘Director Accountability and the Mediating Role of the Corporate Board’ (2001) 79 Wash ULQ 403, 404-405 [hereinafter Blair & Stout, ‘Director Accountability’].

13

To take another example, the view that corporate criminal liability is a quid pro quo for the

benefits of legal personality55 seems to take as its starting point the so-called “concession

theory,” according to legal personality is a privilege granted by the state. It is because legal

personality is a privilege, rather than, say, a reflection of the real nature of business

organizations, that those who choose to avail themselves of it must take the good with the bad.

Like its more orthodox alternatives, the theory of corporate criminal liability to be developed in

Parts 4 and 5 rests upon a particular understanding of the corporation, the basic features of

which I propose to make explicit in the section that follows.

3.2. THE CORPORATION AS A TEAM

My theory of the nature and purpose of the institution of corporate criminal liability rests on a

conception of corporations as teams. In this section, I explain this conception of the corporation

and how it compares with the prevailing economic understanding of corporation.

3.2.1. THE FIRM AS A SET OF RELATIONSHIPS

A conception of the corporation is an answer to the question, “What is a corporation?” The

question is analogous to that sometimes posed by professors to first-year property law

students: “What is property?” We teach our students that property is not a thing, but a bundle of

rights; the same is true of corporations. A corporation is not a thing, but a set of legal

consequences produced by corporate law and the corporation’s constitutive documents.56

In the economic analysis of corporate law, this insight is captured by the metaphor of the

corporation as a nexus of contracts. This metaphor calls attention to the fact that the law of

corporations and the incorporation instruments of individual corporations establish a set of

legal relationships. In terming these relationships “contracts,” the metaphor highlights their

voluntariness and asserts their character as mutual exchanges of value.

3.2.2. MARKETS VERSUS FIAT

Orthodox economic thinking about the firm (of which corporations are one type) goes beyond

the observation that they are voluntary exchange relationships. It also includes a standard

account of the reason for the existence of firms.

According to this account --- one of law and economics’ many debts to Ronald Coase --- firms

exist because of transaction costs.57 If transaction costs were zero, there would be no need for

55 Quaid (n 16) 113-14. 56 By “constitutive documents,” I mean the corporation’s articles of incorporation or association. 57 Ronald Coase, “The Nature of the Firm” (1937) 4 Economica 386.

14

firms and productive activity would instead be coordinated entirely through spot markets. But

because transactions are not costless, resources and control over them are centralized within

firms. Within the firm, a person or a small group exercises fiat.

Economists are divided about how important a phenomenon fiat is — which is a way of saying

that they are divided about how significant transaction costs are, empirically speaking. This is,

for instance, how one may interpret Alchian and Demsetz’s disagreement with Coase.58 The

former famously disputed whether employers wield any real “authority” over employees, or

whether both simply have the right to discontinue their relationship with one another, such that

the employer-employee relationship is more aptly analogized to that between a consumer and a

greengrocer.59 Alchian and Demsetz may be understood to be minimizing the empirical

significance of transaction costs in the employment market.

Similarly, Blair and Stout’s conception of corporate boards of directors as “mediating hierarchs”

depends upon their argument that market forces leave boards with substantial discretion,60

whereas their critics complain that Blair and Stout have underestimated the force of these

markets, particularly the market for corporate control, and consequently overstated the extent

of directorial power.61

Despite these differences, the basic idea of corporations as islands of discretionary power that

are either big or small --- depending on the magnitude of transaction costs in the market that

surrounds them --- is common ground among economic theorists of the firm.

3.2.3. TEAMWORK IN THE STANDARD MODEL

It is equally common among economic theorists to describe firms as vehicles for the

coordination of team production. For instance, Easterbrook & Fischel write:

“Much production is performed in teams. Teams of employees sweep the floor, teams of

engineers design new products, teams of managers decide whether and where to build

new plants. So long as no monitor can determine what each member's marginal

contribution to the team's output is, each member will be a less than perfect

representative of the interests of the team as a whole. Unless one person receives all the

rewards of success and penalties of failure, his incentives are not properly aligned with

those of the venture as a whole. "Let George do it" is a predictable response when any

58 Armen A Alchian & Harold Demsetz, ‘Production, Information Costs, and Economic Organization’ (1972) 62 Am Econ Rev 777. 59 Id 777. 60 See similarly Bainbridge (n 53). 61 E.g., John C Coates, IV, ‘Measuring the Domain of Mediating Hierarchy: How Contestable are U.S. Public Corporations?’ (1999) 24 J Corp L 837.

15

given employee gets some of the benefits of George's hard work and does not get all of

the benefits of his own hard work.62

As can be seen from the quoted passage, the defining characteristic of teamwork is that while

success depends upon the efforts of more than one person, the contribution of any one person

to the success of the team is difficult to measure.63 According to the conventional contractarian

wisdom, market forces lead to the emergence of various monitoring and bonding devices

designed to combat the divergence of interests between the individual members of the team and

the team as a whole.64 In addition, it can be the role of the central authority within the firm to

attempt to monitor and reward accordingly the members of the team. These efforts are bound

to be imperfect and, of course, they entail costs of their own; ultimately, the residual costs

associated with team production are a cost of the firm, to be weighed against the transaction

costs associated with spot markets.

The conventional understanding just described employs a well-known behavioural model: it

assumes that individuals maximize their utility. Private utility calculations drive the decisions of

individuals to shirk on teamwork, to commit resources to the discretion of centralized decision-

makers within firms, to demand and provide monitoring and bonding devices, and to cooperate

outside firms to the extent that the cost of all of the foregoing exceeds the cost of market

transactions.

I do not take issue with the usefulness of the traditional behavioural assumption in accounting

for many features of firms (for instance, the existence of monitoring and bonding devices, and

the observed scale and scope of firms). However, a model is by definition a simplification,

highlighting one aspect of the phenomenon and deemphasizing others. Different models may

therefore be suited to different applications. In particular, for purposes of my argument about

corporate criminal responsibility, I employ a behavioural model that accommodates individual

action upon principles other than the maximization of utility. Let me now turn to an explanation

of this model.

3.2.4. TEAMWORK AND COLLECTIVELY RATIONAL PRINCIPLES

Teamwork, I would suggest, is a context in which individual behaviour is not driven solely by

utility maximization. Instead, the identification of individuals as the members of a team leads

them to adopt collectively rational principles as principles of action.

62 Easterbrook & Fischel (n 8) 9-10. 63 Alchian & Demsetz (n 56). 64 Michael C Jensen and William H Meckling, ‘Theory of the firm: Managerial behaviour, agency costs and ownership structure’ (1976) 3 J Fin Econ 305.

16

I am drawing here upon Amartya Sens critique of the concept of “preference.”65 In particular,

Sen objected to the conventional use of the latter concept to refer both to an individual’s

interests and to that which motivates an individual’s actions. The conventional usage, he argued,

is unhelpfully ambiguous in situations where an individual acts in accordance with a principle,

even at some cost to his or her utility.66

For instance, as Elizabeth Anderson has explained,67 an alternative framework is needed to

comprehend the problem of cooperation in prisoner’s dilemmas. It is conventional to think of

the decision facing the players in a prisoner’s dilemma (PD) as presenting a conflict between

their individual welfare and the welfare of the group. In such a conflict it is difficult to

understand why the group’s welfare would ever prevail --- it scarcely deserves to be called a

dilemma.68 Anderson suggested that, in order to view a prisoner’s dilemma as a true dilemma,

we need to think of it instead as presenting, for each player, a confrontation between two types

of rationality: (1) the pursuit by the player of her own goals, and (2) action in accordance with

collectively rational principles of the group formed by herself and the other players, with which

she identifies. It is rational to pursue one’s goals but it can also be viewed as rational to suspend

the pursuit of one’s goals in deference to collectively rational principles, where one regards

one’s self as part of the group.69

The latter type of rationality is at work in team settings, including firms. As previously noted,

economic theorists emphasize the importance of monitoring and bonding devices to incentivize

effort in team production settings. However, one normally wants team members to devote more

than the bare minimum of effort that will prevent their incurring contractual penalties or

dismissal from the team. To accomplish this goal, strategies beyond bonding and monitoring are

65 Amartya Sen, ‘Rational Fools: A Critique of the Behavioral Foundations of Economic Theory’ (1977) 6 Phil & Pub Aff 317. 66 Id. 67 Elizabeth Anderson, ‘Unstrapping the Straitjacket of “Preference”: A Comment on Amartya Sen's Contributions to Philosophy and Economics’ (2001) 17 Econ & Phil 21. 68 The conventional behavioural model predicts cooperation in apparent prisoner’s dilemmas only through devices that change the payoffs so that they no longer have the structure of a prisoner’s dilemma. For instance, one strategy is to suppose multiple periods, such that the individual payoff from cooperation (taking into account payoffs in subsequent periods) exceeds the individual payoff from defection. Another strategy is to assume that the prisoners’ care about one another’s well-being; this strategy, too, predicts cooperation at the cost of destroying the PD structure of the choice situation. See Gerald Postema, "Morality in the First Person Plural" (1995) 14 Law & Phil. 35, 38, 42-45. 69 Lee, ‘Implications.’ A recent illustration is provided by the decision of the members of a small group of technicians to remain at the Fukushima Daiichi Nuclear Power Station to try to prevent a meltdown. A senior American technician, describing the “consensus” within his profession, said “[Y]ou’re certainly worried about the health and safety of your family, but you have an obligation to stay at the facility…. There is a sense of loyalty and camaraderie when you’ve trained with guys, you’ve done shifts with them for years.” Bradsher and Tabuchi, ‘Last Defense at Troubled Reactors: 50 Japanese Workers’ New York Times (15 March 2011).

17

necessary, such as the cultivation of a sense of mutual loyalty.70 This is why Sen argued that

commitment to an organization is “central to the problem of work motivation…. To run an

organization entirely on incentives to personal gain is pretty much a hopeless task.”71 Successful

production requires, in other words, that the participants be motivated (at least in part) by

loyalty towards the team. It requires that they identify as the members of a team and that they

regard doing their part towards the collective goals of the team as a principle of action.72

3.2.5. THE CORPORATE TEAM

In corporate law, one manifestation of the understanding of the firm as a team is the mandate of

the directors and officers to exercise their powers with a view to the “best interests of the

corporation.” The traditional view is that this duty is a device to control agency costs by

attaching a penalty to the unilateral taking of assets belonging to the firm.73 But if the point is to

deter misappropriation, why does the law impose an affirmative duty --- and a highly open-

textured one at that --- rather than simply prohibiting misappropriation?

I prefer a more straightforward interpretation: the duty reflects the understanding of the

participants in the corporation that the members of the board and the corporation’s officers are

to act in the best interests of the corporation --- that is, upon the collectively rational principles

inherent in the venture.74

It would seem, therefore, that the members of the corporate team include, in principle, its

directors and officers. From the previous section, the team also appears to include some

proportion of its employees.

The situation of the shareholders is surprisingly ambiguous. In principle, shareholders could

regard themselves as members of a collective entity: after all, they are the “citizens” of the

70 Ian B Lee, ‘Efficiency and Ethics in the Debate about Shareholder Primacy’ (2006) 31 Del J Corp L 533, 550. 71 Sen (n 64) 333-35. 72 I am obviously not suggesting that team members do not shirk. Collective rationality is not the exclusive principle of action; it competes with individual rationality. Cooperation in prisoner’s dilemmas is not a given; otherwise they would not be dilemmas. My point is only that team members do not act exclusively on the basis of individual rationality. Doing their part for the team is, for them, also a principle of action. 73 Easterbrook & Fischel (n 8) 92 (“the fiduciary principle … replaces prior supervision with deterrence, much as criminal law uses penalties for bank robbery rather than pat-down searches of everyone entering banks”). 74 I can be agnostic, for present purposes, as to whether these principles are summed up in the concept of shareholder primacy or a broader concept. The latter position is exemplified by Blair and Stout, who argue that directors’ behaviour is influenced by their perceptions of belonging to the same “group” as employees, managers and shareholders: see ‘Director Accountability’ (n 53) 441-42. The former position is exemplified by Bainbridge, ‘In Defense of the Shareholder Wealth-Maximization Norm’ (1993) 50 Wash & Lee L Rev 1423, 1441 (“the shareholder wealth maximization norm is central to management’s socialization”).

18

corporate body politic, possessing collectively the power to select the corporation’s board of

directors.75 Moreover, the premise of many shareholder proposals on corporate social

responsibility matters is that the proponents believe themselves to be morally implicated by

policies and practices undertaken in the name of the corporation.76 The proponents do not see

themselves as bystanders, but as participants in the corporation’s collective enterprise.

However, many shareholders do not view themselves as the members of a collective entity, but

instead, consistently with their portrayal under conventional contractarian analysis, as the

separate suppliers of an input.77 Some shareholders do not view themselves even in these terms,

but instead consider themselves to be mere speculators on security price movements.78

Shareholders of either of these types might well be expected to act, in relation to their

investment, on the basis of the own-utility-maximization principle (or even the own-wealth-

maximization principle) rather than on the basis of principles that would be collectively rational

for all shareholders, let alone for all participants in the venture. So one cannot say that

shareholders, in general, are the members of the corporate team. The most one can say is that

some shareholders in some corporations identify as team-members and adopt collectively

rational principles as a principle of action.

In summary, corporate theory plays an important role in the theorization of the nature and

purpose of corporate criminal liability. The theory to be developed in the following parts of the

article rests on a conception of the corporation as a legal framework within which team

production occurs. To some extent, this conception is consistent with the orthodox conception.

However, for purposes of my argument about corporate criminal responsibility, I emphasize a

particular characteristic of team production --- namely the importance of group identification

and the consequent adoption of collectively rational principles by the members of the team.79

75 See, e.g., Delaware General Corporation Law § 211(b). 76 See Ian B Lee, ‘Corporate Law, Profit Maximization and the “Responsible” Shareholder’ (2005) 10 Stan JL Bus & Fin 31, 64. 77 See Usha Rodrigues, ‘The Seductive Comparison of Shareholder and Civic Democracy’ (2006) 63 Wash & Lee L Rev 1389, 1399. 78 See Mitchell (n 10) 134. 79 The obvious question that arises in relation to this framework: what is the (analytical) benefit of conceptualizing principled action as distinct from utility maximization? The alternative is, of course, to incorporate values such as “cooperating” or “doing the right thing” into individual utility functions. In other writing, I argue for the value of the dual rationality framework ([reference omitted]). However, for purposes of this article, the incorporation approach poses no threat to my argument about corporate criminal responsibility so long as incorporation is done in such a manner that utility maximization mimics the features of principled action which I have described: identification with a group leads one sometimes to act upon collectively rational principles, at some cost to one’s personal goals --- or rather, if one is an incorporationist, to one’s other personal goals

19

4. TEAM-MEMBER RESPONSIBILITY

The thesis of this article is that corporate criminal responsibility is a shared responsibility of the

members of the incorporated enterprise for wrongdoing by one or more of the other members.

This claim rests on the more general claim that, under certain conditions, the members of a

team bear shared responsibility for wrongdoing by their team-mates. It is the aim of this part of

the article to defend the latter claim.

4.1. COLLECTIVITIES AS TEAMS

Sometimes a group is more than the sum of its constituent individuals. What differentiates a

collectivity from a mere collection of individuals? I want to distinguish between two approaches

to answering this question.

A first approach, followed by group agency theorists such as French and Pettit, is to seek in

certain groups the characteristics of individual agents. For both of these authors, a key element

is an “identity” that remains stable despite changes in group membership: that is, a statement

referring to the group has the same meaning before and after a change in its membership.80

Another crucial distinguishing characteristic of a collectivity, under this approach, is an

institutional framework for making decisions.81

This article defends a different manner of thinking about what distinguishes collectivities from

mere collections of individuals. I suggest that we think of a collectivity as being constituted and

maintained by the self-identification of its members with the group. In this definition, the key

concept is neither the group’s identity nor its institutional features but the fact that its members

regard themselves as the members of a collectivity. Collectivities are not quasi-persons, but

teams. As I explained in Part 3, a team exists when its members regard themselves as the

members of a team and adopt collectively rational principles as principles of action.82

As I shall further explain below (Section 4.3), I do not deny that a team can have a stable identity

despite changes in membership, but this is an implication of the definition of a team rather than

a criterion for a group to constitute a team. I also readily acknowledge that group norms often

arise within teams, but I regard such norms as a product of the team members’ joint effort ---

80 French, C&CR (n 17) 13. Pettit (n 17) 172. 81 E.g., Pettit (n 17) 172; French, C&CR (n 17) 13-14. French’s definition includes additional elements, notably the existence of “enforced standards of conduct” applicable to group members (i.e., group norms). 82 This conceptualization of a team may be analogized to Michael Bratman's understanding of shared intention as consisting in a collection of individual psychological attitudes and their interrelationships, rather than in the state of mind of "some fused agent". Michael Bratman, "Shared Intention" (1993) 104 Ethics 102, 107-108. See also Kutz (n 4) 112 ("groups are nothing more or less than agents who intend to participate in collective action").

20

therefore, as a consequence of the establishment of a team --- rather than as a definitional

requirement. All that is required in order to differentiate a collectivity from a group of isolated

individuals is that each of its members identify as a member of a collectivity and, accordingly,

regard doing their part towards the achievement of the goals of the group as a principle of

action.

4.2. NATURE OF TEAM-MEMBER RESPONSIBILITY

Team-member responsibility for wrongdoing committed by another member of the team arises

under two conditions. First, the team’s norms were a motivating factor in the commission of the

wrongdoing. Second, the other members share in the wrongdoing to the extent that they do

their part in the pursuit of the team’s goals.83

Underlying these conditions are two claims. First, when, as members of a team, we do our part

towards a collective goal, we are entitled to share in the credit (and we share in the discredit)

for achievements (failures) resulting from our combined actions. Second, the team’s norms and

the behaviour motivated by them are a positive or negative achievement of the team.

The first claim might seem counterintuitive to some readers, for it assumes a principle of moral

responsibility of each member for the results of the team’s combined actions, rather than only

for the results caused (in a but-for sense) by the member’s individual action. However, I contend

that this principle is implied by the definition of a team and of teamwork.84

To belong to a team is to consider doing one’s part towards a common goal, regardless of

whether one’s action considered in isolation will be pivotal, as a principle of action.85 In

consequence, a team member evaluates her own conduct relative to whether she did her part

and to the achievements of the group resulting from their collective actions, and not only

relative to whether what she individually did produced an achievement. We apply similar

judgments to the participation of others in teams.86

83 The responsibility of a team-member who makes a direct contribution to the commission of the wrongdoing, for instance by knowingly providing assistance to the wrongdoer, does not come within the concept of team-member responsibility. The responsibility of the helper in such circumstances would be analogous to that of an accomplice, and there is nothing about it that depends upon the helper’s status as a team-member. The theory being developed here is of a kind of responsibility that team-members incur by virtue of being the members of a team. 84 Kutz similarly argues that "intend[ing] to do one's part" in a collective action is an adequate basis for individual accountability, regardless of the individual's "particular causal contribution." See Kutz (n 4) 141, 144. 85 See Section 3.2.4. 86 Thomas Nagel, ‘Moral Luck’ in Proceedings of the Aristotelian Society (1976) suppl. vol. I (in making moral judgments about the actions of others, we apply the same principles of moral agency that we use in evaluating our own actions).

21

Let me illustrate this understanding of moral agency in team settings by way of three

hypothetical scenarios.

(1) Suppose that a team of three individuals (A, B, C) seek to achieve a common goal G that

will be realized if two of them to do X. If each individual identifies as a member of the

team, each will take “doing my part towards G” as a principle of action, and each will do

X (barring other reasons not to). G will be realized, and each of A, B and C will share in

the credit for G even though none of their individual actions was the but-for cause of G.

Each will consider: “I did my part as a member of the team, and together we realized G.”

(2) Now suppose a team of six individuals (A, B, C, D, E, F) who seek to achieve common goal

G that will be realized if five of them do X. Suppose that each individual identifies as a

member of the team; each takes “doing my part towards G” as a principle of action.

However, assume that only A, B and C do X, whereas D, E and F fail to do X (because of

weakness of will, self-interest, or some other factor). G will not be realized. By analogy to

(1), D, E and F share in the failure, even though for none of them was his or her inaction

the but-for cause of not-G. There is a sense in which A, B and C do not share in the failure

(for they did their part). Importantly, however, there is also a sense in which they do

share in the team’s failure --- for although they have the individual satisfaction of having

done their part, they do not have the satisfaction of having done their part in a

successful combined action. This is what they mean when they think to themselves, “at

least I did my part.” Yes, they did their part, but no, they did not succeed.

Notice that in both cases, each of A, B and C judges himself or herself as participating in a

success or non-success even though the outcome was the consequence of factors beyond his or

her individual control (the actions of other team members).

(3) Now return to the team of three members. Suppose that C, while doing his part towards

G, commits a wrong against a non-member of the team (V). Do A and B share in the

wrong, or is the wrong C’s alone? My suggestion is that A and B share in the wrong if it is

in part the product of the team’s combined action, which it will be if it is influenced by

the group’s norms. Thus, the combined action of A, B and C towards G produces (among

other team achievements) a set of norms GN. If these norms in turn motivate C to

commit a wrong, then the wrong is a negative achievement of the team.

In each of these three scenarios, individual but-for causation is non-decisive in assessments of

member responsibility for team achievements.

22

In team settings, individual moral agency is better understood in terms of “doing one’s part”

than in terms of being the but-for cause of the outcome. This understanding applies to both the

positive and the negative achievements of the team and, in particular, explains the sense in

which the members of a team share responsibility for wrongdoing committed by another

member, when such wrongdoing can be considered to be a product of the team’s norms and,

therefore, is a consequence of the team-members’ joint effort.

4.3. CROSS-TEMPORAL TEAMS

As previously noted, one feature of group agency theories is that the collective entity persists

even if the identity of the members changes. Is this also true of a team? I would suggest that it is.

Let me begin by recalling the definition of a team. A team is constituted by the self-identification

of the members as the members of a team. Each of them takes doing their part towards the

common goals of the team as a principle of action. This definition does not require that all of the

members be in existence at the same time.

In particular, the appropriate moral assessments in scenarios (1), (2) and (3) from the previous

section do not change if we introduce temporal separation among the members.

(4) Consider a team consisting of A, B, C, D and E who seek to achieve a common goal G that

requires that, in each of three periods, any two members do X. Assume that A, B and C

are alive at T1; B, C and D are alive at T2; and C, D and E are alive at T3. At each moment,

the then-living members of the team take “doing my part towards G” as a principle of

action, and each will do X (barring other reasons not to). This scenario is not materially

different from Scenario (1). G will be realized after T3, and each of the members A

through E will share in the credit for G.

(5) Now suppose that at T1, C, while doing his part towards G, commits a wrong against a

non-member of the team (V). By analogy with Scenario (3), A, B and D share in the

wrong if it is in part the product of the group’s norms. There is a sense in which E does

not share in the wrong (for she did not participate in the generation of the group’s

norms as they stood at T1). Yet, by analogy with Scenario (2), there is a sense in which

she does: for although E has the satisfaction of having done her part and also of not

having contributed causally to the wrong to V, she does not have the satisfaction of

having done her part in an unblemished combined action. The wrong to V is a negative

achievement of the team to which E also belongs.

It may be objected that there is not one team consisting of A through E, but two, or perhaps

three different teams: there is a team consisting of A, B and C; another consisting of B, C and D;

23

and a third consisting of C, D and E. In particular, it may be argued that the wrong to V in

Scenario (5) is a negative achievement of the second, and perhaps the first team, but not of the

third team. E does not share in either the achievements or the failures of the first two teams.

But it is not clear why the members of the team must be simultaneously in existence, any more

than they must be in the same place, or that they must know one another’s identity. On the

definition of the team developed in this article, all that is necessary is that they each take doing

their part towards their common goal as a principle of action.

There is, in fact, something counterintuitive about a requirement of simultaneous membership.

We do not think of a nation, for instance, as thousands of separate entities each existing at a

separate moment in time, and only for a moment, until its membership changes, upon which a

new entity is born. Rather, we think of it as a single entity with a past, present and future.87 The

entity is a collectivity, a team: it consists of people, living at different times, each doing their part

towards their common goals. To the extent that the team exists over time, its members include

people who may not simultaneously be in existence.

It follows that team-member responsibility is compatible with intergenerational responsibility.

Later generations are not moral strangers to the wrongs committed by earlier generations

within the same team. Clearly, later generations are not causally responsible for the wrongs. We

have seen, however, that moral agency in a team setting is not governed by the principle of

individual but-for causation, but of participation in the team’s positive or negative achievements

through doing one’s part. Later generations thus participate in the positive or negative

achievements of their teams; these include achievements that arose through the actions of

members of the team not still in existence.

Some may be uncomfortable with the notion of intergenerational responsibility of the members

of a community. It is important to remember, however, that membership in the team is not only

voluntary; it is also subjective. Teams are constituted by the choice of each member to regard

herself as the member of a team and to adopt collectively rational principles as principles of

action for herself. An individual excludes herself from a team simply by not regarding herself as

a member.

Therefore, in order to participate in responsibility for the wrongs committed by a team-member

living in a previous generation, a team-member living today must regard herself as a member of

a team that includes the earlier generation. This is way of understanding the dilemma

87 An analogy may be made to intrapersonal unity across time. See Postema (n 68) 55-57.

24

experienced by some members of a post-war generation in respect of the crimes committed by

their parents’ generation. Bernhard Schlink described this dilemma as follows:

“[Members of the postwar generation must cope with] learning that the beloved parents,

or admired teachers and professors, or respected pastors and other figures of love and

authority had a dark past, had been involved in crimes. How do you deal with that? You

realize that in a way you are entangled in the guilt of those whom you love or keep

solidarity with. … What should I do when I learn that someone I love is guilty? Do I have

to break with this person? Can I keep this person in my life? These questions are

universal.”88

The children of the wartime generation have no causal responsibility for what their parents did

--- they did not even contribute to the norms that influenced their parents’ behaviour.89 Their

self-interrogation is not for what they themselves did as individuals or as a generation. Instead,

it relates to a forward-looking moral question for the members of the postwar generation: “Are

we right to continue in solidarity with the nation and the intertemporal community which we

share with our parents?”

More generally, the negative achievements of one’s team, including past negative achievements,

may well be a reason why one should consider leaving a team90 or not joining it in the first

place, just as its positive achievements are a reason for entering into or continuing in

membership.

5. FROM TEAM-MEMBER RESPONSIBILITY TO CORPORATE

CRIMINAL LIABILITY

In Parts 3 and 4, I argued that corporations and other firms are teams and that, under certain

relatively undemanding conditions, the members of a team bear shared responsibility for

wrongdoing by their team-mates. It remains to be explained how team-member responsibility

can illuminate the institution of corporate criminal liability.

88 Interview remarks in Cardozo Life [2009:1] 36. Schlink further describes the concept of intergenerational guilt via solidarity in B Schlink, Guilt About the Past (Anansi 2009) 18-22. 89 See, e.g., M Slackman, ‘With Film, Afghan-German is a Foreigner at Home’ New York Times (17 Oct. 2010) (describing a first-generation German-born filmmaker of Afghan origin who, regarding himself as German, “professes to carry the guilt of Germany’s Nazi past”). 90 A corollary of the foregoing discussion is that even a member who dissented from the group norms that contributed to the wrongdoing nonetheless experiences the wrongdoing as a negative achievement of the team, in which she shares, at least if her dissent did not cause her to leave the team. For instance, a vocal pacifist may feel ashamed of her country's participation in a war.

25

As stated at the outset of Part 2, a theory of corporate criminal liability is a theory of the nature

and purpose of the application of criminal punishment to corporate entities. By punishment, I

mean the imposition of negative consequences (such as a requirement to pay a sum of money)

accompanied by condemnation.91 Punishing a team consists in delivering a setback to the

group’s goals together with the punisher’s condemnation.

Why should there be liability of the corporate entity for crimes committed by members of the

corporation in the pursuit of the corporation’s goals? To answer this question, I first need to

articulate an assumption about the purpose of criminal liability, in general. It will then be

possible to explain why that purpose is advanced by the institution of entity liability.

5.1. NORMS-SIGNALLING AND THE CRIMINAL LAW

In Part 2, I suggested that the criminal law seeks to affect behaviour, not only through pricing,

but also through norms-signalling.92 The behavioural assumption underlying this suggestion is

that people’s actions are regulated not only by legal incentives, but by moral and social norms.

The term “moral norm” refers to the behavioral influence of a person’s internalized belief that

she ought to act in a particular way, whereas “social norm” refers to the desire to receive the

esteem of others and avoid their opprobrium.93

It is conventional to assume that prohibitory law carries “moral weight” with citizens.94 This

may be because public officials (such as legislators) are believed to have a good sense of what

does, or does not attract the community’s respect or opprobrium and to be responsive to the

community’s opinion.95 That is, officials are believed to be reliable sources of information about

social norms. It may also be that citizens give weight to society’s opinion in forming their moral

commitments. If so, then the state’s designation of certain conduct as discreditable can be

expected to shape both social norms and the moral norms followed by individuals.

This is one way of understanding how criminal liability is different from civil, regulatory or tax

liability. Criminal liability does merely visit a consequence upon its addressee; it labels that

91 See Grant Lamond, ‘What is a Crime?’ (2007) 27 OJLS 609, 621 (“[p]unishment is … the deliberate imposition of a burdensome liability on an individual for some blameworthy conduct in order to censure that conduct.”) 92 Just as a theory of corporate criminal liability depends upon a theory of the corporation (Part 3), so too does it depend upon a theory of the criminal law --- in this case, a instrumental expressivist understanding of the criminal law. 93 For a similar distinction between moral and social norms, see also Einer Elhauge, ‘Sacrificing Corporate Profits in the Public Interest’ (2005) 80 NYU L Rev 733. 94 Cass Sunstein, ‘On the Expressive Function of Law’ (1995) 144 U Pa L Rev 2021, 2031; see also Dan M Kahan (n 15); Jean Hampton, ‘The Moral Education Theory of Punishment’(1984) 13 Phil & Pub Aff 208; Dau-Schmidt (n 49). 95 Robert Ellickson, ‘The Market for Social Norms’ (2001) 3 Am L & Econ Rev 1, 40.

26

consequence a punishment. Through the magnitude and type of consequence, and the label

attached to it, the state identifies the triggering action as a wrong, in order to adjust individuals’

preferences (via their moral norms),96 and to signal that particular conduct is deserving of social

opprobrium.

5.2. THE PURPOSE OF PUNISHING A CORPORATION

When the wrongdoer is a member of a corporate entity, it is appropriate to punish the entity in

addition to the direct wrongdoer.

Condemnation of the team draws attention to the contributory role that the team’s norms

played in producing the wrongdoing and to the responsibility of the members in relation to the

content of those norms. For the state to condemn the direct wrongdoer alone is to implicitly,

and mistakenly absolve the team. The members of the team bear responsibility in respect of the

wrongdoing, which (I argued in Part 4) is a negative achievement of the team insofar as it

results from the group’s internal norms.

Compare, for instance, two auditing firms. In one, the prevailing ethos is: “think straight, talk

straight.” In the other, the ethos is: “above all, produce profits for the firm.”97 In each firm, the

relevant slogan has attained the status of a local social norm: it guides the actions of the

members, and the actions of the members reinforce the norm. If, in the second firm, dishonest

but profitable conduct is widespread, including criminal conduct by one or more members of

the firm, it is appropriate to punish the firm because the wrongdoing is not the act of an isolated

individual, but a negative achievement of the firm, in which all of the members share because

each of them did his or her part.

In its paradigmatic form --- a fine --- the criminal punishment of a corporation delivers a

monetary setback to the corporation’s goals and, at the same time, signals that a wrong

occurred that constitutes a negative achievement of the corporate team.98 In drawing the

members’ attention and that of other citizens to the commission of a such a wrong, the state

seeks to influence the behaviour of the members as they participate in the generation of the

96 It is possible to conceptualize moral norms as a component of individual utility function. In Sen’s framework, discussed in Part 3.2, moral norms are meta-preferences, which is to say that they are principles against which individuals measure their own conduct. It is not material to the argument of this Part which conceptualization is adopted. 97 Brown and Dugan, ‘Andersen’s Fall from Grace is a Sad Tale of Greed and Miscues’ Wall Street Journal (7 June 2002). The first slogan was the motto of the founder of Arthur Andersen LLP, recited by incoming junior auditors. Later, an Andersen partner said, “produce profits for the firm” became the dominant value. 98 Punishment of the corporate entity is warranted even if the goals of the corporation are innocent (or indeed praiseworthy), just as individual wrongs are punished even if they are committed in the pursuit of worthy individual goals. Social and moral norms constrain the pursuit even of legitimate goals.

27

team’s norms, to alert the members to the existence of grounds for revisiting their continued

membership in the team, and to advertise the relevant moral and social norms to citizens in

general.

I do not argue for the imposition of punishment directly upon the members of teams (other than

the direct wrongdoer).99 On the contrary, the punishment should take the form of a setback to

the team’s goals rather than the infliction of direct harm upon the members. This ensures

symmetry between the way in which punishment is experienced by the members (participation

in a setback of the team) and the members’ relationship to the wrong (participation in a

negative accomplishment of the team).100 Moreover, membership is subjective: an individual is a

member of a team to the extent that she regards herself as a member of the team.101 In the

corporate context, for instance, it would be difficult, as a practical matter, to identify who among

the shareholders and employees possess the required subjective sense of membership.

Within a norms-signalling theory of the criminal law, the magnitude of the setback is one way of

signaling the gravity of the wrong, but signalling is only one of a number of public objectives

that bear on the appropriate magnitude of the penalty imposed. These objectives may

sometimes be in tension, for instance where the burden of the punishment falls upon innocent

third parties.102 To provide a concrete example in a team context, while the awarding of punitive

damages against Catholic dioceses in sexual abuse cases sends the right message to the

dioceses’ members, it also drastically reduces the resources available for charitable activities

and, in consequence, harms the beneficiaries of these activities.103 I do not claim that norms-

99 Kutz's theory of individual responsibility for collective acts leads him, by contrast, to recommend (in principle) individual shareholder civil liability for corporate torts. See Kutz (n 4) 236-53. He does not discuss corporate criminal responsibility. 100 Because punishment does not take the form of the infliction of harm directly upon the team-members, what is proposed here is to be distinguished from common-purpose liability, i.e., the liability of the participants in a joint criminal enterprise even for unforeseen consequences of that enterprise. R v Powell (Anthony) and English (1999) 1 AC 1 (HL); Gillard v The Queen (2003) 219 CLR 1 (HC Aust). The controversy surrounding the latter doctrine, as well as certain features of the doctrine, such as the requirement that the consequence have been regarded as “possible” by the participants even if it was unforeseen by them, illustrate the role that intentionality plays in deontological justifications of the infliction of criminal punishment upon individuals. As discussed above (nn 27-29), these considerations are not self-evidently relevant in the context of corporate liability, where punishment takes the form of a setback to the team’s goals rather than the coercion of the team-members. 101 See Part 4. 102 For instance, a sole breadwinner who commits a crime may be spared incarceration in order to prevent harm to his or her dependants. 103 See Catharine Pierce-Wells, ‘Churches, Charities and Corrective Justice: Making Churches Pay for the Sins of their Clergy’ (2003) 44 BC L Rev 1201, 1202.

28

signalling should prevail over other considerations relevant to the appropriate quantum of

punishment; only that it is one such consideration.104

5.3. REGULATORY OFFENCES

In recent years, many scholars have suggested that the distinction between criminal and civil

liability is becoming blurred in practice.105 In particular, regulatory law seems to defy clear

classification.106 Moreover, in the normative evaluation of regulatory law, utilitarian and moral

judgments sometimes overlap, and sometimes pull in opposite directions. Some may argue, for

instance, that regulatory offences are not true wrongs,107 and that the widespread recourse to

strict liability in regulatory statutes on functional grounds is morally acceptable only if

incarceration is not an available penalty.108 Others might suggest that the paradigmatic

regulatory offence --- the causation of social harm through an absence of due diligence --- is

entirely deserving of being labelled wrongful. That being so, sanctions for regulatory offences ---

even if they are pecuniary --- can be understood to have a punitive character.

This article does not attempt to resolve these debates. I do not aim, for instance, to defend the

view that negligent harm is or is not wrongful (though I think that it is109). My arguments

concern the conditions under which it is an appropriate response to wrongdoing (however

defined) by an individual to visit punishment upon the corporate entity to which the individual

belongs. To the extent that regulatory “offences” are not perceived as wrongs, and that the

“fines” imposed are actually non-condemnatory incentives for behaviour modification, the

imposition of such “fines” on a corporation would not come within my conception of corporate

criminal liability, because it would not be punitive in the sense in which I use the term.110

However, insofar as regulatory offences are understood as wrongful, my conception of 104 It is sometimes argued that the imposition of a criminal fine upon a corporation is unfair insofar as it harms only the current shareholders, and leaves past members of the team (who may be more blameworthy) unharmed. This objection is, in my view, uncompelling, given that liability risk is merely one of many contingencies that the present shareholders acquire from their predecessors. A recent purchaser of shares also benefits when a speculative research investment made years earlier pays off, and suffers a loss when a long-term contract entered into in the past turns out to have been unwise. In principle, the economic contingencies to which shareholders are exposed are priced by the market. 105 See, e.g., John C Coffee, Jr, ‘Does “Unlawful” Mean “Criminal”?: Reflections on the Disappearing Tort/Crime Distinction in American Law’ (1991) 71 BU L Rev 193; Paul Robinson, ‘The Criminal-Civil Distinction and the Utility of Desert’ (1996) 76 BU L Rev 201, 213; 106 Karen Yeung (n 5) 250 and subsequent pages; Celia Wells, ‘Corporate Manslaughter: Why Does Reform Matter?’ (2006) S Afr LJ 648, 651; Celia Wells, ‘Corporate Crime: Opening the Eyes of the Sentry’ (2010) 30 Leg Stud 370. 107 Yeung (n 5) 252 discusses the “moral ambivalence of regulatory offences.” 108 See, e.g., Brudner, Punishment and Freedom (n 27) 169. 109 Moreover, the deontological constraints that may preclude the imposition of physically coercive sanctions such as incarceration on individuals for negligence offences have no bearing on corporate liability as, obviously, such sanctions are not available. 110 That is not to say that corporate liability would be unjustified, only that it would not have the character of corporate criminal liability.

29

corporate responsibility as team-member responsibility supplies a reason why liability should

be imposed upon the entity, and not only upon the direct individual wrongdoer.

6. CONCLUSION

To date, it seems that theorists of corporate criminal liability have been required to choose

between two unsatisfying narratives. On the one hand, some have tried to understand corporate

criminal punishment in retributive terms on the basis of a strained analogy between

corporations and human members of the moral community. On the other hand, many have

embraced a neoclassical understanding of corporate punishment as a means of inducing cost

internalization by the individual participants in corporations, especially shareholders and

managers. The problem with this latter approach has been the difficulty of explaining why other

cost-internalization mechanisms, such as ordinary civil liability and Pigovian taxation, would

not work just as well.

In place of these accounts, I suggest a theory of corporate criminal liability grounded in a

conceptualization of the corporation as a team --- a relationship among individuals each of

whom treats doing her part towards shared goals as a principle of action alongside the pursuit

of her individual goals. Moral agency in team settings is based on the notion that in doing one’s

part towards the team’s goals, one becomes entitled to share in the credit for collective

achievements of the team, and one shares in the discredit for its collective failures. In particular,

when another member commits a wrong under the influence of the team’s norms, the wrong is a

negative achievement of the team, in which the members share.

This understanding of the corporation as a team and of the relationship between individual

agency and responsibility for wrongdoing in team settings illuminates the nature and purpose

of corporate criminal liability. When a participant in an incorporated business organization

commits a crime motivated by organizational norms, the entity and not only the direct

wrongdoer should be punished. The rationale for corporate punishment is to cast opprobrium

upon the organization’s norms so as to influence the behaviour of the participants whose

behaviour contributes to the generation of those norms, and also to give food for thought to the

members of the organization as they evaluate their continued attachment to the corporate team.