Prevention and the Limits of the Criminal Law

Editors: Andrew Ashworth, Lucia Zedner, and Patrick Tomlin
Publisher: New York: Oxford University Press, 2013. 308p.
Reviewer: Alec Walen | September 2013

This book is a collection of papers by 14 prominent scholars who were invited to present their views on preventive justice and the criminal law at two seminars, held at Oxford University in the years leading up to the book’s publication. The editors describe the main objective of the project that gave rise to this book as “develop[ing] an account of the principles and values that should guide and limit the state’s use of coercive preventive techniques.” (1) The book does not itself present any such account. Rather, the articles in it contain information, concepts and arguments useful to anyone looking to develop such an account.

As is always the case with books that collect articles from different people, the material is somewhat uneven in quality. In addition, when the authors come from such a wide range of academic fields—in this case, “criminal law, public law, legal theory, criminology, philosophy, and political theory” (1)—it is to be expected that readers will find certain articles to be more interesting than others in ways that reflect the reader’s background and interests.

To guide the reader who might be interested in reading or dipping into this book, I start by providing a broad-brush overview of the contents. I then engage briefly with each of the essays in the first part of the book. Finally, I indulge my own particular interest by considering in more depth an argument that I take to be at the normative heart of the topic: whether respect for autonomy is in tension with at least the more aggressive uses of preventive justice, such as preventive detention.

1.            An overview of the book

The editors distinguish four main groupings:

1. Chapters 1-5 (by Frederick Schauer, Petter Asp, Markus Dubber, Klaus Günther, and David Dyzenhaus) that “examine the nature and historical antecedents of preventive justice and its relationship with criminal law and punishment” (2);
2. Chapters 6-8 (by Antony Duff, Victor Tadros, and James Nickel) that “focus upon particular exemplars of preventive endeavor” (4);
3. Chapters 9 & 10 (by Doug Husak and Carol Steiker) that “address in different ways questions of how preventive measures should be conceived, justified, and restrained” (5); and
4. Chapters 11-14 (by Peter Ramsay, Matt Matravers, Bernard Harcourt, and Pat O’Malley) that address “very different aspects of the politics of preventive endeavor” (6).

For my own interests, I find it more helpful to regroup the five middle chapters into those (by Duff, Tadros, and Husak—chapters 6, 7, and 9) that present normative philosophical arguments, and those (by Nickel and Steiker—chapters 8 and 10) that provide lawyerly and pragmatic assessments of, on the one hand, a contemporary prevention practice (restraining orders), and on the other hand, a tool for limiting preventive practices (proportionality as a limit on preventive detention). I also think it worth noting that the final four chapters range quite widely in their concerns, from accounts of democracy, to agency, to empirical claims of effectiveness, to the monetization of justice.

2.            Comments on the first five essays (plus a passing reference to Husak’s)
Chapter 1: Schauer starts the book by defending the provocative thesis that the criminal law is not fundamentally different from preventive justice. He defends this in two ways. First, he notes that the criminal law is largely driven by the preventive concern with incapacitating those who are found guilty so that they will not commit future crimes. Second, he argues that the differences between preventive justice and criminal justice come down to the procedures used and the standard of proof required—differences of degree only. If we could be just as “certain that someone would commit a crime of a certain type in the future [as we would need to be to convict them for a crime already committed], it is not entirely clear that the preventive approach [preventive detention, rather than punitive detention] is deficient, and it may, depending on the probabilities, at times be better.” (21)
 
Schauer thinks that his conclusion is consistent with the position in Husak’s chapter, which argues that the criminal law could be extended to allow preventive detention of people who possess “traits that predict future dangerousness.” (182). But Husak emphasizes that the relevant traits must be ones that are “under the control of persons the state preventively detains.” (190) That means that Husak’s traits do not constitute mere “status offenses.” It is not clear that Schauer has any reason to embrace that feature of Husak’s proposal.

This gets to the heart of what is radical about Schauer’s proposal. He disregards the idea that criminal punishment is deserved on the basis of a choice, and that preventive detention jumps the gun on an agent actually choosing. It treats a future choice as something that can be probabilistically assessed just as the odds that someone has done something criminal can be probabilistically assessed. In other words, without recognizing the point, Schauer embraces one side of the debate to which I return in section 3 of this review, and this is why, in my view, his thesis is quite radical.

Chapter 2: Asp offers an important warning about what he calls the growth of “preventionism” in the criminal law. Traditional criminal law had to engage certain key normative questions: Why should certain acts be criminal? How should guilt be established? What is required for someone to have a criminal level of culpability? The shift to preventionism, he argues, undermines those normative questions, and substitutes “empirical ones (what do we need to do in order to …?)” (31)

The second part of his chapter, however, engages in what seems to me a confused discussion of the relevance of intentions for establishing inchoate crimes. Asp uses a concept of intention that is consistent with the person not being “prepared to accept the object of the intention as a consequence of his or her actions.” (40) This confuses entertaining an idea with forming an intention. The fact that one can change one’s mind is perfectly consistent with the fact that, in forming an intention, one has committed oneself to achieving, even if only conditionally, certain goals.

Chapter 3: Dubber describes a distinction between two modes of governance: the law mode and the police mode. His point is that the law mode is based on an idea of legitimacy, which in turn, at least in modern liberal societies, is based on respect for the autonomy of individuals. By contrast, the police mode is based on maintaining peace and order, however that can most efficiently be done. His tracing of these ideas through history was novel, at least to me. Unfortunately, his attempt to offer a critique of the different modes presupposed that the moral foundations for the law mode are merely optional, and therefore fell flat as a normative assessment.

Chapter 4: Günther explores the idea that the international public law norm, the responsibility to protect (RTP) can be taken out of that context and used to understand the current push in democratic societies to develop mechanisms of preventive justice. He worries that what was once an obligation to protect minorities “against a government-supported majority” (88) has been perverted into a majoritarian “rhetorical device” for going after “a minority that is prone to abuse its rights at other people’s expense.” (89) This, he worries, is inconsistent with two core principles of democratic legislation: generality and reciprocity. In other words, by distinguishing good from bad citizens, it fosters a kind of moral and political blindness in which the majority fails to put itself in the shoes of the other. This he says, in his concluding sentence, could prove to be “nothing other than a tacit nullification of the social contract.” (90).

The problem with Günther’s thesis is that he fails to distinguish different kinds of minorities. Surely he is right if we understand the minority in terms of ethnic or religious communities, such as Muslims. If, however, the minority is defined as those who do in fact threaten the rights of others, then it is not so clear that he has given us reason to think that their being preventively restricted is inconsistent with generality, reciprocity, and the social contract. If there is a normative problem with preventive detention, it lies elsewhere, in the failure to respect their autonomy (or so I argue below).

Chapter 5: Dyzenhaus returns to Lon Fuller’s idea that there is an “inner morality of law” (96) that can be captured in eight process principles that are central to the rule of law. He offers a valiant attempt to derive substantive rights from process ideals, in particular the requirement that restrictions on liberty be appropriately justified. This strategy is particularly attractive to lawyers, who are drawn to process and who are, by dint of their training and practice, wary of substantive norms from outside the law. I’m afraid, however, that the project is (or seems to me to be) a hopeless act of self-deception, in which all the work is hidden in the notion of appropriate justification. But then I’m a philosopher first and a lawyer second, and my instincts are to attack substantive normative questions head on.

3.            Discussion of the central normative dispute
I turn, then, to what I think is the core normative question: whether respect for autonomy requires us to use preventive detention much less readily than criminal punishment. Duff takes the affirmative view, in the context of discussing pre-trial detention; Tadros gives us the apostate’s view (he used to hold Duff’s view), in the context of defending the use of control orders (which come close to home arrest, and can be used even if no trial is pending) against the charge that they do not respect the autonomy of those subject to them.

Duff frames his appeal to the respect due autonomous agents in terms of the presumption of innocence (PoI). He distinguishes two senses of the PoI. In the narrow sense, it implies only that a criminal defendant may not be punished unless the state overcomes, with evidence that meets a suitably high standard of proof, the presumption that she is innocent. In a broader sense, however, it requires a substantive sort of trust that can only be revoked after criminality has suitably been proved. Duff appeals to the broader sense to limit pre-trial detention, writing:
[O]ne aspect of a liberal democracy that treats its members as responsible agents is that those members must be willing to accept some kinds of risk, in particular the risk that others will offend against them. The fact that the person we suspect of future criminality has been charged with (but not convicted of) an offense cannot make it reasonable to detain her as if she is already guilty. (p. 130)

The key moral principle here, the PoI, is consistent with requiring those who are reasonably suspected of such a crime to show up for trial, and perhaps even to put up a bond that they will do so. But unless and until they are convicted or caught trying to interfere with their trial or to abscond, they should not lose the presumption that they are innocent.

Despite my being fundamentally sympathetic with Duff, his piece left me with a number of questions. What, for example, should we do with people who are going to trial not to contest whether they are guilty, but to contest the specific crime for which they are guilty? What if there is some question about whether they were trying to interfere with their trial or to abscond? Must there not be a trial about that? Would that not create a problematic regress of proof? What about “the serial offender who is very likely to offend again if freed whilst awaiting trial”? (132) Is there really a dilemma about whether he may be detained pre-trial, or could he have forfeited his right to the PoI in virtue of his earlier offenses?
 
The deepest question, however, the apostate’s question, is whether it is true that “a liberal democracy [must] treat[] its members as responsible agents” deserving of the PoI until proven guilty beyond a reasonable doubt. It is because he responds to this question that I find Tadros’s essay is so interesting.

Tadros usefully distinguishes between “presuming” that a morally responsible agent will act wrongly and “predicting” that she will. Duff’s position is that even if we predict that she will act wrongly, we may not act on the presumption that she will. Tadros notes that we do not mark that distinction with children, and then asks for “an argument why we should act on our predictive judgments with respect to children but not with respect to adults.” (145)

I should disclose at this point that I am Tadros’s first target once he raises that question. He takes me to rest my argument on the claim that if we do not presume that people will act rightly, we treat them “as though they do not have the free will to choose rightly.” (146) But Tadros misunderstands my point (perhaps because I was not clear enough). I do not wish to rest my position on a metaphysical claim that to presume someone will act wrongly is to imply that “he or she is incapable of acting morally.” (146, emphasis added) My point is a moral one, that the presumption is disrespectful. We owe it to other agents who can be held responsible for their choices to presume that they will act morally unless and until they have been convicted of having done otherwise (and even then our responses must be limited by the principle of proportionality). We can take steps to interfere with their liberty, and even to inflict physical harm, based on a prediction that they will otherwise do harm. But we may do so only when the harm they threaten is imminent and no alternative exists. If the alternative of policing a predictably dangerous person exists, and it is likely that he can be caught and prosecuted for an inchoate crime before he actually causes any harm, then he has to be treated, at least when it comes to detention decisions, with the presumption that he will not choose wrongly—the PoI.

But Tadros does more than turn the tables and demand a justification. He has an argument against the PoI position. He offers a very interesting case called “Hostel.”
It is a cold winter’s night. X knocks on the door of a hostel for vulnerable people seeking shelter. Most of the people in the hostel are black. X is well known to have both racist and violent tendencies as well as a lack of self-restraint. Jane is the person who is responsible for deciding who is let into the hostel. For good reason, she believes that X may violently attack one of the vulnerable people staying in the hostel. It will be very difficult to prevent him succeeding if he does mount such an attack. If he is not admitted to the hostel, there is a good chance that X will suffer a serious illness as a result of exposure. (146)

To show that responsibility is not a particularly salient moral quality, Tadros then introduces two versions of X: Xres, who is a responsible agent, and Xnonres, who is psychotic. His point is that we would be equally if not more likely to support Jane’s decision to exclude X if he is responsible than if he is not. Even though such an exclusion would likely cause X serious harm, it seems reasonable given the prediction of what he would otherwise do.

Importantly, the fact that this exclusion fails to accord him the benefit of the PoI seems beside the point. Moreover, Tadros offers an account of why such an exclusion, based on mere prediction, is reasonable: “Xres could have [avoided posing a threat] simply by making better judgments and developing more self-restraint.” (147) In other words, he has brought the prediction and the presumption that he is unacceptably likely to harm others on himself.

This is a fascinating example, and I end with a few thoughts about why I find it unconvincing. First, we assume Jane has a fiduciary duty only to those in the hostel, and not an equal one to X. This makes it very easy to justify excluding X, whether he is a responsible agent or not. That is not the case with most instances of preventive detention, when the person detained has just as strong a claim for the state’s concern as the ones who he predictably threatens.

Second, Tadros makes the example in a way too easy by making the basis for the prediction that Xres is dangerous his prior wrongful behavior. Suppose we make it clear that the prediction is based solely on his well-known racist and violent views, adding that we have no reason to believe that he has ever acted on those views himself. Would it then be so clear that he could be excluded at least as easily as Xnonres? I don’t think so.

Third, taking the other side of that coin, if he is thought dangerous because of his prior racist violence, could it be that Tadros is right only because X has forfeited his right to the PoI? Would that allow someone who has not so forfeited his rights to be treated in the same way? Indeed, if that is what the example turns on, isn’t it fully consistent with insisting on the PoI for those who have not forfeited their right to it?

In order to paint a fair portrait of Tadros’s position, I should point out that he considers another argument, which he thinks might be more successful in raising a normative obstacle to the use of control orders. He thinks the main argument for using them may fail to respect the “separateness of persons.” (153-154) This is an interesting argument in its own right. But I think the normative core remains whether preventive detention, not premised on rights forfeiture, is inconsistent with respect for the dignity of autonomous and responsible persons.

Alec Walen, Rutgers School of Law-Camden; Department of Philosophy, Department of Criminal Justice, New Brunswick