key: cord-0808821-njkh9neq authors: Husak, Douglas title: Retributivism and Over-Punishment date: 2021-09-04 journal: Law Philos DOI: 10.1007/s10982-021-09422-w sha: d89eb1da4fc268329c7ff0257867697af04536d5 doc_id: 808821 cord_uid: njkh9neq Lately it has become a commonplace to complain about the injustice of mass incarceration. I share the sentiment that this phenomenon has been an injustice. But it also has become orthodoxy to allege that the acceptance of a retributive penal philosophy has been one of the chief factors that has brought about mass incarceration in the first place. As a self-proclaimed retributivist, I find these allegations to be troubling and unwarranted. The point of this paper is to take steps to rebut them. I begin by making four conceptual points about retributivism. If I am correct, retributivism comes in countless flavors, and the particular variety to which I am most attracted can be applied to show why some punishments should be less severe than those presently imposed. Next I argue that many persons deserve less punishment than our legal system currently inflicts. Reflection about whether perpetrators should be afforded a complete or partial defense reveals retributivism to be less punitive than conventional wisdom would suggest. (Accepted 7 August 2021) ABSTRACT . Lately it has become a commonplace to complain about the injustice of mass incarceration. I share the sentiment that this phenomenon has been an injustice. But it also has become orthodoxy to allege that the acceptance of a retributive penal philosophy has been one of the chief factors that has brought about mass incarceration in the first place. As a self-proclaimed retributivist, I find these allegations to be troubling and unwarranted. The point of this paper is to take steps to rebut them. I begin by making four conceptual points about retributivism. If I am correct, retributivism comes in countless flavors, and the particular variety to which I am most attracted can be applied to show why some punishments should be less severe than those presently imposed. Next I argue that many persons deserve less punishment than our legal system currently inflicts. Reflection about whether perpetrators should be afforded a complete or partial defense reveals retributivism to be less punitive than conventional wisdom would suggest. Lately it has become a commonplace to complain about the injustice of mass incarceration. I share the sentiment that this phenomenon has been an injustice. But it also has become orthodoxy to allege that the acceptance of a retributive penal philosophy has been one of the chief factors that has brought about mass incarceration in the first place. 1 The mechanisms through which retributivists are said to merit blame for this state of affairs are generally unspecified. 2 To laypersons, retributivism is nearly synonymous with a willingness to punish excessively. To academics, the details of the theory itself are sometimes alleged to be the culprit. 3 On other occasions, the claim is not that retributivism per se is responsible, but that the theory is too easily co-opted by those who seek to inflict harsh treatment. 4 On still other occasions, retributivism is said to be implicated inasmuch as it is unable to offer a persuasive critique of our unacceptable rates of incarceration -which a normative theory of punishment should have the resources to be able to do. 5 Not only have retributivists been held responsible for causing our proclivity to incarcerate too much, they also are accused of being apt to derail any progress that might be made in the future. Rachel Barkow is among a number of contemporary theorists to recommend several practical reforms that would help to make existing systems of criminal justice more rational and cost-effective. According to Barkow, a good deal of imprisonment could be withheld without jeopardizing public safety. An animating assumption of her approach is that the citizenry is receptive to such proposals. But this assumption, as she is aware, may not be correct, and she worries about the potential of retributive thinking to scuttle her reforms. She concludes by writing: 'The arguments in this book are premised on the idea that the public is primarily concerned with public safety as the goal in setting criminal justice policies. It is entirely possible that a significant segment of the public is willing to sacrifice public safety for what it believes to be retributive justice -giving people what they deserve'. 6 I suspect many laypersons are willing to make the trade-off Barkow describes. But at least she does not hold retributivists at fault for causing the crisis. As she must realize, the epidemic in rates of incarceration coincided largely with worries about public safety that were trenchant when crime rates were high. Still, it is fair to say that she fears retributivism to be a potential impediment to progress. As a self-proclaimed retributivist, I find these allegations to be troubling and unwarranted. 7 The point of this paper is to take steps to rebut them. Although I can hardly speak for all retributivists, I am sure I speak for most. We are just as dismayed by the size and scale of our criminal justice system as non-retributivists. To be sure, our favorite examples of our proclivity to over-punish tend to reflect a liberal ideology that wonders why blue-collar criminals, for example, have suffered in contrast to their white-collar counterparts. 8 Despite these differences of emphasis, I doubt that retributivists have done much to contribute to our current predicament. In fact, I believe it is more accurate to state that the punishment theories produced by academics and the punishment policies implemented by politicians bear little relation. As Michael Tonry indicates, 'Punishment theories and policies have marched in different directions in the United States for nearly 50 years'. 9 For better or worse, I detect little influence between academic inquiry and real-world trends, and the gap between the two seems to be widening rather than narrowing. Legal philosophers claim too much credit when they profess to have had a significant impact, either positive or negative, on the actual shape of criminal justice institutions. Even so, I hope it is useful to address this criticism of retributivism on its own terms. The key to my endeavor is not to reject the intelligibility of desert altogether -as some philosophers have done -but to present a variety of grounds on which to conclude that fewer persons deserve punishment than positive law inflicts. And a good many of the persons who do deserve punishment deserve less of it. Progress in reducing the size of our prison population sometimes comes from unanticipated places. The recent surge in criminal justice reform led by the 'Black Lives Matter' movement is one 7 Other philosophers have discussed retributivism and over-punishment. One line of thought I do not discuss is that those committed to punishing in accordance with desert inevitably operate under conditions of uncertainty, and the evils of mistakenly punishing too little pale in comparison to the evils of mistakenly punishing too much. prominent source. The ongoing spread of the coronavirus in prison is another. The latter has done more to accelerate the release of inmates in the United States (as well as in much of the rest of the world) than any arguments ever advanced by legal philosophers or criminologists. Whether these sources of reform facilitate public safety is debatable. Among other worries, early release has led to a predictable backlash from victims who were not even consulted and object that justice has not been done. Criminal justice reforms may even have helped to fuel the troublesome spike in violence (especially murders) throughout much of the United States in the second half of 2020 and the beginning of 2021. 10 Public safety and expressive values apart, however, this experiment should be welcomed by retributivists to the extent that (1) many and perhaps most of the released prisoners were convicted of non-violent crimes, and (2) nonviolent crimes are generally less serious than violent crimes, and thus deserving of less severe punishments pursuant to the principle of proportionality. Both of these propositions may be true, but require empirical evidence and normative support. Reformers will eagerly track whether and to what extent this unprecedented development affects any of the goods said to be produced by incarceration, most notably public safety. We can hardly rely on a crisis of public health or a reaction to police brutality to keep fueling a de-incarceration movement. Philosophers have something to contribute. Conceptual uncertainty about the nature of retributivism has helped to confuse the issue I propose to explore. In what follows, I begin by making four conceptual points about retributivism. If I am correct, retributivism comes in countless flavors, and the particular variety to which I am most attracted can be applied to show why some punishments should be less severe than those presently imposed. But conceptual analysis takes us only so far. Next I argue that many persons deserve less punishment than our legal system currently inflicts. Reflection about whether perpetrators should be afforded a complete or partial defense reveals retributivism to be less punitive than conventional wisdom would suggest. Since I have made something of a career 10 Police Commissioner Dermot F. Shea has linked the significant rise in shootings to the release of inmates from Rikers Island because of measures to limit the coronavirus and the adoption of new laws to limit the use of bail. Predictably, reactions to such allegations divide along partisan lines. See Ashley Southall: 'New York Police Face Scrutiny as Shootings Soar and Arrests Drop', New York Times (July 17, 2020), p. A1. supporting these conclusions, I will draw extensively from my previous work. Although few of them are new, I hope it is worthwhile to collect several of the most important considerations in a single place in order to gain a sense of the overall weight of the many desert-based reasons to withhold liability and lessen punishment. If my arguments are persuasive, I conclude that retributivism can actually be a part of the solution rather than a major source of our current problem. II. FOUR CONCEPTUAL POINTS I begin with four (mostly) conceptual points that do much to undermine the charge that a retributive penal philosophy has been a significant factor contributing to mass incarceration. Those philosophers enamored with analysis may find these points sufficient to establish my case. Even without substantive arguments that purport to show many sentences to be undeservedly harsh -which I offer in Parts II and III -I hope it is clear that retributivists should not be faulted for causing the unacceptably high rates of imprisonment from which we presently suffer. First and most obviously, the retributive theory itself needs to be clarified. Much of the plausibility of the supposition that retributivism fuels mass incarceration is derived from a defective characterization of its underlying nature. As I understand it, retributivism is not really the name of a particular theory of punishment any more than liberalism or conservatism is the name of a particular political ideology. Instead, it refers to a tradition or a group of theories that share loose similarities. 11 Commentators have long recognized that these similarities are hard to capture. 12 Most generally, however, this tradition affords a central and indispensable place to desert (and, perhaps even more importantly, to its absence) in its account of disagree about what desert is, how it figures in the explanation of why penal sanctions are justified, when theories that do not explicitly mention it invoke surrogates or proxies for desert, what it is about criminal behavior that makes persons deserving of punishment, and a host of additional matters. These quarrels among retributivists can be every bit as heated as those between retributivists and philosophers who presuppose an entirely different normative framework. The view that retributivism affords a central and indispensable place to desert must be distinguished from the view that retributivism relies exclusively on desert in its attempt to justify punishment. The latter view is typically attributed to Michael Moore, probably the world's most well-known and distinguished retributivist. According to Moore, 'the distinctive aspect of retributivism is that the moral desert of an offender is a sufficient reason to punish him or her'. 14 Although some retributivists follow Moore on this matter, 15 I am sure that this definition has been quoted more frequently by foes than by friends of retributivism. Even on this account, however, the connection between retributivism and mass incarceration is obscure -as Moore himself takes pains to emphasize in demonstrating his commitment to liberalism. 16 But the supposition that retributivists care only about desert in their justification of punishment fuels the temptation to believe that whatever considerations show mass punishments to be objectionable must be antithetical to the retributive tradition. The better position -which allows desert to play any number of roles even though it is central and indispensable -is obviously more conducive to pluralism about the justification of punishment. 17 On my view, which I regard as a version of retributivism, the criminal law performs several legitimate normative functions in addition to implementing a principle of retributive jus-tice that assigns value to treating offenders as they deserve. 18 Retributivists should not be reluctant to draw from any of these additional functions to lodge their protest against mass incarceration. 19 My second point is closely related to the first. Retributivism is a tradition that emphasizes the role of desert, but a gap exists between the premise that a given punishment is deserved and the conclusion that it should be imposed all-things-considered. This gap can be huge. Desert is only one of several components in a theory of sentencing, and may not even be the most important. In my judgment, a host of considerations are needed to bridge the foregoing gap, nearly all of which are consequentialist. 20 Among these several factors, crime prevention looms the largest in helping to justify the infliction of whatever punishment is deserved. If a given punitive sanction is not needed to increase prevention -or, even worse, is counterproductive in achieving that end -retributivists can be perfectly consistent in recommending that it be withheld. Retributivists can be sensitive, for example, to whether punishment costs too much money, empowers authorities to abuse their power, is too often imposed on the innocent, or exacerbates racial tensions. It was always fantastic to suppose that something as awful as tokens of punishment could be justified irrespective of their consequences. Third, retributivism itself has no implications about the mode or kind of punishment that should be inflicted. 21 Let me elaborate. I assume that retributivists attach special significance to a principle of proportionality. This principle is notoriously difficult to apply and is hard to formulate precisely. 22 As I construe it, proportionality requires (ceteris paribus) that the severity of the punishment that is deserved should be a function of the seriousness of the offense that has been committed. Once the quantum of punishment a given defendant deserves has (somehow) been specified, however, desert has no further implications for the particular form it should take. Defendants who have committed equally serious crimes may each receive a different type of punishment, as long as these modes are comparable in severity. If parity can somehow be established between the many alternative kinds of sanctions that are available, retributivists have no basis to prefer one to another on grounds of desert. Since they have no special allegiance to incarceration as the default mode of punishment, retributivists can even join the growing chorus of penal theorists who advocate prison abolitionism. 23 The very idea that desert is expressed into time behind bars is a relatively recent phenomenon. Retributivists can prefer banishment, exile, corporal punishment, home confinement, a monetary fine, or whatever else they find defensible on grounds independent of desert. Perhaps they can even go so far as to question whether any form of hard treatment or deprivation is always needed to supplement the condemnation and stigmatization that punishment necessarily includes. 24 The only non-trivial constraint that emerges from proportionality is that the extent of any sanction that is deserved must be a function of the seriousness of the crime that is committed. Once this condition is satisfied, however, retributivists can defer entirely to consequentialists in selecting among the various options. If non-retributivists make a compelling case against incarceration, retributivists are welcome to borrow their insights and join their crusade. In fact, almost no retributivist explicitly tries to justify incarceration. Instead, their efforts are aimed at justifying punishment. In both casual and scholarly thought, it is unfortunate that the latter is so often equated with the former. Once they are distinguished, however, it becomes clear that retributivists have almost nothing to say about whether a given offender merits imprisonment. 25 Presumably, 23 incarceration is a more severe type of sanction than any of the alternatives retributivists would tend to allow. But how serious must a given crime be in order to merit incarceration? The draftsmen of state and federal sentencing guidelines struggled with this issue in drawing the in/out line (sometimes called the dispositional line) on their sentencing grids. Defendants who committed crimes of greater seriousness than those along the in/out line became subject to imprisonment. But the exact placement of this line cannot be specified with any confidence. This lack of precision is due largely to the fact (as critics of retributivism never tire of pointing out) that assignments of cardinal proportionality (unlike those of ordinal proportionality) are extremely difficult. 26 Almost certainly, conventions contribute to these assignments. 27 If so, retributivists could bring about a reduction in incarceration simply by adjusting the in/out line on the grids in their sentencing guidelines. Since little in their theory specifies where this line should be drawn, nothing precludes them from drawing it elsewhere. These considerations lead me to suspect that commentators may be misguided to fixate narrowly on incarceration (massive or otherwise). This fixation would clearly be a mistake if critics only manage to replace it with an alternative that is as bad or worse. Imprisonment may be awful, but we should not tolerate a substitute that is equally detrimental. 28 Arguably, then, our target should be on the scope and scale of punishment rather than on that of incarceration. Most of the offenses for which persons are arrested and prosecuted do not result in sentences of imprisonment, but they should not be allowed to fly beneath our radar screen because they cannot land defendants in prison. 29 Penal theorists have begun to pay more attention to the injustice of the lesser punishments imposed for misdemeanors and other non-serious offenses. 30 nized as a multi-billion dollar dehumanizing debacle, it turns out that the misdemeanor behemoth does quieter damage on an even grander scale'. 31 Thus I believe that over-punishment, not mass incarceration, is just as appropriate a target for reformers. Finally and perhaps most importantly, no one can assess the extent to which a given penal philosophy should be faulted for fueling rates of incarceration without a baseline of comparison. I do not mean merely that we need to know what rate of incarceration is optimal before we can be sure that we have too much of it. Although this general point is correct, I am willing to concede that our level of incarceration is too high. 32 Instead, I mean that we need some assurance that a plausible alternative attempt to justify punishment can do better on this score than the retributive view under attack. Of course, retributivists support more severe sanctions than those who would abolish punitive institutions altogether. Short of punishment abolitionism, however, sentences must be imposed pursuant to some normative rationale, however pluralistic or complex. 33 It is one thing to criticize retributivism, and quite another to defend a viable alternative that can do better. It remains to be seen whether any theory of punishment and sentencing can be more successful in reducing the size and scope of punishment while satisfying most of our normative intuitions. 34 It is well beyond my purview to canvass the viable options, but a cursory reminder of the central challenge can be summarized in a single paragraph. Consider, for example, a simple-minded consequentialist theory of punishment that does not include desert. Since the time of Jeremy Bentham, it is telling that no philosopher comes to mind who has explicated the details of such a theory. 35 The obvious difficulty for a consequentialist is familiar to every student in introductory ethics. If the effects of punishing a given person are sufficiently good, why is it 31 relevant whether or not she has committed an offense? Is past misbehavior merely defeasible evidence of future dangerousness, or is it a necessary condition of justified punishment? In other words, consequentialist theories seemingly sever the connection between crime and punishment. Much the same problem resurfaces with rehabilitative theories. Are we really prepared to allow the indefinite confinement of persons who would benefit from rehabilitation, even if they have done nothing wrong? If I am correct, neither of these perspectives would seem to provide a workable recipe for reducing the size and scope of our penal justice system. As bad as retributivism is said to be, its competitors may be even worse. Of course, we cannot be sure until we scrutinize the details of a rival vision. At present, however, few alternatives have been placed on the table for a careful inspection. Most theorists seem content to attack retributivism without specifying what should replace it. I hope that quite a few philosophers regard the above conceptual points as sufficient to acquit retributivism of the charge that it bears a significant degree of responsibility for the epidemic of mass incarceration and over-punishment. But I do not want to rest my case on these four abstract claims alone. Substantive arguments are also needed to bolster my position. We retributivists should seek to show that a great deal of punishment is and has been undeserved. Full support for this conclusion, however, would require nothing less than a comprehensive theory of desert. A systematic effort to match punishment severity with crime seriousness, I hope, would reveal that sentences are too high across the board. 36 If this case could be made, it might be appropriate to take a 'second look' at some of the more punitive sentences inflicted on persons who are presently incarcerated. 37 Unfortunately, however, neither I nor anyone else has such a comprehensive theory -as critics of retributivism are fond of pointing out. 38 Admittedly, this fact is something of an embarrassment. As my fourth and final conceptual point suggests, however, we should not make too much of this difficulty. Just as importantly, however, I believe this problem can be circumvented largely by 36 relying on judgments of comparative desert. Respondents tend to be far more confident that one offender deserves more or less punishment than another than that they can identify a uniquely correct quantum of punishment a given offender merits. Thus I believe that we can make progress using desert-based arguments to limit the size and scope of our penal justice system even without a comprehensive theory that shows many offenders are punished with disproportionate severity. III. SUBSTANTIVE CONSIDERATIONS: DEFENSES As I have indicated, retributivists need not answer all questions about punishment by reference to desert, and can invoke any number of additional considerations to argue against detrimental policies such as mass incarceration and over-punishment. These additional considerations could well be the most important factors in curtailing our punitive policies. Still, quite a bit of headway could be made simply by ensuring that criminal liability and punishment conform to desert. One reason we have high rates of incarceration is not because so many persons deserve prison, but because we often (although not always) punish in excess of desert. 39 We can proceed on at least two fronts in the criminal law when trying to show that the size and scale of penal justice would be reduced if desert were taken more seriously. We could invoke considerations of desert to argue for a narrowing of the scope of offenses, or we could do so to argue for an expansion in the extension of defenses. 40 The first route is the more familiar, but may not be the more efficacious. Although calls to repeal existing offenses are often thought to be the more effective avenue to retard over-punishment, I have come to believe that their capacity to do so has been exaggerated. Efforts to expand the scope of defenses have a greater potential to combat mass incarceration and over-punishment. 39 Occasionally, as in the domain of sexual offenses and white-collar criminality, desert-based con- I will discuss three contexts in which desert-based arguments might bring about reductions in sentences. For at least three reasons, however, it is nearly impossible to quantify the extent of the reductions that might be anticipated through the avenues I explore. First, this determination depends partly on empirical conjectures that are difficult to estimate. How many defendants really qualify for the defenses I examine? This question is partly sensitive to how burdens of proof are assigned -a matter about which I have nothing to say here. Second, courts may already take informal account of the defenses I mention. If the case for complete or partial exculpation is really as powerful as I will suggest, one would expect knowledgeable sentencing authorities to have exercised whatever discretion they possess to reflect the significance of the factors I describe. 41 Finally, as I have indicated, my arguments are comparative. I will contend that one defendant deserves less punishment than another who is similar in all relevant respects except for the variable I isolate. But I confess to uncertainty about how to gauge the quantum of the reduction that is deserved when these comparisons are made. Still, I suspect that quite a bit of progress can be attained in reducing the scale of punishment if some or all of the following suggestions are implemented. In any event, we must try. As Richard Frase has asked, 'can we afford to renounce any major sources of mitigation, given our inflated American penalty scales and overbroad criminal laws?' 42 The first idea is perhaps the most radical, and probably has the greatest potential to achieve impressive results. Elsewhere I have argued that defendants who are ignorant of law should be partly or wholly exculpated. 43 More precisely, a defendant who is ignorant that her conduct is morally wrongful (and not merely that it is illegal) is almost always less culpable than one who is not. Since (ceteris paribus) the severity of the punishment that is deserved should be a function of the seriousness of the crime that is committed, and crime seriousness is partly a function of the culpability of the defendant, it follows that the offender who is unaware that her 41 conduct is wrongful should generally be punished less severely than another who commits the same offense but understands her conduct to be wrong. My position rests on both intuitive and theoretical grounds. To be sure, intuitions about different kinds of situation involving moral ignorance may turn out to vary, so it would be dangerous to generalize too readily from any single case. Still, the following example is instructive. Let me begin by stipulating what I regard as obvious: slavery is and always has been an unjust institution and owning slaves is wrongful. Nonetheless, Hittites who lived thirty centuries ago apparently had no moral qualms about enslaving captives caught in battle. Suppose a slave-owner in some time and place realizes perfectly well that slavery is an immoral institution, but cannot be bothered to do what she knows to be right because she is unwilling to do the hard work her slaves perform on her plantation. How should we assess her blameworthiness relative to that of the ancient Hittite who I again stipulate lacks the relevant moral knowledge? Reasonable minds can and do disagree, 44 but I hold the blameworthiness of slave-owners who know better to be significantly greater than that of slave-owners who are morally ignorant. Arguably, the latter are not blameworthy to any degree. Of course, no one can hope to decide why knowledge that his conduct is wrongful makes one person more deserving of blame and/or punishment than another who is ignorant without at least implicitly invoking some general principles -or perhaps an entire theory -of responsibility. Legal philosophers should struggle to make this commitment explicit, as producing and applying such a theory is the missing piece in nearly all normative investigations into the significance of ignorance of wrongdoing. Without such a theory, we have only our intuitions on which to rely. Unfortunately, however, few projects in contemporary moral philosophy are more unsettled and divisive than that of defending a general theory of responsibility. I hold that questions about whether a person is responsible for her conduct should be resolved by invoking the same framework that shows why a person possesses the capacity for responsibility in the first place. According to the position I accept, agents become eligible for attributions of responsibility generally, and for a given act in particular, by possessing and being able to exercise their capacity to be reason-responsive with respect to that act. Roughly, we become morally blameworthy for wrongful conduct when our capacity to respond to moral reasons is intact, but we utilize it incorrectly. When do we utilize our capacity to respond to moral reasons incorrectly? According to the conception I am inclined to favor, deliberation is deficient in the way that supports blameworthiness most clearly when agents respond incorrectly to the balance of moral reasons according to their own lights. 45 In these circumstances, agents know better than to commit the wrongful act they have performed. When these agents commit wrongs, they merit blame rather than, say, moral education. As a result of my adherence to this theory, I have come to blame far fewer persons for their wrongful conduct than I did at the beginning of my career. Thankfully, I suspect that blameworthiness is not as pervasive as many philosophers are inclined to believe. Admittedly, relatively few moral and legal philosophers share the theory of responsibility that would exempt wrongdoers from blame when they are ignorant their conduct is wrong. Most philosophers, I believe, hold a version of what might be called a quality of will theory of responsibility. According to this class of theories, moral responsibility is not located in an agent's defective practical reasoning -at least as I understand this term. Instead, responsibility for a wrongful act is grounded in the will: agents are blameworthy when their acts proceed from a will that is morally objectionable. Expressed somewhat differently, an individual is morally responsible when her action expresses negative attitudes that reveal something bad about her as a person. Countless variations and permutations of quality of will theories have been proposed, but a sustained evaluation of this family of views is well beyond my present scope. Suffice it to say that their central (but not their only) differences consist in the accounts they offer of the exact factors that make the quality of a will objectionable. The candidates mentioned most often include contempt, hostility, indifference, lack of consideration for the welfare and interest of others, and even disrespect for norms. Several philosophers who hold quality of will theories expressly advertise their ability to resist proposals to exculpate wrongdoers who act in moral ignorance. I suspect that the difficulty of demonstrating why ignorance of wrongdoing does not exculpate has boosted the popularity of these theories; philosophers frequently appeal to quality of will accounts in the course of their efforts to refute views they regard as counterintuitive. Philosophers who hold quality of will theories sometimes express bafflement (or even outrage) about criteria that take seriously the possibility that ignorance of wrongfulness might exculpate as widely as my own account would allow. Thus they would not hesitate to blame the ancient slave-owning Hittites. But I am unmoved. Quality of will theorists should agree that the agent's beliefs about the wrongfulness of her conduct should be included among those factors from which the quality of her will is inferred. In addition to recognizing a partial or complete defense for ignorance of wrongdoing, the severity of punishments could be reduced through a second route: by attaching formal recognition to a variety of mitigating circumstances. By definition, a circumstance mitigates when it lessens the harshness of a deserved punishment relative to some baseline. The systematic implementation of this proposal, however, would require a theory (or at least a set of principles) of mitigation that would include a specification of the baseline. The lack of scholarly effort devoted to this task stands in stark contrast to the overwhelming attention directed to the construction of a theory of complete defenses. 46 Some theorists are so fixated on precluding punishment that they neglect valuable opportunities to lessen its severity. At any rate, this imbalance should be rectified as part of the project of using desert to reduce the severity of punitive sanctions. I have suggested a relatively simple means to identify the mitigating circumstances that should be formally recognized by sentencing authorities. 47 These circumstances amount either to partial excuses or to partial justifications. In short, the severity of the punishment a defendant deserves should be reduced if either he or his act has an analogue in a complete excuse or justification. The clearest (but not the only) kind of analogue exists when the facts that de- scribe him or his act would amount to a complete defense were they greater in degree. For example, if a defendant uses slightly more force than would fully justify his act against a culpable aggressor, he should be granted a partial justification. Or if his mental disease or defect slightly impairs his capacity to conform his behavior to reason, he should be granted a partial excuse. Commentators have long been puzzled by the fact that legal judgments are typically bivalentdefendants are either justified or excused or they are not -even though the facts that gave rise to these judgments are almost always scalar. 48 The recognition of formal mitigating circumstances would go some distance toward responding to this puzzle. One mitigating circumstance without a straightforward analogue in a complete defense merits a more extended discussion: the severity of the punishment a particular defendant would otherwise deserve should be reduced when he has been 'already punished enough' before his official sentence has been imposed by the state. 49 What I mean, roughly, is that a person deserves less punishment when he has already suffered enough for his crime under circumstances in which his suffering is stigmatizing. As we have seen, the proportionality principle states that the extent of the sentence a defendant deserves should be a function of the seriousness of her crime. But nothing in this principle requires that punishment must be imposed by the state. To my mind, the insistence that the only real punishments are state punishments, or that the only punishments that count for purposes of applying the principle of proportionality are those inflicted by the state, has done a disservice to criminal theory in myriad ways. 50 One of the ways this claim has done a disservice is by exempting non-state punishments from being taken into account when the principle of proportionality is applied and defendants are sentenced. Reservations from legal philosophers about accepting the 'already punished enough' plea come from some of the most distinguished legal philosophers of our era. Two merit special mention. First, some influential definitions of punishment would seem to disqualify the relevance of this plea. The most widely-quoted such definition has been proposed by HLA Hart. For present purposes, the pertinent clause in Hart's definition is that a 'standard case' of punishment 'must be imposed and administered by an authority constituted by a legal system against which the offense is committed'. 51 It follows that a deprivation, however onerous, cannot qualify as a genuine punishment unless it is imposed by a legal authority acting on behalf of the state. I see little reason to accept this prong of Hart's celebrated definition. 52 Nothing in the concept precludes genuine punishments from being levied by friends, spouses, vigilantes, or total strangers. 53 Second, Antony Duff famously suggests that the primary function of institutions of penal justice is to call defendants to answer for their crimes. 54 Anything undesirable that happens to a defendant that is not a part of the state's response to his answer lies beyond this function of penal justice, and thus should not be allowed to influence the punishment that is appropriate. Even if Duff has correctly identified the most important function of criminal justice institutions, however, it hardy follows that the severity of the state response that is warranted when defendants provide a deficient answer should not be reduced when parties other than the state have inflicted punishments. I will not seek to further persuade these legal philosophers on the level of theory. Instead, I simply report my intuition that non-legal deprivations should be allowed to offset state punishments under at least two circumstances. First, consider harms that befall to perpetrators in the course of the very criminal incident itself. Suppose, for example, that a drunk driver is seriously injured in the crash of his vehicle. If consigned to a wheelchair for the remainder of his life, why should his sentence be just as severe as that of a drunk driver who escapes unscathed? Second, consider harms that third parties impose on perpetrators for their crimes. Suppose, for example, that the husband of a rape victim locates and beats the offender severely. If permanently disabled, why should the latter's sentence be just as harsh as that of a rapist who is undetected until he is arrested by the police? Although reasonable minds may disagree, I would take these harms into account in calculating the amount of the punishment, if any, that defendants deserve when the state calls them to account. Again, I resort to comparative judgments about offenders who are otherwise relevantly similar in the hope of supporting my position intuitively. It seems callous to treat these defendants as though their harms had not occurred when courts determine what sentence to inflict. Let me provide a specific illustration of the kind of situation in which at least partial exculpation should be extended. At least 833 pediatric deaths have been caused by heatstroke in locked cars throughout the United States since the mid-1990s. 55 In the scenarios I have in mind, the busy parent simply forgets the toddler is in the backseat and is subsequently horrified when he returns to find the child has died. One can debate whether these cases of forgetting involve recklessness or negligence. 56 Whatever level of culpability is involved, however, I believe the awful tragedy endured by the distraught parent should mitigate the sentence that would otherwise be deserved. The parent has already suffered a stigmatizing loss, and probably has already suffered enough -that is, to a sufficient degree to preclude criminal punishment altogether. Many (but not all) law enforcers apparently agree, because prosecutors in many (but not all) jurisdictions elect not to bring charges in these situations. One prosecutor explained his reluctance by stating 'there's nothing as a prosecutor you are ever going to be able to do to that parent that is going to come close to what that parent is going to have to live with for the rest of their [sic] life'. 57 The suppressed premise in his argument is that the suffering experienced by the parent should offset or preclude the severity of whatever punishment the state inflicts. Explicit reduction of a sentence in this circumstance would help to resolve any discrepancy among jurisdictions and formalize the case for partial or complete exculpation. 55 My third and final example of a means by which a defense might be expanded to reduce or preclude the severity of punishment has recently attracted quite a bit of discussion among criminal theorists. 58 It begins with the mundane observation that penal statutes, especially those that are instances of mala prohibita, are inevitably overinclusive. A statute is overinclusive, in the relevant sense, when it is designed to prevent some harm but proscribes tokens of conduct that do not cause or threaten to cause that harm. Presumably, tokens of conduct that fall within an overinclusive statute but do not cause harm are not wrongful. The rationale that justifies the law does not pertain to them. Since virtually all rules are overinclusive to some extent, 59 any number of examples of statutes that prohibit seemingly permissible tokens of conduct could be given. Although many commentators have tried to show that such conduct can somehow be shown to be impermissible after all, I construe the wrongfulness constraint in a theory of criminalization to place it beyond the reach of the penal sanction. 60 Punishment for such conduct would be undeserved. It would be helpful to consider specific examples of the phenomenon under discussion. What is a paradigm case of a token of permissible conduct that is proscribed by an otherwise justified overinclusive statute? Statutory rape and drunk driving have long been among the favorite illustrations. 61 Presumably, not everyone below a given age is too immature to give effective consent to sexual relations, and not everyone above a given BAC is too incapacitated to drive safely. Rather than focus on a particular instance, however, I caution that a single example may not be representative of the entire class. One of many possible grounds on which to contest the usefulness of a given example stems from uncertainty about whether the conduct in question causes whatever harm the statute is designed to prevent. To be confident about this matter, of course, one must be able to specify what that harm is. This problem is formidable; particular statutes may prevent (and be designed to prevent) several different harms, and no authoritative guidance about the statutory objective is likely to be available. 62 Overinclusive statutes need not be repealed to preserve the wrongfulness constraint. No one should reject the justifiability of these laws altogether. In order to salvage them, I would invoke a distinction familiar to constitutional lawyers but relatively undeveloped in the criminal domain: a contrast between a facial challenge to a statute, which seeks to invalidate it in its entirety, and an as-applied challenge, which seeks to invalidate its particular application. 63 A successful 'as-applied' challenge leaves the statute intact so it can continue to be applied to cases in which the defendant causes or risks the harm the statute is designed to prevent. To be clear, I am agnostic about whether the wrongfulness constraint should be accepted as a matter of constitutional interpretation. My only claim is that an overinclusive penal law should not be applied to proscribe and punish tokens of conduct that are permissible, even though the statute itself remains a part of our criminal code. Ideally, the state would succeed in avoiding the punishment of persons whose conduct is permissible while retaining the many practical advantages of drafting overinclusive legislation. Either of two devices to achieve this objective illustrates the fluid nature of the boundary between offenses and defenses. First, the state might extend an explicit defense to those who engage in conduct of the sort I have described. This defense would be codified as a general provision potentially applicable to each offense. What I have in mind is presently (if misleadingly) included as a part of the de minimis defense in the Model Penal Code, 64 which provides: 'The Court shall dismiss a prosecution if, having regard to the nature of the conduct charged to constitute an offense and the nature of the attendant circumstances, it finds that the defendant's conduct ... did not actually cause or threaten the harm or evil sought to be prevented by the law defining the offense'. 65 Second, much the same outcome could be achieved by inserting an implicit provision among the elements of penal statutes. As a result, each offense would be understood to require defendants to cause or threaten more than a de minimis amount of the harm the law is designed to prevent. The first solution creates a general defense; the second adds an element to the complete specification of each offense. These two solutions differ, primarily in their implications for how burdens of proof should be allocated. For present purposes, however, their similarities are more salient. Either device would allow the criminal law to simultaneously satisfy three desiderata: to enact and enforce overinclusive legislation, preserve the wrongfulness constraint, and reduce the amount of punishment the state presently inflicts. If I am correct, quite a few of the punishments presently imposed in the United States are almost certainly undeserved. Desert-based arguments can be used to bring about reductions in the severity of sentences either by narrowing the number and breadth of offenses, or by increasing the scope and application of defenses. Although the latter path is probably explored less frequently, it may have the greater potential to reduce our notorious propensity to be overly harsh. Thus retributivists should be encouraged to promote themselves as part of the solution rather than as a cause of the problem of mass incarceration and over-punishment. But modesty is needed. For two reasons, I do not pretend that the project of showing significant numbers of punishments to be undeserved will have a dramatic impact on the goal of reducing the size and scale of our criminal justice system. First, the amount of leniency that would ensue if my recommendations were implemented is impossible to gauge. We can hazard a rough estimate of the effects of repealing drug offenses, but quantifying the impact of expanding the several defenses I have mentioned is far more speculative. Second, philosophers of law play only a limited role in real-world reforms, as policy makers continue to pay little attention to us in an era in which issues of criminal justice are thoroughly politicized. But we should not despair. In my experience, politicians feel vindicated and more secure in their recommendations if they are confident that academics support their efforts -even if scholarly argument is not what influenced their recommendations in the first place. Thus we retributivists should not hesitate to offer whatever assistance we can to those who aspire to shape penal policy for the better. The Sentencing Commissions Functions Why Punish the Deserving? Kinds of Punishment The Metric of Punishment Severity: A Puzzle for the Principle of Proportionality Smooth and Bumpy Laws Already Punished Enough', in Husak Punishment and Retribution Duff and I have been disagreeing about this issue for nearly as long as we have known one another Playing By the Rules Theorists who allege that this conduct is wrongful after all include Stuart P. Green: 'Why It's a Crime to Tear the Tag off a Mattress: Overcriminalization and the Moral Content of Regulatory Offenses Rethinking the Wrongness Constraint on Criminalization Additional examples are discussed in James Edwards: 'Criminalization without Punishment For example, what harm(s) are various drug proscriptions designed to prevent? See Douglas Husak: Legalize This! The Case for Decriminalizing Drugs For an overview of the distinction and its complexity, see Richard Fallon: 'Fact and Fiction About Facial Challenges The De Minimis 'Defence' to Criminal Liability Special thanks to Antony Duff, Kit Wellman, Alec Walen and Leo Zaibert for helpful suggestions.