Corporate Essence and Identity in Criminal Law

How can we know whether we are punishing the same corporation that committed some past crime? Though central to corporate criminal justice, legal theorists and philosophers have yet to address the basic question of how corporate identity persists through time. Simple cases, where crime and punishment are close in time and the corporation has changed little, can mislead us into thinking an answer is always easy to come by. The issue becomes more complicated when corporate criminals undergo any number of transformations—rebranding, spinning-off a division, merging, changing ownership, changing management, swapping lines of business, etc. These changes are common among all corporations, including those trying to conceal or limit liability for past crimes. This article takes a first step toward developing a workable and philosophically satisfying theory of corporate personal identity and discusses its prospects for fulfilling the retributive, rehabilitative, and deterrent purposes of criminal law.


Introduction
Imagine for a moment that we lived in a world where people were much more. . . fluid. In it, criminals desperate to confound justice would have some strategies available. One option in this fantasy world would be for criminals to divide themselves into two separate people, only one of whom would inherit the criminal taint. Another would be to use the latest technology to alter their personality, perhaps removing any trace of criminal disposition. Or, if neither of those options were satisfactory, a criminal might instead find an innocent person and, with or without consent, fuse with them to form a single, morally ambiguous composite. Sci-fi authors and philosophers can take their time deciding who, if anyone, we should punish after these changes. The fluid world, if it ever arrives for human beings, is far off. Business scholars do not have that luxury (Heenan 2004, p. 100).
The "identity principle" is the basic tenet of criminal justice that whoever did the crime (and only that person) should do the time. Anything else would entail some form of vicarious liability-punishing A for the crimes of B-which would be anathema the most basic values of due process. 1 For individual criminals, the identity principle is pretty straightforward to put into effect. DNA, facial features, fingerprints, social networks, social security numbers, and the like uniquely pick out human beings from birth to grave. But identity is not such a simple matter for corporate people. The sorts of transformations that for individuals are limited to science fiction and philosophers' thought experiments-fissions, fusions, transplants (Parfit 1971, pp. 1-5)-are daily life for corporations-spin-offs, mergers, acquisitionsespecially those trying to manage their liabilities (Solomon 2013). Intuitions about how identity works can start to break down when these processes are in play. But we cannot simply shrug our shoulders at the novelty of such cases. Criminal justice requires us to have a defensible theory of corporate personal identity so we can consistently judge whether the corporation we punish today is the one that committed the crime in the past. 2 It is surprising, then, that theorists in philosophy, law, and business have had relatively little to say about the matter (Rovane 1998, pp. 135-208). This article takes some first steps to fill the gap. 3 It may seem like a fool's errand. Any theory of corporate identity should cohere with our intuitions; at bottom, we have nothing else to rely on (Killoren 2010, pp. 1-35;Shaw 1980, p. 134). But in the absence of robust, welldefined intuitions about complex cases of corporate identity, what could we use to build the theory? This article seeks to systematize and extend a theory premised on the rough intuitions that we do have about simpler cases of corporate identity.
It starts with the philosophy of identity. Though philosophers have yet to talk about corporate personal identity, they have done a good deal to map out the conceptual space for thinking about the identity of various kinds of entities across radical transformations. With the framework from philosophy in hand, the article turns to the most recent data in cognitive science about how people intuitively trace the identity of groups across time. From these data, the contours of a theory of collective identity emerges, according to which collective identity fixes to normatively salient features of the collective. The article then extends and operationalizes this implicit theory for the case of corporate crime, drawing on a wide range of literature in compliance, organizational science, management, and marketing. According to the resulting theory-Criminal Essence Theory-a present-day corporation is identical to a past corporate criminal if (and only if) it retains whatever organizational shortcoming led the past corporation to commit the crime in question. The article closes by discussing how the theory fares with respect to the diverse purposes of criminal law.

Brief Note on Method
By attempting to engage both philosophers and legal academics, this article inhabits an awkward methodological space. Members of the two groups often have very different perspectives on corporate personhood. This has obvious implications for trying to develop a theory of corporate personal identity. In line with long-standing common law jurisprudence, lawyers tend to think of corporate personhood as a legal fiction (Dewey 1926, pp. 655-673;Laufer 1994, p. 650). They think that for pragmatic reasons we pretend that corporations are persons-who can own property, sue, and be sued, etc.-though they really are not. Among philosophers, the issue of corporate personhood is a more open and contestable question of metaphysics. Many philosophers think that, at least so far as moral personality is concerned, complex corporations often have it (Pettit 2007).
Both groups, the philosophers and the lawyers, should agree that we need a theory of corporate personal identity if we are to hold corporations criminally liable. It is an unshakeable commitment of our criminal and moral law that the innocent should not suffer punishment. Some theorists see it as one of the distinctive features of criminal liability, as opposed to other sorts of legal liability. 4 Respecting this maxim requires that we have some framework for assessing which present-day defendants are identical to those who committed past misdeeds. Lawyers and philosophers, however, will assess potential frameworks by different metrics. Lawyers, thinking that corporations are only fictional people, will be more open to a range of possibilities, and will try to identify the framework that best fulfills the purposes of corporate criminal law, e.g., crime prevention. Those philosophers who think corporations really do have moral personhood will want to find the correct framework, that which truly tracks the persistence of corporate moral personhood.
Fortunately, as I argue below, these two perspectives converge. The theory of corporate identity that best promotes the various goals of corporate criminal law is the same theory that best coheres with our other commitments about the nature of corporate personal identity. The different camps will likely find themselves engaged by different parts of the article. In particular, as I develop and evaluate my theory of corporate identity, philosophers who reify corporate personhood will probably find that the "Cognitive Science of Corporate Identity" and "Retribution" sections speak most to them. Lawyers may prefer to skip to the positive statement of my view in the "Criminal Essence Theory" section and the discussion of how it engages corporate incentive structures (the "Rehabilitation" and "Deterrence" sections). Those who, like myself, work at the intersection of law and philosophy, will, I hope, find themselves engaged by it all.
Though I try to steer a metaphysically non-committal course so far as corporate personhood is concerned, I will not be metaphysically neutral about everything having to do with corporations. In particular, I will help myself to the fact that, even if they are not persons, corporations can be causally efficacious, particularly as concerns how the individual person constituents composing them behave. To deny that organization-level features can influence individual behavior (as described in the "Business Literature on Corporate Criminal Essence" section) would be to reject decades of research and data in organizational psychology and sociology. Doing so would be a mistake, and unnecessary, even for someone inclined to reject corporate personhood. Personhood is not a precondition for causal efficacy.

Clarifying the Question: Diachronic Corporate Personal Identity
Criminal and moral responsibility presupposes personal identity. John Locke kicked off the modern interest in that premise (Locke 1694), and philosophers have accepted it, in one form or another, ever since (Butler 1763;Reid 1785;Parfit 1971). Personal identity is central to moral responsibility in two respects. The first is "synchronic," which basically refers to identity at a given time-how can we tell which things in the world make up a person, and by extension what the person is responsible for. Philosophical accounts of synchronic identity for natural people cover a broad range (Parfit 1971, pp. 11-12;Carter 1989, pp. 1-14). The question comes up in criminal law for natural people when we have to distinguish causal forces that are identifiable with the agent (e.g., her volitional acts) from those that are not (e.g., intervening acts of others, psychotic episodes, epileptic fits, remote and unforeseeable turns of events). We generally only punish people for the things they, rather than some other external force, did (Lafave and Scott 1986, § 3.2(c)).
Synchronic identity issues are common in corporate criminal law as well. There is not much controversy about what makes up a corporation, e.g., a corporate charter, shareholders, directors, and employees. But things can get complicated when we turn to causal forces and want to know when the corporation, as opposed to merely one of its human-sized parts going on a spree of her own, has done something. Academics have had a lot to say on that issue, and criminal courts resolved long ago in favor of the doctrine of respondeat superior. 5 Respondeat superior puts some limits on the synchronic identity of corporations-they are identified with their employees (and the things they do) only so long as the employees are working within the scope of their employment and intend to benefit the corporation (O'Sullivan 2016, p. 157).
A different sort of personal identity, the topic of this article, is equally important for criminal law-"diachronic" personal identity. Diachronic identity is identity over time (Gallois 2016); in other words, when can you say that a person at one time, and a person at a different time, are one and the same continuous person (French 2016, p. 579). Dystopian sci-fi movies aside, 6 in criminal law, we punish someone only after she commits a crime. This immediately raises the diachronic identity question-even if we know who committed the crime and who we are about to punish, how can we be sure they are the same person? We rarely worry about this for individual human defendants since we generally accept that continuity of personhood follows bodily continuitysame fingerprints, same DNA, same face, same person (Olson 1997). Serious concerns about diachronic identity for natural person defendants only arise in rare cases of dissociative identity disorder (Birmingham 1998, pp. 117 − 24;Braude 1996;Hacking 1995;Radden 1996;Wilkes 1981), personality-changing brain trauma (Harlow 1868), and the recesses of philosophical fancy where people's minds and bodies can split and fuse (Parfit 1971).
Diachronic identity is equally important, but much harder to conceptualize, when it comes to punishing corporate defendants. While we may question the relevance of philosophers' fission and fusion scenarios for natural persons, the active world of corporate spin-offs, mergers, and acquisitions makes fission and fusion impossible to ignore for corporate persons. The corporate equivalent of splitpersonalities or radical changes in personality are common too, as corporations may have separate lines of business or wholesale changes in management.
What is worse is that we do not have the same reservoir of strong intuitions about corporate identity as we do for individual identity. We seem fairly confident about the extremes. If a corporation has undergone few changes since it committed a crime-no major alterations in structure, purpose, employment base, ownership, management, public brand, headquarters, etc.-then whatever corporation currently has those unchanged basic features is the same as the one that committed the crime. At the opposite end of the spectrum, if the corporate criminal has fundamentally altered all these features, it has probably changed its identity, and there may be no present-day corporation it would be appropriate to punish. The trouble is in finding where to draw the line between these poles-how much change, and which changes, can a corporation undergo without changing its identity and shedding its responsibility for past crimes? Is there, as Aristotle would have put it, a clear distinction between "accidental" (non-identity-altering) and "essential" (identity-altering) changes? (Matthews 1990, pp. 215-252). Or are there some more complicated interrelationships between changes that, in the right constellations, can result in a change of corporate identity?
These are the questions of diachronic personal identity for corporations, and the issues on which this article focuses. Wherever the article refers to corporate "identity" below, it means diachronic personal identity. The article will not attempt to consider every possible change a corporation could undergo. But what it does address-mergers, spinoffs, and changes in ownership, management, and employment-should be enough to set up a framework for thinking about what it does not.
There are other forms of legal liability that frequently ignore the identity principle. But in the criminal law, the identity principle is a bedrock commitment. In principle, criminal law could try to abandon the identity principle, at least so far as corporations are concerned. But this would be a dangerous precedent. The various domains of criminal law are not hermetically sealed from each other, and developments in one domain can bleed into others. How we treat corporations in criminal law has implications-if not in theory and policy, in terms of the organic development of criminal law-for how we treat ourselves. In any case, abandoning the identity principle in corporate criminal law would be a solution in search of a problem-as argued below, criminal justice policy favors sticking to the identity principle for corporations.

Groundwork for a Theory of Corporate Identity
The ancient Greek hero Theseus had a ship (Plutarch, First Century A.D.). After every adventure, Theseus would have his crew inspect the ship and replace any overly worn planks, sails, or riggings. After many years, not a single original splinter, nail, or thread of the original ship remained; everything had been replaced. Did Theseus still have the same ship, or a new one? Would it still be the same ship if Theseus sold it to a merchant who then used it for commerce? Does it matter that one of Theseus' greatest admirers managed to collect each ship part as it was discarded and painstakingly reconstructed "the original ship of Theseus"? This is a famous paradox, and people's intuitions differ about which, if any, of the ships is the same ship Theseus had at the start. The paradox arises because ships are complex entities. Radical alterations (like building them again from new materials) that intuitively would change their identities can occur gradually, and then intuitions weaken. Corporations are similarly complex. Philosophers have grappled with identity paradoxes for millennia. At some point, we may just want to throw up our hands and say, "Who cares? Every interested party knows which ship to board each morning." When it comes to corporate identity, we cannot dismiss the question so easily. For one thing, corporate stakeholders care a lot about whether their corporation is on the hook for some past criminal misconduct. For another, any fair-minded citizen should care that the criminal justice system targets the right defendant.
One benefit of identity paradoxes like the Ship of Theseus is that generations of philosophers have developed a framework for thinking about identity across a range of cases. That philosophical work has also provoked some preliminary work in cognitive science about group identity. The approach some philosophers take, and the one adopted here, is to start with simpler cases where we have firmer intuitions about diachronic identity for the type of entity under consideration. From these intuitions, we can try to disentangle an implicit, general theory, which, if fleshed out, may lead to answers in more complex cases. This section begins with some general philosophy of identity. It then turns to recent cognitive science about collective identity and the general theory it implies. Lastly, it draws on scholarship in organizational psychology and business for insights that can help operationalize the general theory as applied to corporations.

The Philosophy of Personal Identity
Philosophers have developed various conceptual frameworks for identity that can serve as the starting point for the corporate case. Aristotle, for example, distinguished between "accidental" and "essential" traits, where only a change in the latter brings about a change in identity. These concepts offer a helpful way to frame the present inquiry: Which of a corporation's traits are essential, and which are merely accidental?
Even with the distinction between accidental and essential traits, there is still something conceptually bizarre about the sorts of merging and spin-off cases that can arise with corporations. Fusion and fission cases (as they are called in the philosophical literature), where entities combine or divide, respectively, raise familiar puzzles in the philosophy of identity. If, after a merger, the successor is identical to one of the its predecessors, what happens to the other predecessor? Surely it is not simply extinguished. Whatever makes the successor corporation identical to one predecessor could make the successor identical to the other predecessor too. But this sounds like it entails a logical contradiction. Identity is a transitive relationship, 7 and this network of identity relationships seems to violate transitivity: the two predecessors are identical to the successor, but not identical to each other. 8 At this stage, we need an open-ended framework for theorizing identity that can accommodate these sorts of cases without necessitating any particular answer. Philosophers of identity have worried about fission and fusion problems for a long time, initially with respect to objects (Lewis 1971(Lewis , 1986, and later with respect to natural people (Nagel 1979). Some philosophers think they have found a solution in an approach to identity called "four-dimensionalism." (Parfit 1971;Lewis 1976;Noonan 2003, pp. 139-142). Fourdimensionalism views entities as things extended through time. Accordingly, four-dimensionalists prefer not to talk about the identity of an entity at a single time, but of a thing existing through time. This gives them the resources to accommodate fission and fusion cases.
Suppose two entities merge in such a way that they are both seemingly identical to the third composite entity that results. Four-dimensionalists would describe what happened in terms of three entities. The first two are the initial entities, E1 and E2. These each occupy their separate spatiotemporal paths until the moment of merger, after which point their paths overlap. The third entity is the composite, E3, which, according to the four-dimensionalist, existed before the merger, but in two parts as E1 and E2. At the point of merger, the spatiotemporal paths of the two parts of E3 join and E3, thereafter, occupies a single path in space-time. 9 Four-dimensionalism does require some mental gymnastics. For example, note that after the merger, there may be three entities inhabiting the same path through space-time: E1, E2, and E3. But this kind of basic accounting is a small cost for having the tools to speak coherently about identity for entities that can merge and split.

Cognitive Science of Corporate Identity
Embracing a four-dimensionalist framework for thinking about corporate identity is only a start. If we must be open to the possibility that corporations can merge, spin-off, and undergo other transformations while retaining their identity, we must equally entertain the prospect that they can sometimes go through these changes and lose their identity, i.e., become different corporations. To figure out whether a corporation survives these modifications and, in the case of spin-offs, to determine which, if not both, of the resulting corporations it survives as, we need to know what the essential features of a corporation are. Any successor corporation that retains the essential features of a predecessor corporation will be identical to it.
How can we determine what are the essential features of a criminal corporation? We really only have two data points available to kick off this sort of inquiry: observation and intuition. We can observe criminal corporations (imaginatively through hypos, if not in fact) as they change various of their features, and gauge our intuitions about when their identity shifts. A theory of corporate identity may be implicit in these data points, or, to the extent our intuitions conflict, reflective equilibrium could help reconcile them (Rawls 1971). Not everyone will agree with this approach. From those who, as fictionalists about corporate moral personality, are feeling baffled right now, I ask for patience. It turns out that the theory of corporate identity that results from this process makes good criminal justice policy, so far as rehabilitation and deterrence are concerned. To those who are realists about corporate moral personality, I humbly submit that I know of no other way forward. Identity is not a directly observable relation, and no available theory of corporate personhood, so far as I am aware, has any answer to the question of diachronic identity.
This still leaves the question of whose intuitions we should use. I distrust my own. They have been poisoned by years of work on corporate criminality and my personal investment in the theories I have developed. We could ask other philosophers for their intuitions, but these raise similar concerns. What remains are ordinary folk intuitions. Building a theory off these is particularly attractive in criminal justice. Criminal law is a law for all people, not just for the theorists and philosophers. It largely derives its legitimacy and its efficacy from the legitimacy it is perceived to have (Robinson 2013). This, in turn, depends on how well it coheres with ordinary people's intuitions about desert and responsibility. 10 Cognitive scientists have some insight into how people intuitively think about diachronic identity generally (Rips et al. 2006). Early data confirm that people really do distinguish between accidental and essential traits (Strohminger and Nichols 2014). There is a pattern to how they make the distinction. The sorts of traits that people judge to be accidental tend to be superficial, while essential traits tend to be deeper and possibly hidden from view (Blok et al. 2001;Hall et al. 2003). This difference is about more than mere location in physical space, inner versus outer. The traits we judge to be essential traits are those hidden features that are causally efficacious in determining how the entity behaves (Rips and Hespos 2015).
More recent studies have focused on how normative valence influences our intuitions about whether a trait is essential or accidental. Studies about individual human beings suggest that normative valence plays an important role (Newman et al. 2014(Newman et al. , 2015. The newest data extend these results to collective entities too (De Freitas et al. 2016). For example, if a group undergoes a change that causes it to lose its positively valenced traits (e.g., a nation loses its egalitarian policies), then people are likely to judge that it has lost its identity. More importantly for present purposes, cognitive scientists have found that "[i]f an entity is explicitly described as having a bad essence, it will be viewed as losing its identity if it improves" (p. 12). 11 Cognitive scientists asked subjects, for example, about variations on a scenario that described a school abandoning its initial mission of teaching Nazi ideology in order to focus on traditional subjects. The studies conclude that subjects are more likely to judge that a collective loses its identity if it undergoes a change that fundamentally alters its negatively valenced traits.
Combining the insights about the importance of causal efficacy and normative valence suggests an implicit theory of identity applicable to criminal corporations: The essential trait of a criminal corporation is whatever trait caused it to commit the crime. In the context of a criminal trial, this is likely to be the most salient, normatively valenced, and causally efficacious trait of the corporate defendant. For short, I refer to this trait below as a criminal corporation's "criminal essence." If a criminal corporation retains its criminal essence to the present day, our intuitions should be that the corporation retains its identity. If, however, through some transformative process, the corporation improves upon and eliminates its criminal essence, the data suggest our intuitions will be that whatever corporation emerges is no longer identical to the corporation that committed the crime.

Business Literature on Criminal Essence
To integrate the insights of cognitive science into a workable theory of corporate identity, we need an understanding of what sorts of ground-level features of a corporation could account for its criminal essence. Business scholars have a lot to say about the organizational causes of corporate misconduct and what changes are more likely, which less, to alter them. Like cognitive scientists studying diachronic identity, business scholars have come to distinguish between superficial corporate elements, such as external branding symbols (Simoes 2005; Perez and del Bosque 2012; Theunissen 2014), and deeper, causally efficacious, intrinsic traits (Balmer 1998;van Reiel and Balmer 1997;Kiriakidou and Millward 2000). This section considers what specifically those intrinsic traits might be.
One of the initial premises of systems theory is that "[o] rganizations are systems. . . not just aggregations of individuals" (Fisse and Braithwaite 1988, p. 479;Corneliessen et al. 2007, p. S8). This means that it is often the corporate organization itself, rather than the individuals within it, that is causally effective. Individuals within a corporation adapt to its procedures, rules, and culture once they join it (Coleman 1990, p. 427). As a result, "personnel changes will seldom lead to real changes in the organization's behavior and work processes" (Lederman 2000, p. 688). This is true throughout the corporate hierarchy, from managers to the assembly line (Ermann and Lundman 1996). Of course, this assumes only gradual changes in personnel. Wholesale replacement of large groups of employees may alter causally effective corporate traits, but probably only because the changeover makes changes in other, essential traits possible. As a general rule, though, change in the personnel of a criminal corporation will not reflect a change in its criminal essence. Conversely, if it is the organization rather than the personnel that matters, continuity in personnel need not necessarily entail that a criminal corporation has retained its criminal essence.
The same is true of a corporation's shareholders-continuity or change in ownership is not indicative of a change in a corporations' criminal essence. Shareholders do Footnote 10 (continued) the processes of reflective equilibrium and is no longer with us. Had the philosopher kings had their way, deodand may have had a speedier demise, and that would probably have been a good thing. But any interposition between folk intuition and criminal justice that goes beyond speeding the steady course of reflective equilibrium raises a worryingly undemocratic specter-not that of the philosopher king, but that of the philosopher tyrant. 11 To test this effect, scientists presented subjects with one of four scenarios describing a school in Nazi Germany. In two of the scenarios, the school initially shirks traditional academic subjects to focus on Nazi ideology, and, after a turnover in administration, either continues teaching Nazi ideology or switches to teach exclusively academic subjects. In the other two scenarios, the school initially teaches traditional academic subjects and, after the administrative turnover, continues with academic subjects or switches to Nazi ideology. theoretically have the power to affect the internal operation of a corporation (Alexander and Cohen 2011). For larger corporations, shareholder power in theory rarely reflects power in fact. Shareholders are a dispersed group with divergent interests, making coordination difficult (Stout 2002(Stout , p. 1191. Even when coordinated, shareholder power to influence board decisions or composition is limited; their power to influence managers even more so. This is not to deny that shareholders can have an effect, especially large institutional shareholders or shareholders of smaller firms. But, even in these cases, since shareholders are generally not involved in the operation of a corporation, they have little correlation with the causally effective normative traits that define a criminal corporation's essence.
If the identities of personnel and shareholders are not reliably correlated with the presence or absence of a corporation's criminal essence, what is? The short answer is, we do not know for sure. But there are some compelling hypotheses available. Corporate culture or ethos is one supra-individual feature that has attracted the attention of academics in business (Corneliessen et al. 2007, o. S7;Melewar and Karaosmanoglu 2006) and law (Bucy 1991(Bucy , pp. 1099(Bucy -1100Fisse 1991;Foerschler 1990Foerschler , pp. 1300Foerschler -1302Moore 1992, pp. 759-760), and policymakers (U.S. Department of Justice 2008, § 9-28.000). The idea here draws on the insight mentioned at the start of this section that organization-level features influence how individuals within a group perform, including whether they commit crimes (Needleman and Needleman 1979). For example, a high-pressure environment oriented toward quotas and production goals with little emphasis on legal or ethical limits can foster malfeasance, even among individuals not otherwise disposed to it (Reckard 2013). Factors that impact a corporation's ethos include its hierarchy, goals and policies, treatment of prior offenses, efforts to educate employees on compliance with the law, and compensation scheme (Bucy 1991(Bucy , p. 1101.
Another trait that understandably receives a lot of attention is a corporation's compliance program (Buell 2011, p. 93). By definition, compliance programs seek to prevent corporations from violating the law (Baer 2009, p. 956;Laufer 1999Laufer , p. 1345. Compliance programs are related to, but ultimately different from corporate ethos. They focus on formal operation procedures designed to prevent, detect, and remedy criminal conduct within the corporation. The sorts of techniques currently emphasized in the compliance literature are mostly commonsense: "promulgation of codes of behavior, the institution of training programs, the identification of internal compliance personnel and the creation of procedures and controls to insure company-wide compliance with legal mandates" (Rostain 2008, p. 467). But they need not stop there. Some scholars, including William Laufer, are calling for a more "progressive," data-driven and technically sophisticated approach to compliance (Laufer 2017 (Laufer and Robertson 1997, pp. 1029-1030, the scientific study of compliance is still in its infancy. There is good reason for optimism that with each passing year, we will know more about what sorts of programs are effective at neutralizing corproate criminogenic traits.

Criminal Essence Theory
It is time to weave together these threads from philosophy, cognitive science, and organizational psychology. We know from philosophy that we can coherently speak of diachronic corporate personal identity, even in contexts where mergers and spin-offs may seem to lead to transitivity failures. We also know from cognitive science that we intuitively see corporate identity as being tied to causally efficacious, normatively valenced traits. In the context-corporate criminal trials-that trait is whatever feature of the corporation was causally responsible for its criminal conduct; what I call its "criminal essence." It stands to reason that the theory of corporate identity that should cohere best with our intuitions is the "Criminal Essence Theory": A past corporate criminal is identical to, and only to, the present-day corporation or corporations, if any, that retains or retain the same criminal essence. According to the Criminal Essence Theory, if a transformation somehow enables a criminal corporation to shed its criminal essence, the resulting corporation would not be identical to the criminal corporation.
Applying the Criminal Essence Theory to determine which, if any, present-day corporation is identical to a past corporate criminal is a two-step process. The factfinders must first determine what trait-whether a poisonous corporate ethos, a gaping compliance deficiency, or something else entirely-was causally responsible for the corporate criminal's misconduct. That is to say, the factfinders must isolate the corporation's criminal essence. Second, the factfinders must determine whether that trait continues in any present-day corporation. If it does, whatever corporation (possibly more than one) has the trait is identical to the criminal corporation (so long as there is causal continuity of the trait from the criminal to the present corporation). If it does not, no presentday corporation is identical to the corporate criminal. 12 12 Criminal Essence Theory leaves open the question of what to do if a corporation is charged while it is reforming its criminal essence and before it has finished the job. Any effort to draw a bright conceptual line-when precisely has it reformed enough to become a new corporation-would bump up against familiar problems in the philosophy of vagueness. As will be clearer after the discussion of deterrence drawing the line closer to the start or the end of the process of reform will influence the trade off between fulfilling criminal law's rehabilitative and deterrent purposes. Such line-drawing is beyond the scope of this article.
The test is simple and predictable across a range of cases, so long as the organizational traits that caused the crime are discernable. While a change in personnel will not generally extinguish corporate criminal essence, it may where it causes or is accompanied by modifications in corporate ethos or compliance (assuming one of those was causally responsible for the crime). When a corporate criminal spinsoff a division, it will be identical to whichever of the successors (possibly both) inherits the criminal essence. And when a corporate criminal merges with another corporation, its identity will persist into the successor if the criminal essence survives the merger. But, where the ethos or compliance programs of the counterparty corporation overwhelm and replace that of the corporate criminal, the merger may extinguish the criminal identity. 13

Putting the Criminal Essence Theory to Work
I developed Criminal Essence Theory in the last two sections to be a theory of identity for criminal corporations that would best cohere with our intuitions. The original motivation for finding a theory, and the reason we should not rest without one, is that criminal justice requires it. Recall that the identity principle prohibits us from punishing a defendant (incorporated or otherwise) for a past crime unless it is the same person who committed it. Now that I have a theory in hand, I should check how well it can fulfill the purposes of criminal law. It would be troubling if it turned out that there were a bad fit between the best theory of corporate criminal identity and these purposes. Fortunately, Criminal Essence Theory is well-positioned to advance the goals of corporate criminal law.
Though criminal law has many objectives, its most basic purposes fall into three general categories: retribution, rehabilitation, and deterrence. 14 Individual criminal law theorists often have purposes they personally favor, sometimes to the near exclusion of others. (Moore 1987;Huigens 2002, p. 5;Alexander and Cohen 2011, p. 11;Cahill 2011). 15 Additionally, I noted above that one's view on the metaphysics of corporate personhood will likely influence which purposes one finds most relevant to corporate criminal law. Those who believe in the moral personality of corporations will want to pay most attention to the next subsection, on retribution. Those who are skeptics will likely be less moved by retributive considerations, and may find the pragmatic arguments on rehabilitation and deterrence more compelling. I will not pick between these purposes. Rather, I will show that the Criminal Essence Theory can advance all three of corporate criminal law's most basic goals.

Retribution
Retributivists think that punishment should give criminals their just deserts (Moore 1987), and that the justice deserved is proportional to the seriousness of the offense (von Hirsch 1993, pp. 6-19). The most straightforward way to assess Criminal Essence Theory in retributive terms would be to compare it to the right theory of corporate desert. While some philosophers have made heroic efforts to develop such a theory (List and Pettit 2013), no proposal has found general acceptance. In any case, available theories of corporate desert focus on the question of synchronic identity: When do the constituents of a corporation commit a wrong that is attributable to the corporation? They do not answer the diachronic question: Assuming a corporation did something that deserves punishment, how do we tell whether that same corporation is still around to punish today?
In the absence of a theory that can help with the diachronic question, I have argued that our best option for vindicating corporate criminal law's retributive goals is to align the law with our intuitions on the matter. 16 By design, Criminal Essence Theory incorporates the best data we have about how to achieve the retributive purposes of corporate criminal law. According to the identity principle, we should punish a present-day corporation when and only when it is the same corporation that committed the crime. The Criminal Essence Theory builds on what we know of our intuitions about when a corporation satisfies this condition.
Even if Criminal Essence Theory is retributively appropriate within the limited scope of corporate criminal law, we should make sure it does not introduce retributively unacceptable spillover to other areas of law. I see two possible concerns. One is that Criminal Essence Theory will produce different results for corporate defendants than our current criminal law allows for individuals. For example, under Criminal Essence Theory, a criminal corporation could extinguish its criminal liability by removing the organizational vulnerabilities that led it to commit its crime. Individuals do not have this option. This raises the possibility that 13 The Department of Justice has exhibited some interest in implementing such an approach informally. See, e.g., U.S. Department of Justice (2008 Criminal Essence Theory gives corporations the option of a free pass that is unfair vis-à-vis individuals. It is not clear whether this criticism gets off the ground since corporations and individuals are such different types of entities-the first is extremely malleable, the other has familiar biological constraints. But if, as cognitive science seems to suggest, we think about corporate responsibility in ways similar to how we think about individual responsibility, I need a response. 17 Though individuals cannot, through reform, escape criminal liability as a matter of law, we do seem to think reform can mitigate their moral responsibility. "I'm a different person now" is a common reply to a moral accusation-"I am no longer the irresponsible youth who would hurt you in that way." What such replies usually mean is that the speaker no longer has whatever motivations or attitudes led to the misconduct. Assuming the speaker can say this credibly, we often think it carries moral weight.
Why has this intuition about individual moral responsibility not found its way into the law of individual criminal liability? To some extent, I think it has. This may be part of what is behind statutes of limitations for criminal liability. After a long enough period of time without a recurrence of the same misconduct, the statute of limitations runs and a person becomes immune from prosecution. One intuitive explanation is that as time passes without re-offense, it becomes increasingly likely that an individual who committed a past crime has relevantly different motivations and attitudes, i.e., is a now "different person." There are also epistemological barriers to implementing something like the Criminal Essence Theory for individual criminals. It is very difficult to tell when an individual's motivations and attitudes really have changed. While we can get in and look for changes in the organization-level features that constitute criminal essence for corporations, we do not have the technology to do the same for individuals. We can assess compliance programs and compensation metrics, but know very little about the neurological bases of individual criminal essence. Lacking the necessary neuroscience, the criminal law must turn to less reliable but easier-to-observe indicia of diachronic individual identity, like continuity of physical form and behavior over long periods of time.
A second retributive spillover concern arises from the fact that Criminal Essence Theory would allow corporations to escape punishment by reforming themselves. Retributivists often seem to think that, in an ideal world, no crime should go unpunished. If Criminal Essence Theory allows some corporate crimes to go unpunished, that could leave a retributive residue in need of response.
Part of the retributive residue may be addressed by convicting and punishing individuals. Though it has not been a focus of this article, I do not mean to suggest that corporate punishment should be the exclusive white-collar remedy. Individuals must face criminal justice as well, regardless of whether the corporations of which they are a part have sufficiently changed their identities to avoid their own punishment.
This response can only go so far. There have been cases where a corporation commits a crime even though no individual within it has committed a crime (O'Sullivan 2016, pp. 176-177). In these cases, should the corporation subsequently reform its criminal essence, thereby escaping liability, there would be no suitable object of punishment on the Criminal Essence Theory, corporate or individual. A crime would go totally unpunished. This could initially seem like something that would worry a retributivist, but it must be balanced against another retributive concern-avoiding punishment of innocents. 18 Those who ultimately bear the brunt of a corporate sanction are usually innocent corporate stakeholders, like individual shareholders and employees (Alschuler 2009(Alschuler , pp. 1366(Alschuler -1367. That is retributively problematic and something retributivists should want to avoid unless there are compelling justifications. I turn to these now in considering how Criminal Essence Theory advances criminal law's interests in rehabilitation and deterrence.

Rehabilitation
We hold corporations to account for many of the most common corporate crimes-e.g., securities fraud, anti-competitive practices, environmental violations-because of the negative consequences that conduct has for our economic and social wellbeing. Preventing such crimes is an uncontroversial purpose of corporate criminal law. What may be more controversial is how best to accomplish this. According to rehabilitation theorists, one effective approach is to structure criminal law so that it promotes the reform of criminals. Criminal Essence Theory does just that.
While the law could reform criminal corporations after the fact through sentencing (Diamantis 2018), it would be more efficient if the law could encourage corporations to reform their vulnerabilities ex ante. The problem is that corporations have strong incentives not to take the steps necessary to uncover and reform criminal essence. Monitoring and internal investigation raises the probability that vulnerabilities and misconduct will be detected, and, if detected, leaked to authorities. It may be safest for the corporation to remain blissfully ignorant. Since corporations cannot fix problems they do not know about, we should expect relatively little self-initiated reform unless the law can overcome these disincentives.
Criminal Essence Theory flips these incentives around, giving corporations' strong reasons to detect and remedy any criminal essence they may have. Once the trait is discovered, corporations have a double incentive to fix it-to prevent future misconduct and to nullify their liability for the past misconduct. By remedying its criminal essence, a criminal corporation would, according to the Criminal Essence Theory, change its identity for purposes of criminal law. The resulting corporation, not being identical to the criminal corporation, would emerge free of criminal liabilities. This potential benefit also gives corporations a robust incentive to detect compliance weaknesses in the first place. Once they do, they have taken the first step to eliminating their liability.
These incentives to detect and remedy compliance vulnerabilities would persist to the successors of spin-offs and merger scenarios. Any successor corporation that inherits a criminal essence from a predecessor will want to find and fix it. If it fails to do this, it risks facing liability not only for any future violations the criminal essence causes, but also for the past violations the essence generated with the predecessor. This reasoning extends to innocent counterparties in merger and consolidation cases even as the terms of combination are being negotiated. Should they discover a problem, innocent counterparties will want to cover some of the costs of subsequent reform through more favorable terms for the merger or consolidation.

Deterrence
Another common way of thinking about how to structure the criminal justice system so as to prevent corporate crime is deterrence theory (Huigens 2002, p. 5). According to deterrence theory, the purpose of criminal law should be to threaten sanctions that will disincentivize criminal conduct (Alexander and Cohen 2011, p. 11). Criminal Essence Theory may seem to be weakest with respect to this purpose. Will corporations really be deterred from misconduct if they can commit a crime, reap the benefits, and then insulate themselves from liability by shedding their criminal essence?
To some extent, this is less a weakness of the Criminal Essence Theory than a reflection of the basic tension between criminal law's rehabilitative and deterrent purposes (Robinson 2017, pp. 91-92). Fully achieving one can only come at the expense of fully achieving the other. If we punish every criminal violation regardless of efforts at subsequent reform (as would be optimal from the perspective of deterrence), we reduce the incentives criminals have to better themselves. Alternatively, if we allow criminals to reduce their prospective sanction by reforming themselves, we risk diminishing their incentives not to commit crime in the first place.
Something like the Criminal Essence Theory holds out the possibility of striking a balance between deterrent and rehabilitative purposes. Under Criminal Essence Theory, crime would still be risky for corporations. If authorities detect misconduct before a criminal corporation reforms its criminal essence-either because the corporation failed to detect it or did not have time to reform-then the corporation would be subject to prosecution and the sanctions that follow. The risk of sanction under Criminal Essence Theory is indeed lower because the corporation may have time to reform before detection. But if the threat of sanction becomes too low for purposes of deterrence, the criminal justice system has the option (at least in theory) of dialing it up by raising the severity of the sentence a corporation will face if successfully prosecuted. So we should still be able to approximate optimal deterrence while adopting Criminal Essence Theory.
There is one very aggressive form of corporate gamesmanship that Criminal Essence Theory could incentivize. A corporate mastermind might find some way to induce corporate crime, reap the profits, and then reform the organization immediately after, extinguishing its liability. That seems like a troubling recipe for generating risk-free criminal profits. Fortunately, under Criminal Essence Theory, it would be a rare, oneoff opportunity for corporations. Complete corporate reform is crucial to the success of the gambit. If a corporation truly loses its criminal essence, it will have removed any propensity toward similar misconduct, and along with it the opportunity to benefit from a repeat performance. If whatever trait led the corporation to try to game the criminal justice system remained, the corporation would not have lost its criminal essence and would remain liable for the original crime. The success of the maneuver forecloses the possibility of repeating it.

Conclusion
According to the Criminal Essence Theory proposed here, a criminal corporation's identity attaches to the organizational trait that led to its misconduct, the corporation's "criminal essence." Should that trait disappear, perhaps because the corporation detected and reformed it, the corporation would emerge for purposes of criminal law with a new identity, free from criminal taint. But so long as the criminal essence remains, regardless of any other transformations the corporation may have undergone, the criminal identity persists. Any present-day corporation that inherits that essence-whether through a spin-off, merger, or just continuous operationwould, under Criminal Essence Theory, be liable for the criminal conduct it caused.
This new approach aligns with folk judgments about corporate identity, so it excels in retributive lights. It also provides strong incentives for criminal corporations to reform their criminal essence, thereby achieving criminal law's rehabilitative goals. While the proposal might weaken the deterrent effects of current corporate criminal law, the loss on this front should be modest.