theories of corporation personality

Crim Law and Philos
DOI 10.1007/s11572-013-9232-y
ORIGINAL PAPER

Social Engineering as an Infringement of the Presumption of Innocence: The Case of Corporate
Criminality
Douglas Husak

Ó Springer Science+Business Media Dordrecht 2013

Abstract I examine how deferred-prosecution agreements employed against suspected corporate criminality amount to a form of social engineering that infringes the presumption. I begin with a broad understanding of the presumption itself. Then I offer a brief description of how these agreements function. Finally I address some of the normative issues that must be confronted if legal philosophers who hold retributivist views on punishment and sentencing hope to assess this device. My judgment tends to be favorable.
More importantly, I caution against the facile assumption that any means to infringe the presumption is necessarily an illegitimate part of penal practice.
Keywords Presumption of innocence Á Retributivism Á Corporate criminality Á Deferred prosecution agreements Á Social engineering

Introduction
The presumption of innocence is among the handful of doctrines that laypersons think they understand about our criminal justice system. Their confidence, however, is largely misplaced. Knowledgeable commentators appreciate that positive law has eviscerated this presumption in the United States and that its significance is far less than civics lessons would lead us to suppose. The full impact of the presumption is blunted easily and routinely. Indeed, its full impact has been eroded even more frequently than commentators acknowledge. My ambition in this paper is to describe and assess a relatively recent innovation in criminal justice that infringes the presumption of innocence as broadly construed. The existence of an offense is used to create the opportunity to engage in social engineering, that is, to provide an occasion for criminal justice officials to seek to improve the behavior of alleged wrongdoers and thereby reduce the likelihood of subsequent

D. Husak (&)
Department of Philosophy and Law, Rutgers University, New Brunswick, NJ, USA e-mail: husak@rci.rutgers.edu

123

Crim Law and Philos

offenses.1 Although occasions to engage in social engineering are becoming more prevalent in our penal justice system, I focus on one such opportunity in particular. I argue that a novel procedure used against corporate criminality amounts to a form of social engineering that infringes the presumption of innocence—when this presumption is interpreted widely. The device I describe is well-known to experts who specialize in corporate criminality, but has received almost no attention from legal philosophers. By turning our attention to this relatively recent procedure, we may succeed in reaching new insights about more familiar issues in the philosophy of criminal law.
Although the failure to relate the device I describe to the presumption of innocence has several possible explanations, two should be mentioned at the outset. First, many commentators may believe that the normative principles governing penal sanctions have no application to corporate criminality. Retributivism—the general tradition I presuppose— has rarely been applied to questions about the legitimacy of punitive measures employed against corporations. Perhaps retributivists think their view has nothing to say about this important topic. If this belief were true, it would represent an important limitation on the scope of retributive thought—a concession that members of this school have not explicitly acknowledged. Presumably, consequentialists admit to no such limitation, but the terrain of sanctions for corporate defendants should not be conceded to them without a fight. It is at least worthwhile to speculate about whether and how the fundamental principles of criminal law—including the presumption of innocence—might apply to some recent developments in the state’s approach to corporate criminality. Moreover, if these principles allow a given kind of social engineering in the corporate context, we should at least be open to the possibility that they permit it elsewhere.
Second, many commentators approve of the procedure I examine. It may seem counterintuitive to allege that a means to infringe the presumption of innocence—a doctrine venerated by theorists on all points along the political spectrum—might be defensible. In the final analysis, I will contend that our judgment about the permissibility of this device is complex, turning both on empirical as well as on principled controversies.2 The important point, however, is that legal philosophers should not be quick to condemn a given infringement of the presumption as pernicious or unjust. The weight of the presumption might vary from one context to another, and may have minimal force in the corporate context—even though it is not altogether negligible. Although there is ample ground for skepticism about the legitimacy of any procedure to infringe such a central pillar of AngloAmerican jurisprudence, calls to abolish the device I describe and preserve the presumption of innocence have substantial costs that should not be borne lightly.
This paper contains three parts. In Part I, I briefly discuss the presumption of innocence itself. I make no effort to be comprehensive. Apart from a quick digression about the nature of guilt, my efforts are designed almost solely to support my judgment that the practice I subsequently describe indeed infringes the presumption. My interpretation of the presumption is broad—far broader than the truncated version adopted in the United States as a result of decisions by our Supreme Court. In Part II, I examine how deferred-prosecution agreements (DPAs) employed against suspected corporate criminality amount to a form of social engineering that implicates the presumption. I offer a brief description of how these agreements function. In Part III, I address some of the normative issues that must be confronted if legal philosophers who hold retributivist views on punishment and sentencing
1

I do not intend the term social engineering to have negative connotations.

2

Because the device I describe may be justifiable, I speak of an infringement rather than a violation of the presumption of innocence.

123

Crim Law and Philos

hope to assess this device. My judgment tends to be favorable. More importantly, I caution against the facile assumption that any means to infringe the presumption is necessarily an illegitimate part of penal practice.

Formulating the Presumption of Innocence
No one should purport to assess whether a given device is or is not compatible with the presumption of innocence without specifying the content of the presumption itself. This task is of special importance in the present context, since the use of DPAs against corporate criminality is hardly the clearest example of an infringement—especially in light of judicial decisions (which I soon describe) that have narrowed the scope of the presumption in the United States. How might we formulate a robust, broad version of the presumption?
According to Shima Baradaran, the ‘‘lost’’ and ‘‘greater meaning’’ of the maxim ‘‘innocent until proven guilty’’ is that ‘‘the defendant was protected against any inferences or findings of guilt before trial’’ (Baradaran 2011, p. 1). Andrew Stumer goes even further. According to him, the idea ‘‘underlying’’ the presumption of innocence is ‘‘that the state must not take coercive action against any individual unless it has been proved that the person is guilty of a criminal offence and is properly subject to punishment’’ (Stumer 2010, p. xl). My own version resembles these accounts. On my view, a broad understanding of the presumption includes two parts. First, it contains a burden of proof: persons must be spared from penal sanctions unless the state proves them to be guilty. Second, it establishes a standard of proof: persons must be spared from penal sanctions unless their guilt is proved beyond a reasonable doubt. The combination of these components yields the complete formulation of the presumption of innocence as broadly construed: persons must be spared from penal sanctions unless the state proves their guilt beyond a reasonable doubt.
Since these two components are so easily stated together, why bother to disaggregate them? The answer is that the separation allows us to appreciate two distinct ways in which the presumption might be infringed or violated. First, a system of criminal justice might allow a person to be punished even though the state does not prove his guilt. In other words, it might alter the burden of proof. Second, even a system of criminal justice that preserves the burden of proof might allow a person to be punished on the basis of something less than proof beyond a reasonable doubt. In other words, it might relax the standard of proof.3 Both devices infringe or violate the presumption of innocence, but do so in different ways. One virtue of separating these components is that devices to alter the burden of proof might (or might not) be more objectionable than those that relax the standard of proof. In any event, each of these two kinds of infringement or violation seems to be occurring with increasing frequency in systems of criminal justice around the AngloAmerican world (Ashworth 2006).4
Before proceeding, we might pause to ask: what is the nature of guilt that must be proved beyond a reasonable doubt? Efforts to answer this question reveal complexities in the content of the presumption of innocence that seldom are recognized. Perhaps the
3

Some scholars have openly called for a relaxation of the standard of proof in criminal trials (Laudan
2011).

4

Ashworth (2006), p. 63 distinguishes four types of example. The presumption of innocence can be threatened by ‘‘confinement, by defining offences so as to reduce the impact of the presumption; erosion, by recognizing more exceptions; evasion, by introducing civil law procedures in order to circumvent the rights conferred on accused persons; and side-stepping, by imposing restrictions on the liberty of unconvicted persons but not depriving them of their liberty.’’.

123

Crim Law and Philos

obvious answer to this question is that the presumption demands proof beyond a reasonable doubt that the defendant has committed the crime of which he is accused. The problem with this answer is not simply that crime cannot be whatever the legislature says it is. This problem, I believe, is widely appreciated, even if it is not easily solved. States cannot be allowed to manipulate the content of offenses so that components normally regarded as elements are treated as defenses and thus are placed outside of what the prosecution must prove to establish guilt (Jeffries and Stephan 1979). Instead, the more fundamental problem is that this ‘‘obvious’’ answer is false. Consider the many cases in which a defendant admits to having engaged in a given kind of conduct, but denies that this conduct amounts to the crime for which he is charged. In Keeler v. Superior Court,5 for example, the defendant beat and kicked his pregnant ex-wife with the intention of killing her fetus.
When tried for murder, he conceded both his act and his culpable state, but insisted that the intentional killing of a fetus did not amount to murder under California law. If indeed the presumption requires proof that the defendant committed the crime of which he is accused, one might naturally suppose it would require the prosecution to prove beyond a reasonable doubt that the intentional killing of a fetus is murder. Almost certainly, however, the presumption makes no such demand. Even when a question is raised at trial about the law
(e.g., about whether killing a fetus is murder), it is for the judge to declare what the law is rather than for the prosecution to prove it. Perhaps, then, it is more accurate to say that the presumption requires proof beyond a reasonable doubt only of the actus reus and the mens rea which the prosecution alleges to constitute the crime charged.
But I put these important complexities about guilt aside in order to highlight features of my formulation of the presumption of innocence that are more salient for my purposes. I intend my interpretation to be unusually wide in two (or perhaps three) respects. First, it is broad as to whom it applies. On my interpretation, the presumption does not only attach to defendants charged with a crime or whose cases actually come to trial. Instead, it attaches to all entities subject to criminal liability. In other words, the presumption protects anyone the state may punish. These entities include natural and corporate persons. The supposition that corporations are entitled to some of the legal protections extended to natural persons is riddled with controversy, to say the least.6 But even though this supposition is enormously problematic in several contexts, I know no commentator who denies that the presumption applies to corporate defendants. Although I will not pursue the matter here, it would be highly instructive to ponder why this concession is made so readily. If the presumption applies to corporate defendants, it cannot be for the very same reasons it applies to natural persons. Do we protect moral autonomy, show respect for persons, or preserve any of the other traditional values associated with the presumption when we grant it to corporate defendants? Entirely different rationales may be needed to defend the application of the presumption of innocence to corporations.
Second, my formulation of the presumption is broad as to when it applies. On my interpretation, the presumption applies at all times—from the moment of birth to the time of death. If indeed there is no moment at which the state may impose penal sanctions on a person unless it first proves her guilt, it is clear that the presumption is at odds with legal realities. In the United States, the vast majority of persons who are punished are not proven guilty by any standard, let alone beyond a reasonable doubt. Instead, punishment usually follows from a guilty plea (Lippke 2011). Of course, prosecutors need to offer huge
5

470 P.2d 617 (1970).

6

The controversy has boiled in the wake of Citizens United v. Federal Election Commission, 558 U.S.
08–205 (2010).

123

Crim Law and Philos

incentives to induce defendants to plead guilty. Generally, the sentences defendants receive when they reject plea bargains and are found guilty at trial are far more severe than those that would have been imposed had they accepted the deal. Because defendants are aware of these ‘‘trial penalties,’’ they take enormous risks if they invoke the presumption of innocence and require the state to prove their guilt beyond a reasonable doubt.7 Even when they know themselves to be innocent or to have a plausible defense, they frequently are rational to plead guilty to avoid the severe sentences that would be inflicted if they erroneously were found to be guilty at trial (Stuntz 2004).
Perhaps I need to contend that my formulation of the presumption is broad in yet a third respect in order to support my claim that the device I describe infringes it. Again, I hold that the presumption is infringed or violated whenever the state imposes penal sanctions on persons whose guilt has not been proved beyond a reasonable doubt. My account refers to penal sanctions rather than to punishments, and this difference may or may not be substantial. The Supreme Court has held that the presumption applies only to defendants who are subject to punishment. Not all restraints on liberty imposed by the state—even those that are onerous—amount to punishments that implicate the presumption.8 In order to resolve whether a given restraint qualifies as a punishment ‘‘in the constitutional sense,’’ a court must decide whether it is imposed for the purpose of punishment or is merely an incident of some other legitimate governmental objective. Thus, if a particular condition of pretrial detention is reasonably related to a legitimate state purpose—such as prevention—it may not be an instance of punishment.9 In any event, it is clear that the type of social engineering
I describe does not involve a paradigm case of punishment. I tend to believe that this mode of social engineering is a form of punishment rather than an alternative to it, but I need not press this controversial point. In the cases of corporate criminality I will discuss, prosecutors use their statutory power as an opportunity to try to improve the behavior of the alleged offender and reduce the likelihood of subsequent offenses. The state secures this result by threatening prosecution and punishment unless the defendant agrees to accept its terms.
Thus the state has leverage to obtain an outcome it prefers but probably could not achieve by resorting to more conventional sanctions. Ultimately, it may not matter for our normative purposes whether we conceptualize these outcomes as types of punishment.
Thus my formulation of the presumption of innocence is broad in two or perhaps three respects. As a matter of settled law, however, my wide construction—in particular, my view about when and to whom the presumption applies—encounters a crushing objection: it has been explicitly rejected by the Supreme Court. In a series of cases challenging the conditions of various pre-trial interventions, the Court has held ‘‘the presumption of innocence is a doctrine that allocates the burden of proof in criminal trials.’’10 In short, the presumption has nothing whatever to say about arrestees or persons whose cases do not come to trial. As a number of scholars have pointed out, this narrow understanding saps the presumption of much of its normative vitality (Baradaran 2011). But if the presumption is construed more broadly—a huge assumption I henceforth make—its potential normative implications are enormous.
7

The tension between plea-bargaining and the presumption of innocence has rarely been noticed. One
European commentator, however, observes ‘‘plea-bargaining itself may be seen as an infringement of the presumption of innocence’’ (Maffei 2004, p. 1055).

8

Bell v. Wolfish, 441 U.S. 520, 537 (1979).

9

Id., at 538.

10

Id., at 533. Thus (for example) the presumption ‘‘has no application to a determination of the rights of a pretrial detainee during confinement before his trial has even begun.’’

123

Crim Law and Philos

I am inclined to seek an intermediate position about the normative status of the presumption of innocence in order to provide a more nuanced account of its legal implications. Philosophers are accustomed to supposing that various rights and standards exert different weights in the different contexts in which they apply. A right to free speech, for example, is generally believed to be more stringent in political than in commercial contexts
(Shiner 2003). I suggest the presumption of innocence might be conceptualized similarly.
The foregoing decisions by the Supreme Court indicate that the presumption has a single, narrow formulation that either applies (for example, in criminal trials) or does not (for example, in plea-bargaining). But why not construe the presumption broadly, as I have proposed, while allowing its weight to vary in the several areas of the penal law to which it applies? If the weight of the presumption varies, more stringent normative considerations would be needed to override it in some domains than in others. Of course, this suggestion would require a detailed defense. We would need criteria to determine whether and why the presumption is more forceful in some contexts than in others. But we should not assume uncritically that the presumption cannot be formulated as broadly as I propose here because it must always have the weight it exerts in criminal trials.
In what follows, I explore a few of the normative implications of the presumption of innocence as broadly construed. In particular, I discuss a means by which our criminal justice system infringes the presumption. As I have indicated, persons who accept pleabargains almost always receive more lenient sentences. But a defendant need not plead guilty in order to evade the more severe punishments the state would impose after a conviction at trial. In the circumstances I describe—involving corporate criminality—the defendant need not plead guilty because he need not plead at all. Indeed, he has not even been prosecuted; a prosecution has only been threatened. In order to avoid the severe punishment to which the threat almost certainly would lead, the corporation agrees to comply with a state practice of social engineering by changing its behavior in specified respects. This procedure is not promoted as paternalistic; improving behavior is not said to be the ultimate goal of social engineering. Instead, these changes are advantageous to the state because they are alleged to reduce the risk of subsequent offenses.11 If the presumption of innocence is infringed when defendants plead guilty to avoid severe sanctions, the same result obtains when persons alter their conduct to accomplish the same end.
Before turning to the details of the particular infringement I describe, it would be helpful to briefly locate my thoughts about the presumption of innocence in existing scholarship.
This task turns out to be surprisingly difficult. Many theorists contrast formal from substantive interpretations of the presumption (Tadros and Tierney 2004). The presumption is construed as formal when it imposes no constraints on the content of criminal statutes; it allows the criminal law to proscribe anything at all. But once a criminal statute has been enacted, the presumption bars conviction unless the state proves guilt, as defined by the relevant statute, beyond a reasonable doubt (Roberts 2005). By contrast, the presumption is construed as substantive when it disallows conviction for an offense that fails to conform to one or more normative constraints.12 A substantive interpretation is easy to motivate. After all, if the point of the presumption is to protect persons from unjustified punishment, the risk of being convicted for conduct that is not worthy of punishment should elicit much the same concern as the risk of being convicted for a crime they did not commit (Tomlin 2003).
11
Moreover, this change is beneficial because it reduces the cost of reducing the risk of subsequent offenses. 12
A normative interpretation (although not exactly corresponding to the description I provide) is advanced by Duff (2005).

123

Crim Law and Philos

Obviously, opportunities to violate the presumption are multiplied to the extent it is understood substantively rather than formally.13 In any event, the particular infringement I will describe is not easily categorized as formal or as substantive. Although it might be possible to characterize this device solely in procedural terms, it is available only for corporate entities suspected of perpetrating given kinds of offense. Here, as elsewhere in penal practice, the line between form and substance is elusive and unclear (Roberts 2011).
It should be relatively clear that the procedure I describe in Part II indeed infringes the presumption of innocence as I have characterized it. The state infringes this presumption by engaging in social engineering—using its leverage to change behavior in ways it perceives as desirable in order to reduce the likelihood of subsequent offenses. Pursuant to the procedure I discuss, corporate entities secure binding agreements from prosecutors that they will not be charged if they alter their conduct in specified respects. I conclude by examining some of the normative issues that must be addressed in order to decide whether this device is justified, notwithstanding its infringement of the presumption of innocence.
By focusing on the application of the presumption to the fairly unusual case of corporate criminality, I hope to gain a deeper understanding both of the presumption itself and of the legitimacy of some novel sanctions employed by systems of criminal justice.

Corporate Crime and Deferred Prosecution Agreements
My illustration of social engineering in criminal justice involves corporate wrongdoing.14 I select this example largely because I believe it provides a plausible kind of case in which the state might be permitted to impose penal sanctions on an entity without proving guilt beyond a reasonable doubt.
Textbooks suggest that the role of the prosecutor is to bring charges against persons suspected of criminality. Whatever may be true elsewhere, however, prosecutors in the
United States engage in forms of social engineering in the domain of corporate criminality.
They increasingly use their power not to seek conventional punishments, but to change the ways corporate entities do business. Prosecutors pursue this objective by threatening to bring charges and by negotiating the conditions under which they commit themselves not to prosecute. The terms they negotiate typically are called non-prosecution agreements
(NPAs) or (more commonly) deferred prosecution agreements (DPAs). These devices suspend a pending prosecution as long as the corporation complies with the terms of the agreement. The threat is terminated when compliance is judged to be complete. Until that time, any substantial breach of the DPA authorizes prosecutors to resume formal charges without regard to the statute of limitations.15
13
Subtle interpretations of the presumption that are intermediate between a formal and substantive interpretation might be generated. Thus one commentator (Tadros 2007) invokes the presumption against what he calls divergent offenses.
14
Corporate criminality has produced a massive literature. Most recently, see the Symposium on corporate crime in 41 Stetson Law Review (2011).
15
No data indicate that substantial numbers of corporations have failed to comply with the terms of DPAs and are prosecuted as a result. I could find only one such instance. In 2007, Vetco Gray International, Inc. and three subsidiaries all pleaded guilty to bribing Nigerian officials. Another subsidiary, Aibel Group, Ltd., a U.K. corporation, entered into a deferred prosecution agreement. In 2008, Aibel Group Ltd. reported that it had continued to violate the law, thus breaching the DPA. Having breached, the firm pleaded guilty and was ordered to serve two years of organizational probation requiring periodic reports on implementation of antibribery measures. Thanks to Brandon L. Garrett for calling this case to my attention (Garrett 2011).

123

Crim Law and Philos

These agreements have become increasingly popular for a wide array of corporate crimes: health and safety violations, environmental offenses, antitrust conspiracies, bribery and corruption in government contracts, and securities fraud, to name just a few (Alexander and
Cohen 2011). In light of the breadth of the offenses for which they are invoked, it should not be surprising that the content of DPAs varies dramatically. Most involve provisions that historically fall solely within the authority of boards of directors or shareholders. Many dictate that the firm must make specified changes to membership of its board, sometimes adding or deleting a named individual. Others mandate novel arrangements to separate powers, appoint a new General Counsel, require the resignation of senior management officers, or necessitate fundamental changes in business practices (Arlen and Kahan forthcoming). Some DPAs go so far as to include provisions that force corporations to work directly toward socially positive outcomes barely related to their wrongdoing. For example, the DPA for an alleged securities violation by Bristol–Myers Squibb required it to endow a program in business ethics and corporate governance at a neighboring law school. Wal-Mart’s settlement of potential criminal liability for a stampede at one of its stores mandated contributions of 1.8 million dollars to the local Youth Board and United Way (Epstein 2011). More typically, however, these agreements restructure corporations to minimize the risk of subsequent offenses.
According to Samuel Buell, ‘‘criminal DPAs now routinely require firms to reorganize business operations, adopt compliance measures, submit to enhanced monitoring for legal violations, and create systems to encourage and protect whistle-blowers’’ (Buell 2011, p. 93).
Clearly, these sanctions are not conventional punishments.16 It is fair to say that prosecutors who negotiate these terms use state power to engage in modes of social engineering.
Philosophers of criminal law have said almost nothing about this important development in penal justice. Some will be introduced to this procedure for the first time. I have long complained that criminal theorists tend to focus almost exclusively on so-called core offenses committed by natural persons (Husak 2004). A similar complaint is that philosophers of punishment concentrate almost solely on traditional types of sanction such as fines and imprisonment. Corporate offenses and DPAs fall outside of these parameters. The failure to consider these non-core offenses and unusual sanctions impoverishes legal philosophy by leading theorists to miss some of the most exciting and controversial developments in criminal justice. Strategies for dealing with corporate criminality provide an occasion to rethink several of the principles legal philosophers hold dear. This relatively unexplored field may help us to reach new insights about the presumption of innocence and its importance—insights we are likely to miss if we remain stuck within the core of crime and fixated on familiar modes of punishment.17
Even though there is some precedent for these arrangements—as I discuss below—it may be important to understand why DPAs are a relatively recent innovation in criminal justice. No single answer explains their current popularity; a combination of at least three factors accounts for their emergence. First, DPAs are the product of creative thinking among Attorneys General and their Deputies in the Department of Justice.18 Sometime
16
Some of these sanctions resemble requirements that defendants perform community service, although it is hard to see how such measures improve the behavior of defendants and reduce the risk of subsequent criminality. 17
For example, suppose we hold that punishment is justified because culpable wrongdoers incur duties to the victims and to society. The question is: what is the content of these duties; how are they best discharged?
Can the existence of these duties provide an occasion for social engineering? For a defense of the ‘‘duty view,’’ see Tadros (2011).
18
Of particular salience are the memos circulated by the (then) Deputy Attorneys General Eric Holder in
1999 and Larry Thompson in 2003.

123

Crim Law and Philos

around 1990, a broad consensus emerged that too little was being done to deter corporate crime. New ideas were needed because existing efforts to sentence individual wrongdoers within the firm had proved ineffective in retarding corporate criminality. Second, defendants risk punishments of unprecedented severity that create huge incentives to bargain and negotiate. Sentences in the United States, both inside and outside of the corporate context, are extraordinarily harsh relative to those in other countries. In particular, corporations convicted of federal crimes are subject to fines that can reach hundreds of millions of dollars, and enormous civil penalties may be imposed as well (Arlen 2011). The stakes of conviction are so vast that potential corporate defendants have ample reason to reach bargains with prosecutors. Finally, the phenomenon of overcriminalization makes it difficult even for conscientious firms to avoid committing offenses. Statutes creating corporate crimes are phrased so broadly that opportunities for prosecution are nearly ubiquitous (Husak 2008). Any number of examples might be given. John L. Diamond explains how prosecutions for such crimes as embezzlement, false pretense, extortion, bribery, and obstruction of justice apply vague statutory definitions, encompass behavior commonly viewed as legitimate, and pose a danger of being politically motivated (Diamond 2010). The charge-stacking these offenses allow make incentives to bargain nearly irresistible (Dervan 2011). Justifiable or not, these three factors help to explain why DPAs became popular in the twenty-first century rather than in an earlier era.
Assessing DPAs poses a challenge that legal philosophers have barely begun to undertake. It is noteworthy that almost all of the scholarship about these arrangements has been written by commentators who presuppose a consequentialist normative framework.
Academic specialists in corporate criminality tend to be trained in economics rather than in philosophy. It is not surprising that many of the theorists who adopt a cost-benefit perspective are relatively enthusiastic about DPAs (Alexander and Cohen 2011). Despite his well-known reservations about corporate criminality generally,19 Richard Epstein concedes that ‘‘the strongest argument in favor of accepting unbounded DPAs is that the mutual gain for the parties (who could otherwise not agree) normally results in a social improvement’’
(Epstein 2011, p. 48).
To be sure, consequentialists have expressed reservations about DPAs and returned mixed verdicts about their justifiability (Arlen and Kahan forthcoming). Some critics worry about abuse and contend that too many DPAs include terms that lie beyond the scope of legitimate state authority (Epstein 2011). The most pervasive complaint, however, is that prosecutors do not have the manpower or competence to bring about the structural reforms that would reduce subsequent criminality. Prosecutors typically lack the business experience to compile and evaluate the complex data needed to monitor compliance with the detailed provisions DPAs often include (Khana 2011). Officials in regulatory agencies have far more expertise than prosecutors in making these judgments reliably (Arlen 2011).
Moreover, the very success of DPAs as deterrents might distract states from the more urgent need to improve regulatory oversight (Griffin 2011). But these complaints, however trenchant, do not amount to a principled objection to DPAs. Instead, they mostly allege that the goals sought by these arrangements could be achieved more effectively if they were supplemented by additional efforts and if regulatory agencies cooperated with prosecutors in negotiating them.
In what follows, I propose to put aside these in-house disagreements among consequentialists. Although I will not defend my position here, I hold that retributivism provides
19
‘‘Congress and state legislatures…should scrap the doctrine of corporate criminality altogether.’’ Epstein
2011, p. 58.

123

Crim Law and Philos

the more defensible framework for evaluating institutions of punishment and sentencing.
Thus the more relevant question is: how do DPAs fare from a retributivist perspective?
Legal philosophers have written a great deal about foundational issues in corporate criminality.20 In particular, they have examined at length whether corporations are the kinds of entity that bear moral responsibility. The debate unfolds in two steps. First, legal philosophers prepare lists of the properties an entity must possess in order to be a moral agent. Next, they debate whether corporations do or do not possess the properties allegedly required (Hasnas 2010). I will bypass these foundational issues here and make the huge assumption that corporate criminal liability is defensible within a retributive framework.
As I have indicated, retributivists have said very little about the sanctions imposed on corporations—or about the principles that should be used to assess their justifiability. But retributivists should not concede that their penal philosophy has no application to this important area of penal law. Those retributivists who are willing to extend their thought to corporate sanctions might be expected to be somewhat more skeptical than consequentialists of DPAs.
Without engaging in foundational debates, how might retributivists proceed in deciding whether DPAs are justified? All retributivists, I believe, place the principle of proportionality somewhere near the heart of penal justice. Do DPAs preserve proportionality by tailoring the severity of the sanction to the seriousness of the corporate offense? If not, it is doubtful that retributivists should share the enthusiasm about these arrangements expressed by many consequentialists. Of course, several of the reservations about proportionality can be brought against the practice of plea-bargaining more generally. Why think that bargains will produce proportionate punishments? For better or worse, however, the institution of plea-bargaining is here to stay, and criminal justice could scarcely survive without it. But
DPAs, as I have indicated, are a relatively recent phenomenon we could easily ban.
Moreover, they are employed even before a prosecution is formally initiated. Does the use of bargaining to defer prosecution rather than to induce a plea of guilty after prosecution raise special difficulties that go beyond broader concerns about plea-bargaining? I turn now to such questions.

Normative Considerations
I have described an area of law in which the criminal justice system engages in social engineering. The state uses a penal offense as an opportunity to try to improve the behavior of suspected wrongdoers in order to reduce the risk of subsequent criminality.
The entities subjected to these forms of social engineering have not been prosecuted
(let alone convicted), so I hope it is clear that this device infringes the presumption of innocence as broadly construed. The state imposes a penal sanction even though it has not proved guilt beyond a reasonable doubt. Is it just for the state to possess and use the leverage these devices provide? If not, why not? If so, might this power ever be exercised outside of the corporate context? Legal philosophers must wrestle with these normative questions. Unfortunately, the answer is uncertain, depending on a great many complex factors that are exceedingly difficult to weigh and balance. Many (but not all) of these contested issues are empirical.21 Despite my ambivalence and lack of certainty, my
20

See, for example, the entire content of Issue 1 of XIX Journal of Law and Policy (2010), pp. 1–230.

21

Indeed, the editors of the most important book on corporate criminality conclude that ‘‘empirical inquiry should be the primary aim of further research’’ (Barkow and Barkow 2011).

123

Crim Law and Philos

judgment tends to be favorable. Although the legitimacy of each mode of social engineering must be evaluated on its own merits, I conclude that the type represented by DPAs has advantages and should not be rejected simply because it infringes the presumption of innocence. One might resist my conclusion and adopt an extreme position on the normative question I raise. That is, one might hold that the presumption of innocence establishes a conclusive reason not to subject any entity to a penal sanction, corporate or otherwise, unless the state first proves it to be guilty of an offense beyond a reasonable doubt. This position might be held even by theorists who concede the benefits of social engineering.
Although this extreme position should not be dismissed out of hand, I reject it for two reasons. First, it suggests that the normative importance of preserving the presumption of innocence as broadly construed is paramount, trumping all (or nearly all) competing considerations. I see no ground to suppose that the presumption enjoys this lofty status relative to other goods that might be achieved by a system of criminal justice. Moreover, if
I am correct to suggest that the normative weight of the presumption may be greater in some domains than in others, we should not uncritically assume it has the same strength when used for corporate social engineering as when applied to natural persons at criminal trials. In short, despite its central role in Anglo-American jurisprudence, the presumption of innocence is merely one factor among many that must be taken into account in assessing a given procedure. Second, we should be mindful of the consequences if we conclude
DPAs to be unjustified. Are the alternatives preferable? More traditional kinds of punishment have proved stubbornly resistant to justification. We should not pretend that our more familiar methods of dealing with suspected wrongdoers, corporate or otherwise, are morally unproblematic. Because of the well-known difficulties in attempts to justify state punishment, we should always be receptive to new ideas about how to sanction offenders and reduce threats of future crime. Our final judgment about various types of social engineering must be made against the background of the normative problems we all know to plague the status quo.
Arguably, however, it would be preferable for the state to resort to social engineering only after persons had been prosecuted and convicted. If this option were followed, worries about the presumption of innocence would dissipate. In many respects, this alternative would be ideal. Defendants might be allowed to choose at sentencing whether they preferred to take part in whatever mode of social engineering is available. The most realistic response to this objection is comparable to that raised against challenges to plea-bargaining
(Young 2013). Quite simply, our state lacks the resources to prosecute everyone it has probable cause to think has committed a crime. If a suspected offender really believes himself to be innocent, he can always refuse to negotiate and take his chances at trial. As I have said, however, most defendants are rational to negotiate with the state. Moreover, if we believe that the state would be justified in subjecting these entities to traditional punishments if they are tried and found to be guilty, acceptance of the terms negotiated in lieu of prosecution probably satisfies our tests of voluntariness (Westen 2004, especially
p. 183). On what basis should third parties oppose a socially beneficial procedure to which both the state and potential defendants voluntarily consent? DPAs may be acceptable even though we are able to imagine a preferable alternative in a possible world in which criminal justice officials possess unlimited resources. In the actual world we inhabit, DPAs may be the best we can do.
Presumably, consent should be needed to undertake any mode of social engineering; we must remember that corporations agree to the terms in deferred prosecution agreements.
Unfortunately, not all justificatory questions are resolved by conceding that consent is

123

Crim Law and Philos

voluntarily given by the entities subjected to social engineering.22 The wrongs proscribed by the criminal law are (or ought to be) public; members of the communities in which these harms are perpetrated have an interest in ensuring that justice is done (Duff 2007, especially pp. 140–146). Unlike in civil law, the willingness of both defendant and victim to settle a dispute out of court does not terminate the public interest when the criminal law is involved. This worry is trenchant whenever one recommends a non-traditional sanction for a penal wrong.23 It is not clear, however, why the public interest is not served by the particular mode of social engineering I have described. Members of communities should welcome a response to corporate crime that improves the behavior of suspected wrongdoers and minimizes the likelihood of subsequent offenses.
What, then, should retributivists conclude about this novel procedure? Two related questions loom large, but I can barely scratch the surface in addressing them. In the first place, one might worry about what does not happen when DPAs are negotiated. These devices enable corporate entities to evade the traditional punishments that would have been imposed had they been prosecuted and convicted. The fact that the vast majority of guilty individuals manage to avoid liability has always raised tricky questions for retributivists.
For all but the most serious crimes, the typical offender is neither arrested nor charged. To be sure, corporations that evade prosecution through DPAs do so as a matter of deliberate state policy, not because their offenses are undetected. If nearly all of those whose behavior is altered by social engineering could have been proved guilty of a crime beyond a reasonable doubt—as seems likely—how could any retributivist be sanguine about the failure to convict them?
It is manifestly unclear, however, what sentences retributivists should believe to be deserved when corporations are convicted of wrongdoing. Proportionate punishments are hard enough to identify in ordinary cases, but corporate wrongdoing raises troublesome issues not replicated in examples of core criminality. In the first place, proportionality seemingly requires (ceteris paribus) that the suffering or deprivation inherent in punishment should be a function of the blame of the offender (Husak 2000; but see Berman
2013). Of course, corporations cannot literally suffer or be deprived of liberty. Thus we have little option but to devise a novel sanction in this domain. In addition, proportionality seemingly requires the severity of the sentence to be a function of the culpability of the defendant (von Hirsch and Ashworth 2005). Many of the statutes corporations breach impose strict liability—a type of liability about which retributivists have generally been critical.24 Finally, traditional punishments imposed on corporate wrongdoers produce
22

One might think that voluntary consent to a DPA rebuts the presumption of innocence. I believe, however, it is more accurate to say that consent waives the presumption. In fact, I am unsure whether the presumption of innocence can be rebutted because it is not a ‘‘true presumption’’ as that term is traditionally understood by criminal law theorists. According to Joshua Dressler, ‘‘presumptions operate in the following manner: Upon proof of Fact (or set of Facts) A, a factfinder must (or ‘‘may,’’ depending on the language of the instruction) presume Fact B.’’ Understanding Criminal Law (Lexis/Nexis, 5th ed., 2009), p. 79. The presumption of innocence, by contrast, does not require proof of any initial ‘‘fact A.’’ Instead, it applies before any evidence is adduced. Thus it is better described as a default rule: a rule describing how an official is required to behave in the absence of proof. In the absence of proof of guilt (beyond a reasonable doubt), the factfinder is required (and not merely permitted) to pronounce the defendant innocent and to render a verdict of acquittal.
23
Thus many schemes of restorative justice many not protect the public interest in sanctioning criminal wrongs. 24
Even though the statutes in question do not require culpability explicitly, we might inquire what level of blame is involved in typical cases covered by DPAs. Since much of the wrongdoing in these circumstances takes place despite the objections of the company’s owners and investors—and is prosecuted pursuant to the

123

Crim Law and Philos

collateral consequences that typically dwarf those arising in cases of core criminality.25
Impositions of the ‘‘death sentence’’ on huge firms such as Arthur Anderson & Company have cost thousands of innocent employees their livelihood, devastated the economic position of hundreds of retirees, and caused huge losses to unsuspecting shareholders. Thus we should not be too confident about the justifiability of the punishments that would have been imposed had DPAs not been negotiated.
In addition, retributivists need not believe that desert provides a very weighty reason in favor of either inflicting punishment or of selecting the particular mode of punishment to be inflicted. Legal philosophers can retain their retributive credentials by embracing the weaker thesis that desert provides a reason that permits the state to use defendants for instrumentalist ends. Any sensible version of retributivism will provide ample room for consequentialist factors in criminal law and sentencing. The state allows enormous discretion to police and prosecutors not to arrest or bring charges at all, or to decide what charges to file when they are brought.26 Retributivists should be no more opposed when discretion is used to negotiate DPAs.
In the second place, what should retributivists say when DPAs are used? Retributivists should oppose a quantum of punishment that is either too severe or too lenient. It is hard to know whether the terms in DPAs are too severe, but is doubtful that they are too lenient.
DPAs generally include huge monetary fines—ranging between $33 and $203 million from
2003 to 2010 (Arlen and Kahan forthcoming). Even rumors of a possible prosecution can have devastating reputational effects on a corporation, leading its profits and market share to plummet (Gilchrist 2013). The communicative effect of a pending prosecution can reverberate throughout the entire corporate culture and frequently is perceived as threatening the very survival of the firm (Buell 2011). Even the negotiations that culminate in
DPAs are stigmatizing. It is not clear why retributivists should ask for more.
But we might pursue an altogether different strategy in assessing the mode of social engineering represented by DPAs. Perhaps we can make progress on this difficult normative issue by thinking about whether and under what circumstances social engineering would be justifiable if used elsewhere—against natural persons. In short, if social engineering is a sensible response to suspected corporate wrongdoing, why shouldn’t we employ it in cases of individual wrongdoing? Social engineering outside the corporate context might have several advantages. Most importantly, it might help to overcome the charge of barbarism sometimes leveled against retributive theories of punishment. Can punitive sanctions really be justified even when they involve deprivations that produce no further instrumental goods? Is supposed conformity with an abstract principle of retributive justice all that is needed to defend punishment? No one doubts that the prevention of crime is among the central duties of the liberal state. Why not design our system of penal justice to achieve positive outcomes—to minimize the likelihood of future criminality by improving the offender himself?

Footnote 24 continued doctrine of respondeat superior—many offending corporations have simply failed to implement methods to detect and prevent criminal activity by their lower-level managers. Intuitively, these omissions do not involve massive amounts of culpability even when they cause large amounts of social harm. For a nuanced account, see Skupski (2013).
25
Commentators are increasingly persuaded that the magnitude of these collateral consequences should be a factor in gauging the severity of a punishment (Kolber 2012).
26
To fully evaluate the use of such discretion, better data are needed about how and why it is exercised. For an interesting study concluding that discretion seeks retributive objectives, see Minzer (2012).

123

Crim Law and Philos

A partial reply to these questions begins by pointing out that social engineering is already used against natural persons in Anglo-American systems of criminal justice (Berman and
Bibas 2006). Thus the procedures I describe are not entirely novel; most jurisdictions have mechanisms that serve as precedents for DPAs, even though they may differ from them in important respects. All such devices might be described as ‘‘hidden sentence proceedings’’ inasmuch as they take place before a formal sentencing hearing (or, indeed, even before prosecution), increasing or decreasing an offender’s punishment (Appleman 2007). A few states, for example, allow for probation before judgment. Although the details of this procedure vary from one jurisdiction to another, probation before judgment typically allows adjudication for a few specified charges to be deferred pursuant to conditions overseen and supervised by a judge. The court enters a verdict of not guilty if the defendant successfully completes the terms of his agreement without further violations. In some jurisdictions, for example, a person arrested for his first offense of drunk driving can avoid conviction by complying with the conditions specified in the probation before judgment.
Anti-social behavior orders (ASBOs) are a prime example of social engineering that has stimulated a great deal of critical commentary in Great Britain (Burney 2005). The procedures that lead to the issuance and enforcement of these orders might be thought to infringe the presumption of innocence to the same degree as those that lead to DPAs
(Ashworth and Zedner 2010). Thus a thorough assessment of ASBOs must decide how much weight to afford to the presumption of innocence. In any event, these orders are subject to a host of additional objections that may be even more devastating. In particular, the types of behavior that allow such orders to be issued need not be criminal. Effectively, then, conduct is criminalized whenever a person engages in whatever conduct is prohibited by a valid anti-social behavior order. Breach of the ASBO then renders the person liable to punishment not for the original conduct that led to the imposition of the ASBO, but for that very breach—that is, for conduct made criminal by the imposition of the ASBO itself. But since the conduct initially proscribed by such an order need not have been criminal, the presumption that persons are innocent unless proven guilty of a crime would seem to be violated more directly by ASBOs than by DPAs.27 By contrast, the conduct that leads to a
DPA is expressly prohibited by a penal statute. Legal philosophers need not be worried that the use of DPAs involves a hidden expansion in the scope of the substantive criminal law.
Arguably, the several kinds of ‘‘problem-solving’’ courts—and drug courts in particular—involve the very best examples of social engineering in positive law (Nolan 2001).
Frustrated by passing drug offenders through the revolving doors of penal justice, all fifty states in the United States have sought to decrease rates of recidivism by trying to help substance abusers to overcome their weaknesses. Although generalizations are perilous, many jurisdictions agree to defer prosecution for given kinds of drug offenders if suspects enter a treatment regime designed and overseen by the judge. Charges are dropped if the defendant completes the treatment regime by remaining abstinent for a specified period of time. Although the mode of social engineering pursued by these counts is an important development in penal justice, I have examined their rationales elsewhere and will only summarize one of my many reservations here (Husak 2011a). Among the most worrisome factors are the untested and unproven empirical assumptions on which these institutions depend. Despite constant assurances that ‘‘drug courts work,’’28 almost no reliable data
27
It is noteworthy that ASBOs are being replaced by ‘Criminal Behaviour Orders’ pursuant to Anti-Social
Behaviour Bill 2012, Part II.
28
‘‘We know that drug courts outperform virtually all other strategies that have been attempted for druginvolved offenders.’’ See http://www.ndci.org/courtfacts.htm.

123

Crim Law and Philos

supports the truth of this allegation. In much of the literature on drug courts, anecdote substitutes for evidence of effectiveness; this area is dominated by testimonials rather than by sound data.29 Admittedly, drug courts may reduce rates of recidivism more effectively than traditional punishments, but this conclusion sets the bar remarkably low. Few areas of criminal justice have evoked more criticism than our policy of punishing drug offenders.
Hopefully, the evidence that shows DPAs to be effective in reducing future criminality is more impressive. This result would not be surprising. Drug treatment specialists have struggled for generations to find viable solutions to the complex problems of substance abuse; it would be remarkable if untrained judges have managed to identify an effective strategy when legions of experts have failed. In short, efforts to improve the behavior of natural persons have frustrated generations of theorists committed to rehabilitation. But the changes in corporate structure I sketched in Part II are easier to implement and, I suspect, more likely to succeed. Of course, these issues are empirical. Other matters, however, are normative. For example, the content of DPAs seems less vulnerable to the charge that the behavior of offenders is changed in ways incompatible with moral autonomy and state neutrality. Even if social engineering is not entirely unprecedented in criminal justice, we should not exaggerate its growth. Several of the devices designed to protect the public from future criminality are not modes of social engineering and should not be confused with it. Moreover, they would seem to be far more difficult to justify than the procedures I have mentioned here.
Few systems of preventive detention, for example, involve social engineering as I have characterized it. Although preventive detention seeks to protect the public from future criminality, it does so simply by incapacitating persons believed to be dangerous.30 Few of these systems seriously purport to protect the public by improving the behavior of the persons detained. Oftentimes the detainees are not even believed to have committed crimes. If such detention is construed as penal, it goes without saying that these systems of preventive detention infringe the presumption of innocence. Moreover, the persons subject to preventive detention do not consent to their treatment—a key factor in assessing the legitimacy of any mode of social engineering. Notwithstanding their crucial differences, we might be able to reach insights about social engineering by thinking about the justifiability of systems of preventive detention. If we allow the state to preventively detain dangerous persons—as some theorists would allow—it is hard to see why we should prohibit the state from using social engineering to protect the public from future criminality.31

Conclusions
I have formulated the presumption of innocence broadly, construing it to oppose the imposition of penal sanctions on entities (natural or corporate) the state has not proven to be guilty of crime beyond a reasonable doubt. If this wide characterization is accepted, it should be clear that quite a few procedures in criminal justice infringe this presumption— in particular, the use of DPAs against corporate criminality. Penal statutes are used to
29
Since drug courts have existed for about two decades, one might expect that they would begin to affect the national rate of addiction. The incidence of addiction in the United States, however, has been largely unchanged despite the growth of the drug court movement.
30
See the several papers in the Symposia on the Morality of Preventive Restrictions of Liberty in 48 San
Diego Law Review 1075–1395 (2011b).
31

For a suggestion about how to improve a system of preventive detention, see Husak (2011, 2013).

123

Crim Law and Philos

provide the opportunity to engage in social engineering: efforts to improve the conduct of suspected offenders to reduce the likelihood of a subsequent offense.
The most interesting question about this novel procedure is whether retributivists should applaud or condemn it. I have sought to identify a few of the questions retributivists must address to resolve this normative issue, but have barely scratched the surface. A complete answer would need to identify the weight we should attach to the presumption of innocence in the corporate context. In addition, our assessment depends partly on whether we are able to identify a better solution that would be possible to implement in the real world. DPAs may be the best we can do to punish and deter corporate criminality within a retributive penal philosophy. Although I admit to ambivalence and my ultimate verdict is uncertain, I conclude that we should not condemn this procedure simply because it infringes the presumption of innocence. Finally, I have suggested that a perspective on this justificatory question might help us to decide whether and under what circumstances social engineering is defensible outside of the corporate domain.
Acknowledgments The author would like to thank Roger Shiner and to the several participants at the
‘‘Fraying the Golden Thread’’ conference at Aberdeen, Scotland in February, 2012.

References
Alexander, Cindy R. and Mark A. Cohen. (2011). The Causes of Corporate Crime. In Anthony S. Barkow and Rachel E. Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate
Conduct. New York: New York University Press, p. 11.
Appleman, Laura I. (2007). Retributive Justice and Hidden Sentencing. 68 Ohio State Law Journal 1307.
Arlen, Jennifer. (2011). Removing Prosecutors from the Boardroom. In Anthony S. Barkow and Rachel E.
Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct.
New York: New York University Press, p. 62.
Arlen, Jennifer and Marcel Kahan. (forthcoming, 2013). Corporate Governance Regulations through NonProsecution.
Ashworth, Andrew. (2006). Four Threats to the Presumption of Innocence, 123 South African Law Journal
63.
Ashworth, Andrew and Lucia Zedner. (2010). Preventive Orders: A problem of Undercriminalization? In
R.A. Duff, et.al., eds.: The Boundaries of the Criminal Law. Oxford: Oxford University Press, p. 59.
Baradaran, Shima. (2011). Restoring the Presumption of Innocence. 72 Ohio State Law Journal 1.
Barkow, Anthony S. Barkow and Rachel L. Barkow. (2011). Conclusion. In Anthony S. Barkow and Rachel
E. Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct.
New York: New York University Press, p. 249.
Berman. Mitchell N. (forthcoming, 2013). Rehabilitating Retributivism. Law and Philosophy.
Berman, Douglas A. and Stephanos Bibas. (2006). Making Sentencing Sensible. 4 Ohio State Journal of
Criminal Law 37.
Buell, Samuel W. (2011). Potentially Perverse Effects of Corporate Civil Liability. In Anthony S. Barkow and Rachel E. Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate
Conduct. New York: New York University Press, p. 87.
Burney, Elizabeth. (2005). Making People Behave: Anti-Social Behaviour, Politics and Policy. Cullompton:
William Pub.
Dervan, Lucian E. (2011). Overcriminalization 2.0: The Symbiotic Relationship between Plea Bargaining and Overcriminalization. 7 Journal of Law, Economics and Policy 645.
Diamond, John L. (2010). Reviving Lenity and Honest Belief at the Boundaries of Criminal Law. 44
University of Michigan Journal of Law Reform 1.
Duff, R.A. (2005). Strict Liability, Legal Presumptions, and the Presumption of Innocence. In A.P. Simester, ed. Appraising Strict Liability. Oxford: Oxford University Press, p. 125.
Duff, R.A. (2007). Answering for Crime. Oxford: Hart Pub. Co.
Epstein, Richard A. (2011). Deferred Prosecution Agreements on Trial. In Anthony S. Barkow and Rachel
E. Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct.
New York: New York University Press, p. 38.

123

Crim Law and Philos
Garrett, Brandon L. (2011). Globalized Corporate Prosecutions. 97 Virginia Law Review 1775.
Gilchrist, Gregory M. (forthcoming, 2013). The Expressive Cost of Corporate Immunity. 64 Hastings Law
Journal.
Griffin, Lisa Kern. (2011). Inside-Out Enforcement. In Anthony S. Barkow and Rachel E. Barkow, eds.:
Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct. New York: New
York University Press, p. 110.
Hasnas, John. (2010). Where is Felix Cohen When We Need Him: Transcendental Nonsense and the Moral
Responsibility of Corporations. XIX Journal of Law and Policy 55.
Husak, Douglas. (2000). Retribution in Criminal Theory. 37 San Diego Law Review 959.
Husak, Douglas. (2004). Crimes Outside the Core. 39 Tulsa Law Review 755.
Husak, Douglas. (2008). Overcriminalization. Oxford: Oxford University Press.
Husak, Douglas. (2011a). Retributivism, Proportionality, and the Challenge of the Drug Court Movement. In
Michael Tonry, ed.: Retributivism Has a Past: Has It a Future? Oxford: Oxford University Press,
p. 214.
Husak, Douglas. (2011b). Lifting the Cloak: Preventive Detention as Punishment. 48 San Diego Law Review
1173.
Husak, Douglas. (2013). Preventive Detention as Punishment? Some Possible Reservations. In Andrew
Ashworth, Lucia Zedner, and Patrick Tomlin, eds.: Prevention and the Limits of the Criminal Law.
Oxford: Oxford University Press, forthcoming.
Jeffries, John C. and Paul B. Stephan. (1979). Defenses, Presumptions and Burdens of Proof in the Criminal
Law, 88 Yale Law Journal 1325.
Khana, Virkamaditya. (2011). Reforming the Corporate Monitor? In Anthony S. Barkow and Rachel E.
Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct.
New York: New York University Press, p. 226.
Kolber, Adam J. (2012). ‘‘Unintentional Punishment,’’ 18 Legal Theory 1.
Laudan, Larry. (2011). The Rules of Trial, Political Morality, and the Costs of Error: Or, Is Proof Beyond a
Reasonable Doubt Doing More Harm than Good? In Leslie Green and Brian Leiter, eds.: Oxford
Studies in Philosophy of Law. Oxford: Oxford University Press, Vol. 1, p. 195.
Lippke, Richard L. (2011). The Ethics of Plea Bargaining. New York: Oxford University Press.
Maffei, Stefano. (2004). Negotiations ‘On Evidence’ and Negotiations ‘On Sentence’. 2 Journal of International Criminal Justice 1050.
Minzer, Max. (2012). Why Agencies Punish. 53 William & Mary Law Review 853.
Nolan, James L. Jr. (2001). Reinventing Justice: The American Drug Court Movement. Princeton: Princeton
University Press.
Roberts, Paul. (2005). ‘‘Strict Liability and the Presumption of Innocence: An Expose of Functionalist
Assumptions.’’ In A.P. Simester, ed. Appraising Strict Liability. Oxford: Oxford University Press,
p. 151.
Roberts, Paul. (2011). Groundwork for a Jurisprudence of Criminal Procedure. In R.A. Duff and Stuart
Green, eds.: Philosophical Foundations of Criminal Law. Oxford: Oxford University Press, p. 379.
Shiner, Roger A. (2003). Freedom of Commercial Expression. Oxford: Oxford University Press.
Skupski, George. (forthcoming, 2013). The Senior Management Mens Rea: Another Stab at a Workable
Integration of Organizational Culpability into Corporate Criminal Liability. Case Western Reserve
University Law Review.
Stumer, Andrew. (2010). The Presumption of Innocence: Evidential and Human Rights Perspectives.
Oxford: Hart Pub.Co.
Stuntz, William J. (2004). Plea Bargaining and the Criminal Law’s Disappearing Shadow. 117 Harvard Law
Review 2548.
Tadros, Victor. (2007). Rethinking the Presumption of Innocence. 1 Criminal Law and Philosophy 193.
Tadros, Victor. (2011). The Ends of Harm. Oxford: Oxford University Press.
Tadros, Victor and Stephen Tierney. (2004). The Presumption of Innocence and the Human Rights Act.
Modern Law Review 67.
Tomlin, Patrick. (forthcoming, 2013). Extending the Golden Thread? Criminalisation and the Presumption of Innocence. The Journal of Political Philosophy.
Von Hirsch, Andrew and Andrew Ashworth. (2005). Proportionate Sentencing. Oxford: Oxford University
Press.
Young, Michael. (forthcoming, 2013). A Moral Defense of Plea Bargaining.
Westen, Peter. (2004). The Logic of Consent. Burlington, Vt.: Ashgate Pub. Co.

123

References: Alexander, Cindy R. and Mark A. Cohen. (2011). The Causes of Corporate Crime. In Anthony S. Barkow and Rachel E Appleman, Laura I. (2007). Retributive Justice and Hidden Sentencing. 68 Ohio State Law Journal 1307. Arlen, Jennifer. (2011). Removing Prosecutors from the Boardroom. In Anthony S. Barkow and Rachel E. Arlen, Jennifer and Marcel Kahan. (forthcoming, 2013). Corporate Governance Regulations through NonProsecution. Ashworth, Andrew. (2006). Four Threats to the Presumption of Innocence, 123 South African Law Journal 63. Ashworth, Andrew and Lucia Zedner. (2010). Preventive Orders: A problem of Undercriminalization? In R.A Baradaran, Shima. (2011). Restoring the Presumption of Innocence. 72 Ohio State Law Journal 1. Barkow, Anthony S. Barkow and Rachel L. Barkow. (2011). Conclusion. In Anthony S. Barkow and Rachel E Berman. Mitchell N. (forthcoming, 2013). Rehabilitating Retributivism. Law and Philosophy. Berman, Douglas A. and Stephanos Bibas. (2006). Making Sentencing Sensible. 4 Ohio State Journal of Criminal Law 37. Buell, Samuel W. (2011). Potentially Perverse Effects of Corporate Civil Liability. In Anthony S. Barkow and Rachel E Burney, Elizabeth. (2005). Making People Behave: Anti-Social Behaviour, Politics and Policy. Cullompton: William Pub. Dervan, Lucian E. (2011). Overcriminalization 2.0: The Symbiotic Relationship between Plea Bargaining and Overcriminalization Diamond, John L. (2010). Reviving Lenity and Honest Belief at the Boundaries of Criminal Law. 44 University of Michigan Journal of Law Reform 1. Duff, R.A. (2005). Strict Liability, Legal Presumptions, and the Presumption of Innocence. In A.P. Simester, ed Duff, R.A. (2007). Answering for Crime. Oxford: Hart Pub. Co. Epstein, Richard A. (2011). Deferred Prosecution Agreements on Trial. In Anthony S. Barkow and Rachel E Garrett, Brandon L. (2011). Globalized Corporate Prosecutions. 97 Virginia Law Review 1775. Gilchrist, Gregory M. (forthcoming, 2013). The Expressive Cost of Corporate Immunity. 64 Hastings Law Journal. Griffin, Lisa Kern. (2011). Inside-Out Enforcement. In Anthony S. Barkow and Rachel E. Barkow, eds.: Prosecutors in the Boardroom: Using Criminal Law to Regulate Corporate Conduct Hasnas, John. (2010). Where is Felix Cohen When We Need Him: Transcendental Nonsense and the Moral Responsibility of Corporations Husak, Douglas. (2000). Retribution in Criminal Theory. 37 San Diego Law Review 959. Husak, Douglas. (2004). Crimes Outside the Core. 39 Tulsa Law Review 755. Husak, Douglas. (2008). Overcriminalization. Oxford: Oxford University Press. Husak, Douglas. (2011a). Retributivism, Proportionality, and the Challenge of the Drug Court Movement. In Michael Tonry, ed.: Retributivism Has a Past: Has It a Future? Oxford: Oxford University Press, Husak, Douglas. (2011b). Lifting the Cloak: Preventive Detention as Punishment. 48 San Diego Law Review 1173. Husak, Douglas. (2013). Preventive Detention as Punishment? Some Possible Reservations. In Andrew Ashworth, Lucia Zedner, and Patrick Tomlin, eds.: Prevention and the Limits of the Criminal Law. Jeffries, John C. and Paul B. Stephan. (1979). Defenses, Presumptions and Burdens of Proof in the Criminal Law, 88 Yale Law Journal 1325. Khana, Virkamaditya. (2011). Reforming the Corporate Monitor? In Anthony S. Barkow and Rachel E. Kolber, Adam J. (2012). ‘‘Unintentional Punishment,’’ 18 Legal Theory 1. Laudan, Larry. (2011). The Rules of Trial, Political Morality, and the Costs of Error: Or, Is Proof Beyond a Reasonable Doubt Doing More Harm than Good? In Leslie Green and Brian Leiter, eds.: Oxford Lippke, Richard L. (2011). The Ethics of Plea Bargaining. New York: Oxford University Press. Maffei, Stefano. (2004). Negotiations ‘On Evidence’ and Negotiations ‘On Sentence’. 2 Journal of International Criminal Justice 1050. Minzer, Max. (2012). Why Agencies Punish. 53 William & Mary Law Review 853. Nolan, James L. Jr. (2001). Reinventing Justice: The American Drug Court Movement. Princeton: Princeton University Press. Roberts, Paul. (2005). ‘‘Strict Liability and the Presumption of Innocence: An Expose of Functionalist Assumptions.’’ In A.P Roberts, Paul. (2011). Groundwork for a Jurisprudence of Criminal Procedure. In R.A. Duff and Stuart Green, eds.: Philosophical Foundations of Criminal Law Shiner, Roger A. (2003). Freedom of Commercial Expression. Oxford: Oxford University Press. Skupski, George. (forthcoming, 2013). The Senior Management Mens Rea: Another Stab at a Workable Integration of Organizational Culpability into Corporate Criminal Liability Stumer, Andrew. (2010). The Presumption of Innocence: Evidential and Human Rights Perspectives. Stuntz, William J. (2004). Plea Bargaining and the Criminal Law’s Disappearing Shadow. 117 Harvard Law Review 2548. Tadros, Victor. (2007). Rethinking the Presumption of Innocence. 1 Criminal Law and Philosophy 193. Tadros, Victor. (2011). The Ends of Harm. Oxford: Oxford University Press. Tadros, Victor and Stephen Tierney. (2004). The Presumption of Innocence and the Human Rights Act. Tomlin, Patrick. (forthcoming, 2013). Extending the Golden Thread? Criminalisation and the Presumption of Innocence Von Hirsch, Andrew and Andrew Ashworth. (2005). Proportionate Sentencing. Oxford: Oxford University Press. Young, Michael. (forthcoming, 2013). A Moral Defense of Plea Bargaining. Westen, Peter. (2004). The Logic of Consent. Burlington, Vt.: Ashgate Pub. Co.