Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

A Response to Professor Kleinfeld’s Reconstructivism: The Place of Criminal Law in Ethical Life

Professor Joshua Kleinfeld’s reconstructivism is profoundly ambitious. It is meant to describe the criminal law’s core function; to answer perennial questions in the philosophy of criminal law, such as those pertaining to criminalization and punishment; to synthesize sources from disparate traditions ranging over 150 years of writing on crime and punishment; and to provide an attractive framework for evaluating existing criminal justice institutions.1×1. See Joshua Kleinfeld, Reconstructivism: The Place of Criminal Law in Ethical Life, 129 Harv. L. Rev. 1485 (2016). So Kleinfeld’s paper sets out a remarkable challenge for itself.

Does it succeed in meeting that challenge? For my part, I suspect we need to see more argument from Kleinfeld on each of these fronts before we jump on board. First, Kleinfeld purports to give us an account of what criminal law’s core function is. But his functional analysis cannot explain why the role he has identified is “the” central role for criminal justice institutions, nor why it is so important that this role take the specific form of punishment. Second, Kleinfeld vacillates between two different claims — that the criminal law is constitutive of a society’s way of life (and in that sense “embodies” it), and the quite different claim that the criminal law is a means of protecting a society’s way of life. Without a clearer account of the basic architecture of the view, it is difficult to assess Kleinfeld’s use of reconstructivism to legitimate punishment. Third, I raise some concerns about the relation between reconstructivism and its historical forebears as well as its present-day interlocutors. Finally, does reconstructivism serve to effectively insulate a society’s potentially oppressive use of the criminal law from criticism? Kleinfeld’s protestations notwithstanding, I suspect it might.2×2. In the interest of full disclosure, I have written on these issues, and defend a rather different view than Kleinfeld does. See Vincent Chiao, What Is the Criminal Law For?, 35 L. & Phil. 137 (2016) [hereinafter Chiao, What Is the Criminal Law For?], http://link.springer.com/article/10.1007/s10982-015-9247-8 [http://perma.cc/5Y2P-2ZPY]; Vincent Chiao, Two Conceptions of the Criminal Law, in The New Philosophy of Criminal Law 19 (Chad Flanders & Zachary Hoskins eds., 2015) [hereinafter Chiao, Two Conceptions].

Function: What Is Punishment For?

Kleinfeld seeks to answer the question: what is crime? In doing so, he also seeks to answer another question: what is the criminal law? These are good questions to be asking, and Kleinfeld is, in my view, absolutely right that the philosophy of criminal law has not paid sufficient attention to these questions.3×3. But see, e.g., Grant Lamond, What Is a Crime?, 27 Oxford J. Legal Stud. 609 (2007) (discussing potential answers to these questions). But Kleinfeld’s answer to these questions is incomplete in at least two important dimensions. First, he does not establish that expressive nullification of crime (or, as he sometimes puts it, restitching the social fabric),4×4. Kleinfeld, supra note 1, at 1537–39. even if a function of the criminal law, is its central function. Second, even if it were, he nevertheless owes us a further argument that no other social institution could do this job better than the criminal law.

With regard to the first point, Kleinfeld acknowledges that the criminal law plays roles beyond sustaining a society’s way of life.5×5. Id. It deters crimes, vindicates rights, and rehabilitates wrongdoers, at least sometimes. But in light of this plurality of functions, what justifies the assertion that restitching the social fabric is the central role for the criminal law? Surprisingly, Kleinfeld does not consider this question. But without some such account one could agree with everything Kleinfeld says about the criminal law’s role in restitching the social fabric while nonetheless rejecting his conclusions. After all, even if the criminal law were uniquely situated to restitch the social fabric, one might believe that its other functions — functions that Kleinfeld concedes it serves — are equally or more important. So one might think that when it comes to evaluating the performance of criminal justice institutions we ought to pay as much, if not more, attention to how these institutions do at deterring crime, reforming criminals, and fulfilling other purposes than to how they do at restitching the social fabric. In other words, for Kleinfeld to show that reconstructivism provides a normatively attractive framework for evaluating the criminal law, he needs to do more than show that the criminal law has a large and dramatic role to play in sustaining a society’s way of life by way of expressive nullification of crime. He also has to show that this role is so important that it overshadows the other functions the criminal law serves.

Moreover, even if the criminal law were a necessary means for sustaining a society’s way of life, it still would not follow that this end is what the criminal law is for, much less that it is the standard by which the criminal law is to be evaluated. You might not be able to watch cat videos on YouTube without eyes, but that does not show that what your eyes are for — much less the standard by which they are to be evaluated — is enabling you to watch cat videos on YouTube. Nor does it follow that this is the standard by which eyes should be evaluated. Similarly, even if nullifying wrongdoing requires punishment, it would not follow that nullifying wrongdoing is necessarily punishment’s most important or fundamental purpose.

On the other hand, one could start the opposite way around: not by taking our institutions as given and asking what role(s) they serve, but instead by specifying a functional role — sustaining a society’s way of life — and asking: what kinds of institutions or policies would contribute to realizing this function in practice? One might point out in Kleinfeld’s defense that it is quite plausible that the criminal justice system has a hand in realizing this function in practice. Hence, if we agree that this role is important, indeed essential, then doesn’t it follow that the criminal law is also essential? No, it does not. The following argument is a non sequitur:

  • (1) Sustaining a society’s way of life is an essential social function;
  • (2) the criminal law contributes to sustaining a society’s way of life;
  • (3) hence, the criminal law is an essential social institution.

Even if we grant (1) and (2), (3) only follows if a society’s way of life could not be preserved without the criminal law. But it is not clear why we should believe this to be true. After all, the criminal law is hardly the only institution that serves to preserve a society’s way of life. Families, schools, churches, mass media, and other institutions all play a vital role in reproducing the values, norms, and attitudes that prevail in a given society at a given time. What (2) says is that the criminal law also contributes to this end, perhaps even in a distinctive manner; but (2) does not say that the end could not be achieved without criminal law. Again, it might be true that punishment provides a distinctive means for preserving a society’s way of life. But to establish that this is true is not yet to establish that the criminal law is necessary to that end; it is not yet to show that a society’s way of life could not be preserved but for a significant reliance on expressive nullification of crime through punishment.

In short, reconstructivism’s functional account of crime and punishment relies on at least two crucial and undefended premises: that the criminal law is essential to sustaining a society’s way of life, and that sustaining a society’s way of life — rather than promoting some other value — is the central function of the criminal law.6×6. I join Kleinfeld in defending a functionalist approach to the criminal law, but on rather different terms. See Chiao, What Is the Criminal Law For?, supra note 2; Chiao, Two Conceptions, supra note 2.

Content: Is the Criminal Law a Means to an End or an End in Itself?

Generally, a normative theory is characterized as teleological insofar as it specifies an end (a value) and evaluates various proposed means as more or less effective at achieving or promoting that end. For any given means, one can always ask whether it is a good one and whether there might not be better ones. In contrast, nonteleological theories typically insist that the object of the theory — the criminal law, for instance — is not a means of promoting any particular value, but is constitutive of respect for that value. From a constitutivist point of view, there is little sense to the question of whether X is the best way of promoting the value in question, since the claim is that doing X is required to respect that value, not that it is a means of promoting it.7×7. For a recent discussion of the difference between teleological and nonteleological forms of retributivism, see Mitchell N. Berman, Punishment and Justification, 118 Ethics 258, 267–87 (2008).

Kleinfeld’s official view is that reconstructivism is teleological, in that it shows how the criminal law furthers human welfare by virtue of its role in helping a society sustain its way of life.8×8. Kleinfeld, supra note 1, at 1531–34. As a teleological view concerned with human welfare, it should make sense to ask whether there might be, now or in the future, some better way of sustaining a society’s way of life than punishment. This question is an important one, given the obvious downsides of a criminal justice system. Yet Kleinfeld does not consider this question, and it is also not clear whether he would think it a sensible one for a reconstructivist to ask. If, as Kleinfeld insists, crime just is a rending of the social fabric, or the “de-actualization” of a norm,9×9. Id. at 1498–500. and if punishment just is the “restitching” of that fabric or “re-actualization” of the norm, then it makes no sense to ask whether crime should be met with punishment. Punishment just is the kind of response that crime calls for, quite independent of whether it does anything to further anyone’s interests. It is “simply right.”10×10. Id. at 1526. This is to turn reconstructivism into a constitutivist theory. This kind of view is systematically insensitive to the costs of a criminal justice system, and as I have argued elsewhere, there is good reason to be less than sanguine about the ability of this type of theory to provide plausible guidance about the design of actual criminal justice institutions.11×11. Vincent Chiao, Mass Incarceration and the Theory of Punishment, Crim. L. & Phil. 4–5 (July 16, 2015), http://link.springer.com/article/10.1007%2Fs11572-015-9378-x [http://perma.cc/S9ZL-ZH5R].

This may seem like an esoteric point, but it is in fact crucial to evaluating whether reconstructivism provides a plausible theory of punishment. The criminal law may be a unique way of promoting social solidarity, namely expressive nullification of wrongdoing. However, as a teleological theory, it is not clear that reconstructivism can rule out the possibility that there are better means of promoting social solidarity that do not work by means of expressive nullification, or even that the upsides of expressive nullification necessarily outweigh its downsides. If they do not, then by reconstructivism’s own lights there would be no good reason to punish. In other words, Kleinfeld must show that (a) there is, and can be, no less destructive, humiliating, and rights-violating way of preserving a society’s way of life than punishment, and (b) our reason for valuing the preservation of a society’s way of life is so powerful that it outweighs the substantial costs of a criminal justice system. Since Kleinfeld does not defend either of these propositions, much less both of them, taken as a teleological theory — that is, taken on its own terms — reconstructivism does not provide us with sufficient reason to ever criminally punish anyone for anything.

In response to these concerns, Kleinfeld insists that criminal punishment is a human universal, and that abolishing punishment altogether is “quixotic, utopian, unworkable.”12×12. Kleinfeld, supra note 1, at 1520. Perhaps. But the point is not about abolishing punishment altogether; it is about reducing our reliance on it and relying more heavily on other means of promoting social welfare. Similarly, Kleinfeld suggests that crime is inevitable, and that he can therefore focus exclusively on reconstructing the social fabric in its wake.13×13. Id. Kleinfeld is right: crime is inevitable. But in the context of the modern administrative state, every dollar spent on punishing crime after the fact is a dollar not spent on possibly more effective and less punitive means of minimizing its overall incidence. Presumably, if repairing a norm after it has been violated conduces to social solidarity, then so too does ensuring that the norm isn’t violated to begin with. From the point of view of sustaining social solidarity, nonpunitive social programs that have the effect of reducing crime ex ante are a substitute for punishment of crime ex post.14×14. See Tapio Lappi-Seppälä, Trust, Welfare, and Political Culture: Explaining Differences in National Penal Policies, 37 Crime & Just. 313 (2008) (“There is an inverse relation between the commitment to welfare (the generosity of welfare provisions) and the scale of imprisonment.” Id. at 356.); Lance Lochner & Enrico Moretti, The Effect of Education on Crime: Evidence from Prison Inmates, Arrests, and Self-Reports, 94 Am. Econ. Rev. 155 (2004) (finding that “schooling significantly reduces the probability of incarceration and arrest,” id. at 155). Yet reconstructivism provides no guidance on how to weigh the costs and benefits of punitive and nonpunitive ways of dealing with crime. Perhaps this could be forgiven if it were a constitutivist theory. But as a teleological theory, this is a serious shortcoming.15×15. It is notable that Kleinfeld here falls back on an appeal to what “our culture” “encod[es]” as the appropriate response to crime. Kleinfeld, supra note 1, at 1520. If “our” refers here to the contemporary United States, apparently “our” culturally fixed response to crime is mass incarceration. This elides both the anomalous character of mass incarceration in American history and undermines Kleinfeld’s own claim that reconstructivism is able to critically evaluate conventional morality, a point I discuss at greater length below.

Kleinfeld might fall back on his claim that the “primary purpose and . . . competence” of the criminal law is to “restitch a torn social fabric”;16×16. Id. at 1538. perhaps other social institutions may contribute to sustaining social solidarity in other ways, but expressively vindicating the status of the victim is the unique job of the criminal law. As I noted above, I am dubious of this claim. But let us grant it for the sake of the argument. From a teleological point of view, restitching the social fabric is of value insofar as it contributes to human welfare, namely by helping a society preserve its way of life. But, as a teleological theory, reconstructivism cannot claim to have justified punishment unless it can show not just that restitching the social fabric is a means to sustaining social solidarity, but that there are no better means of achieving that very same end. To insist that he does not need to provide any such argument because of the intrinsic connection between crime and punishment is to succumb to the “magical” thinking characteristic of deontological retributivism that Kleinfeld elsewhere decries.17×17. See id. at 1527.

For my part, I think Kleinfeld does well to embrace a teleological view of the criminal law: the criminal law is not an end unto itself, but is rather a means for promoting other ends we have reason to care about. But taking this thought seriously places real pressure on reconstructivism as a normative theory. The thought that not every problem of social deviance is a problem for the criminal law is not a fanciful one. Crime is indeed inevitable. But how we respond to it — our preferred mix of prevention and punishment — is not. We have reason to embrace nonpunitive policies insofar as they are successful in sustaining pro-social attitudes while avoiding punishment’s serious human costs. But this thought, reaching as it does beyond the four corners of reconstructivism, calls into question reconstructivism’s ambition to explain when and why we should punish people. To adequately explain how public institutions should choose among punitive and nonpunitive responses to crime requires a fully political justification: an accounting, acceptable to people with widely diverging comprehensive doctrines, of how and why the criminal justice system allocates the benefits and burdens of social cooperation.18×18. Vincent Chiao, Criminal Law in the Age of the Administrative State (forthcoming 2018).

Context: Hegel and Durkheim, Duff and Brudner

Reconstructivism asserts a necessary conceptual relation between crime and punishment. Here, one might ask how Kleinfeld purports to know what “the” nature of crime or punishment is, or even that they have a nature. Kleinfeld’s answer takes the form of an appeal to the theories of Hegel, Durkheim, and Hampton. There are, however, significant questions about Kleinfeld’s interpretation of these figures’ views, at least in the case of Hegel and Durkheim. For instance, with regard to Hegel, Kleinfeld simultaneously attributes to Hegel the proto-welfarist view that punishment “serves human welfare”19×19. Kleinfeld, supra note 1, at 1526–27. as well as the non-welfarist view that punishment is an “imperative of justice,”20×20. Id. and that since “wrongdoing cannot be allowed to stand, . . . punishing wrongs is not prima facie evil but simply right.”21×21. Id. at 1526 (emphasis added). For a more developed interpretation of Hegel on this point, and an explicit rejection of the idea that Hegel is a welfarist about punishment, see Alan Brudner, Punishment and Freedom 45–58 (2009).

With regard to Durkheim, Kleinfeld relies heavily on The Division of Labor in Society to support his claim that Durkheim viewed punishment as a privileged form of reproducing a society’s “collective consciousness.”22×22. Kleinfeld, supra note 1, at 1493. In The Division of Labor in Society, Durkheim famously claimed that punishment was a privileged means of holding a society together if that society was characterized by “mechanical” rather than “organic” solidarity. According to Durkheim, societies exemplified by mechanical solidarity are held together by peoples’ similarities to each other rather than, as in the case of societies exemplified by organic solidarity, the kinds of complex cooperation made possible by peoples’ differences from each other.23×23. See Emile Durkheim, The Division of Labor in Society 31–87 (W.D. Halls trans., Macmillan Press 1984) (1893). In other words, Durkheim’s thesis was not that punishment was an all-purpose way of fostering solidarity, but rather that it served that purpose in the context of undifferentiated societies, a role that is gradually overtaken by more complex forms of social cooperation in heterogeneous societies such as the contemporary United States. Hence, it is far from clear that Durkheim’s views in The Division of Labor support reconstructivism, at least when applied to complex societies marked by a plurality of interdependent social roles and identities.24×24. See generally id. at 31–87, 101–25. Durkheim claims that in modern societies, “the ties binding us to society, which spring from a commonality of beliefs and sentiments, are much fewer than those that result from the division of labour.” Id. at 101. (Indeed, Chapter 5 is entitled “The Increasing Preponderance of Organic Solidarity and its Consequences.”) Durkheim also suggests that the nature of the solidarity that results from enforcing in-group norms against outsiders “bind[s] men together less strongly than does organic solidarity.” Id. at 105 (emphasis added). This suggests that insofar as the aim of reconstructivism is to bind people together, Durkheim’s proposal would be to enable people to cooperate as full members of a complex and interdependent civil society rather than by punishing them and excluding them from the polity.

Turning to more contemporary authors, reconstructivism evinces some notable parallels with the views of Professors Alan Brudner and Antony Duff, veritable giants in the field. The expressivist aspects of Kleinfeld’s theory, including its emphasis on the victim as well as the wrongdoer, are reminiscent of Duff’s communicative theory of punishment.25×25. See R.A. Duff, Answering for Crime (2007). The punishment-annuls-crime claim is reminiscent of Brudner’s (nonteleological) interpretation of Hegel in Punishment and Freedom.26×26. See Brudner, supra note 21, at 50 (attributing the claim to Hegel); see also Arthur Ripstein, Force and Freedom 300–24 (2009) (endorsing the view, and attributing it to Kant). To be sure, Kleinfeld’s reconstructivism departs from Duff and Brudner in that Kleinfeld describes his theory as teleological in orientation, whereas Duff and Brudner are quite clear that their views are, in the terms used here, constitutivist rather than teleological.27×27. See R.A. Duff, Retrieving Retributivism, in Retributivism: Essays on Theory and Policy 3 (Mark D. White ed., 2011); Brudner, supra note 21, at 37–41, 46; Ripstein, supra note 26, at 300–24, 301. However, as we have seen, it is not clear just how teleological Kleinfeld’s view really is, particularly since much of his argument seems to turn on the thought that expressive punishment is a uniquely appropriate moral response to crime rather than on the thought that punishment turns out to be, by comparison to available alternatives, the best way of ensuring that societies don’t fall apart in the face of crime.

Critique: Does Reconstructivism Amount to Conventional Morality Writ Large?

Kleinfeld insists that reconstructivism is consistent with a critical attitude toward a society’s norms and values, even those reflected in its criminal law. He suggests that “Thayerian restraint” would ensure that minority ways of life are not oppressed by majority preferences.28×28. Kleinfeld, supra note 1, at 1560–61. However, insofar as this is plausible, it is only because he limits the reach of his theory to “basically decent” societies29×29. Id. at 1556. — presumably, that is, societies in which the majority does not have an interest in oppressing dissentients. This restriction comes close to begging the question. The question is what reconstructivism would allow us to say in response to people who want to use the criminal law to preserve an oppressive way of life. To say that it only applies to people who do not want to use it for that purpose does not show that reconstructivism allows us to be critical about a society’s prevailing norms and values. Indeed, it amounts to a concession that it does not.

Hence it remains, I think, very much an open question just how critical an attitude reconstructivism allows us to take toward conventional morality. Kleinfeld’s remarks on Judge Patrick Devlin in this context are telling. Kleinfeld insists that Judge Devlin was a “problematic figure”30×30. Id. at 1488 n.2. in the sittlichkeitian tradition, but it is hard to see why. Consider Judge Devlin’s infamous claim that if society regards homosexuality as “a vice so abominable that its mere presence is an offence,” then homosexuals should be punished — not because homosexuality is wrong but simply because a society is held together by its shared value, including values about sex.31×31. See Patrick Devlin, The Enforcement of Morals 16–17 (1965). Suppose that Judge Devlin was right about his society’s values, at least at that point in time: that the man on the Clapham omnibus would have found the idea of two men having sex revolting, would have rejoiced in seeing gay men publicly humiliated, and would have felt that his own status was thereby made more secure. Isn’t punishment here doing exactly what a reconstructive theory says it is supposed to do?

Kleinfeld tries to establish the critical bona fides of reconstructivism by claiming that “fellow-feeling has had a central place in the democratic tradition,” citing the “modern Europe[an]” conception of solidarity.32×32. Kleinfeld, supra note 1, at 1563. This is an ironic invocation, given the current resistance of European publics to accepting the masses of Syrian refugees clamoring at their borders, and, more generally, the rise of xenophobic political parties premised on (as they see it) preserving the folkways of Europe’s member states from becoming diluted and co-opted by foreigners. One might think the problem here is precisely an excess of solidarity. The flipside of solidarity, after all, is the exclusion of outsiders, whether at border crossings, or from schools, labor markets, and “good” neighborhoods. (Prisons, of course, are only modern society’s most overt form of social exclusion.)

In the end, Kleinfeld’s insistence that sustaining a shared way of life justifies a society in punishing people who act in ways inconsistent with it is fundamentally illiberal, in that it authorizes a society to enforce the majority’s moral opinions against dissentients simply because they are the majority’s opinions.33×33. Requiring a supermajority, as Kleinfeld at one point suggests, id. at 1553–54, does not fundamentally change this point. Notwithstanding Kleinfeld’s invocation of Rawls, this is in fact a profoundly anti-Rawlsian thought.

Conclusion: Teleological, Revisionary, Critical

In my own work, I have defended an approach to the criminal law — an approach I have labeled the public law conception — that is resolutely teleological, revisionary, and critical. The public law conception is teleological, in that it treats the criminal law as a means to an end, namely stabilizing patterns of cooperation with public institutions.34×34. Chiao, What Is the Criminal Law For?, supra note 2; Chiao, Two Conceptions, supra note 2. It is revisionary, in that it does not seek to lard historically contingent labels for certain institutions (“crime” versus “tort,” for instance) with philosophical content, but rather asks: is ex post punishment the optimal means of stabilizing cooperation with public institutions?35×35. A point I develop in Punishment and Permissibility in the Criminal Law, 32 L. & Phil. 729 (2013). And it is critical, because in highlighting the role of the criminal law in supporting public institutions, it insists that justifying the criminal law cannot be prized apart from justifying the institutions it supports.36×36. See Chiao, supra note 18. The criminal law entrenches an allocation of social advantage across a population, and hence to justify the criminal law is to justify that allocation of social advantage. The public law conception thus stands in sharp contrast to Kleinfeld’s account of the criminal law as a society’s preferred means for sustaining “fellow-feeling” among its citizens by punishing those who violate a community’s shared values. On the public law conception that I defend, we cannot hope to justify the criminal law as a self-standing institution. Rather, because the criminal law is an integral part of society’s basic structure, it calls for a fully political justification: one that is acceptable to free and equal citizens living under conditions of reasonable pluralism.37×37. I develop just such an account in Chiao, supra note 18. See also Chiao, Two Conceptions, supra note 2.

Kleinfeld would no doubt regard this as so much “Enlightenment-humanist” revulsion at suffering, a well-intentioned “fantasy of [what] a different and gentler society” than ours might be like.38×38. Kleinfeld, supra note 1, at 1564. To which, in the face of the ultra-harsh realities of contemporary American criminal justice, I can only plead: guilty as charged.


* Faculty of Law, University of Toronto. My thanks to the editors of the Harvard Law Review Forum for inviting me to respond to Professor Joshua Kleinfeld’s provocative paper, as well as to Chad Flanders and Stephen Galoob for detailed comments.