Rethinking Punishment in the Era of Mass Incarceration 1138047791, 9781138047792

Citation preview

Rethinking Punishment in the Era of Mass Incarceration

One of the most important problems faced by the United States is addressing its broken criminal justice system. This collection of essays offers a thorough examination of incarceration as a form of punishment. In addition to focusing on the philosophical aspects related to punishment, the volume’s diverse group of contributors provides additional background in criminology, economics, law, and sociology to help contextualize the philosophical issues. The first group of essays addresses whether or not our current institutions connected with punishment and incarceration are justified in a liberal society. The next set of chapters explores the negative effects of incarceration as a form of punishment, including its impact on children and families. The volume then describes how we arrived at our current situation in the United States, focusing on questions related to how we view prisons and prisoners, policing for profit, and the motivations of prosecutors in trying to secure convictions. Finally, Rethinking Punishment in the Era of Mass Incarceration examines specific policy alternatives that might offer solutions to our current approach to punishment and incarceration. Chris W. Surprenant is Associate Professor in Philosophy and Director of the Alexis de Tocqueville Project in Law, Liberty, and Morality at the University of New Orleans, USA. He is the author of Kant and the Cultivation of Virtue (Routledge 2014), co-editor of Kant and the Scottish Enlightenment (Routledge 2017) and Kant and Education: Interpretations and Commentary (Routledge 2011), and has written numerous articles on various aspects of Kant’s moral and political philosophy.

Routledge Studies in Contemporary Philosophy For a full list of titles in this series, please visit www.routledge.com

84 McTaggart’s Paradox R. D. Ingthorsson 85 Perspectives on Ignorance from Moral and Social Philosophy Edited by Rik Peels 86 Self-Reflection for the Opaque Mind An Essay in Neo-Sellarsian Philosophy T. Parent 87 Facts and Values The Ethics and Metaphysics of Normativity Edited by Giancarlo Marchetti and Sarin Marchetti 88 Aesthetic Disinterestedness Art, Experience, and the Self Thomas Hilgers 89 The Social Contexts of Intellectual Virtue Knowledge as a Team Achievement Adam Green 90 Reflective Equilibrium and the Principles of Logical Analysis Understanding the Laws of Logic Jaroslav Peregrin and Vladimír Svoboda 91 Philosophical and Scientific Perspectives on Downward Causation Edited by Michele Paolini Paoletti and Francesco Orilia 92 Using Words and Things Language and Philosophy of Technology Mark Coeckelbergh 93 Rethinking Punishment in the Era of Mass Incarceration Edited by Chris W. Surprenant

Rethinking Punishment in the Era of Mass Incarceration Edited by Chris W. Surprenant

First published 2018 by Routledge 711 Third Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2018 Taylor & Francis The right of the editor to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-1-138-04779-2 (hbk) ISBN: 978-1-315-17060-2 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents

Acknowledgement

Introduction: Why Do We Punish?

vii 1

CHRIS W. SURPRENANT

  1 The Problem of Punishment

15

JOHN HASNAS

  2 Unconscionable Punishment

34

MICHAEL HUEMER

  3 The Coproduction of Justice

49

NATHAN GOODMAN

  4 The Certainty of Punishment and the Proportionality of Incarceration

69

CHRIS BARKER

  5 Imprisonment and the Right to Freedom of Movement

89

ROBERT C. HUGHES

  6 Are There Expressive Constraints on Incarceration?

105

BILL WRINGE

  7 Punishment, Restitution, and Incarceration

122

DAVID BOONIN

  8 Communicative Theories of Punishment and the Impact of Apology EDDY NAHMIAS AND EYAL AHARONI

144

vi Contents   9 A Reparative Approach to Parole-Release Decisions

162

KRISTEN BELL

10 Restorative Justice in High Schools: A Roadmap to Transforming Prisons

180

JOHANNA C. LUTTRELL

11 Reforming Youth Incarceration in the United States

194

CARA H. DRINAN

12 Policing for “Profit”: The Political Economy of Private Prisons and Asset Forfeiture

209

ABIGAIL R. HALL AND VERONICA J. MERCIER

13 Why Paternalists and Social Welfarists Should Oppose Criminal Drug Laws

225

ANDREW COHEN AND WILLIAM GLOD

14 The Need for Prosecutorial Guidelines

242

JOHN F. PFAFF

15 Prison Tunnel Vision

262

JOSH DOHMEN

16 Exile as an Alternative to Incarceration

277

BRIANA MCGINNIS

17 Corporal Punishment as an Alternative to Incarceration

294

JASON BRENNAN

18 The Potentials and Limitations of De-Incarceration

309

DANIEL J. D’AMICO

List of Contributors Index

329 333

Acknowledgement

This publication volume would not have been possible without a generous grant from the Charles G. Koch Charitable Foundation. I am incredibly thankful for their support of this project and the scholars who have contributed to this volume, as well as their support for other scholars working on a wide range of projects in the area of criminal justice reform.

Introduction Why Do We Punish? Chris W. Surprenant

In June 2016, the United States found itself in the middle of a national sentencing controversy. 20-year-old Brock Turner, a star swimmer at Stanford University, was found guilty of sexually assaulting an unconscious 22-yearold woman behind a fraternity house dumpster. Turner was sentenced to six months in prison, three years on probation, and life registration as a sex offender. Public consensus appeared to be that Turner’s punishment was far too lenient and that justice demands a more severe penalty for such violent and reprehensible acts, both to punish the perpetrator and to send a message generally that this type of behavior will not be tolerated. Around the same time, another sentencing controversy was playing out in New Orleans. There, 34-year-old Jacobia Grimes faced a sentence of 20 years to life for stealing $31 in candy from a convenience store. Having been convicted of petty theft four times previously, the district attorney chose to prosecute Grimes as a repeat offender, subjecting him to certain mandatory minimum sentencing. “Twenty years to life for a Snickers bar, or two or three or four,” Judge Franz Zibilich wondered aloud during Grimes’s arraignment. “Isn’t this a little over the top?”1 This volume examines what I believe to be the most important topic in applied moral and political philosophy in the US: criminal justice reform, and, specifically, how we should go about addressing our problem of mass incarceration. As I write this introduction in February 2017, the US is second in the world in per capita incarceration, behind only the small island nation of the Seychelles. 2.3 million people currently occupy American prisons and jails, 0.71% of our population or approximately 1 out of every 140 people. This problem extends beyond jails and prisons and to our criminal justice system more generally. 820,000 Americans are currently on parole, which is a conditional release from prison, and another 3.8 million people are on probation, which often serves as an alternative to prison or jail. In total, a staggering 7 million people—1 out of every 46 Americans—are under the control of the US criminal justice system. While these numbers seem staggeringly high, it may be the case that this outcome is just and that these individuals are being punished appropriately. But it is difficult to think clearly about just and unjust punishment without

2  Chris W. Surprenant considering the question of why we punish, specific examples of punishment, and if those punishments aim at their intended goals. When thinking about justice and punishment our first thoughts likely go to the criminal justice system that we are most familiar with and whether or not we think that certain laws, and the penalties associated with breaking those laws, are fair. We rightly associate justice and fairness, and determining what is fair in any particular situation seems to depend on the circumstances relevant to that case. Factors such as why a person acted contrary to an explicit or understood rule, if the actor intended to violate that rule or cause serious harm, and the likelihood of that person acting badly again in the future all seem as if they should influence what punishment, if any, is the fairest response to undesirable behavior. That it seems necessary to take circumstances beyond the act itself into account when determining an appropriately just punishment does not mean that punishments always are handed out in this way. In nearly every developed country, elected or appointed leaders gather to discuss principles of conduct that the citizens should live by, codify these principles in juridical law, and assign appropriate penalties for individuals who break those laws. Even if we stipulate that these legislators conduct these discussions with the very best intentions, strict adherence to the law and distributing punishments required by the letter of the law appears frequently to miss the mark in achieving justice. The alternative, leaving punishment decisions up to what we hope is the good sense of individual judges or juries, can produce unjust outcomes as well, although perhaps less frequently. No matter which approach is taken, some punishments seem too harsh, others seem too lenient, and some happen to be just right. It may be difficult to put our finger on the cause or suggest obvious paths for reform. For example, by and large our courts seem to get things right. The vast majority of people prosecuted for crimes in the US are guilty of what they have been accused of doing, if not worse. If people are in fact guilty of this supposed criminal behavior, the easy (or only) solution to the problem of mass incarceration is to decriminalize certain behavior, eliminate incarceration as a punishment for offenders who do not pose a significant danger to other people, or take other steps along these lines. The problem, of course, is that taking any steps in this direction seems to present significant practical and political challenges. While the focus of this book is on the justification of incarceration as a method of punishment and possible alternatives to incarceration, these discussions cannot take place without thinking some about punishment generally, including the question of why we punish. Here, my aim is to establish the foundation for the remainder of the discussion in this book by considering the dominant theories of punishment and the goal or goals at which each of these theories is supposed to aim. But trying to decide which of these theories is correct, or at what goal or goals just punishment ought to aim, appears problematic. While it may seem most appropriate to focus

Why Do We Punish?  3 on retribution in some instances, it may seem just as appropriate to focus on rehabilitation in others. How do we decide which principles to weigh more heavily? How we answer this question may be a function of judgment, which involves combining aspects of reason, culture, and experience, things that may be difficult to codify into law.

Theories of Punishment One of the earliest recorded accounts of a system of laws that aims at just punishment dates back to 18th-century BC Mesopotamia and the Babylonian King Hammurabi. For Hammurabi, the role of the state and function of juridical law was “to bring about the rule of righteousness in the land, to destroy the wicked and the evil-doers; so that the strong should not harm the weak” (King translation). Hammurabi’s Code contained 282 “laws of justice” for addressing topics as wide ranging as adoption and inheritance, the appropriate amount of payment for certain services, and penalties for harming other citizens or causing damage to their property. For a code of law that aims at justice, what may be most striking is how severe many of these punishments appear to be. Twenty-eight laws prescribe death as the appropriate punishment for behavior ranging from murder, kidnapping, and negligent homicide to theft, making false accusations, adultery, and withholding compensation from mercenary soldiers. While it may not be surprising to see death prescribed as the appropriately just punishment for murder, it may be surprising to see it as the just punishment for withholding pay from mercenaries. Perhaps more surprising is that for some crimes, such as when a man strikes a free-born woman and causes her to die (#210), death is prescribed as the appropriate punishment not for the perpetrator, but for one of his children. Beyond these 28 laws that prescribe death as the appropriately just punishment, other laws prescribe punishments that may strike us as even more barbaric: amputation of ears and hands, bone breaking, eye gouging, blows from an ox-whip, indentured servitude, and fines to be paid directly to the people who were harmed. Noticeably absent from Hammurabi’s Code are the types of punishment that we see most frequently in our own society—fines paid to the state and imprisonment by the state. Although the Code makes reference to prisons, these prisons appear to have been private and were used for holding individuals who owed debts and were working those debts off through forced labor. The Code provided certain protections for these prisoners and outlined penalties for any individuals who mistreated them or caused their death while in custody. Although it is not clear why the early Babylonians did not utilize fines paid to the state or state imprisonment as methods of just punishment, one reasonable explanation is that they believed these two types of punishment failed to contribute to the goal of restoring the appropriate balance to the community that was upset by the unlawful behavior.

4  Chris W. Surprenant How we determine if any specific instance of punishment is appropriate is connected to what we believe punishing another person is supposed to accomplish. Perhaps we believe, similar to the aim of Hammurabi’s Code, that the aim of punishment is to restore an appropriate balance to society that was upset by the performance of the punishable act. But we could also have other answers to the question of why we punish and how we determine whether or not a particular punishment is just. Generally speaking, there are a handful of dominant approaches to answering this question.2 Some people believe that just punishment must be backward-looking and focus on what an individual has done to deserve being punished. Perhaps the most well-known backward-looking theory of punishment is retributivism. Retributivists generally believe that an individual should suffer penalties proportionate to the harm they inflicted. No harm, no penalty, which would make it unjust to punish an individual for so-called victimless crimes. It may also be difficult to identify appropriately proportional penalties. What is the appropriately proportional punishment for assaulting someone else? A prison term of some fixed length? A fine paid to the victim or the state? Community service? Being similarly assaulted? It is not at all obvious what kind of punishment is appropriately just, never mind the quantity of that punishment. This concern is not insignificant. Other approaches to just punishment, such as deterrence or rehabilitation, are forward-looking and aim to prevent similar bad acts in the future, either by the offender in particular or by others in the community generally. Deterrence theorists focus on punishment as a disincentive to performing bad acts. They argue that the threat of punishment, and then following through on that threat, is most likely to prevent bad behavior in the future. Although this view may seem reasonable in practice, there is little empirical evidence supporting the claim that punishments generally, or increasingly harsher punishments specifically, deter people from behaving badly. Rehabilitation theorists are similarly forward-looking. They argue that our response to bad behavior should focus on identifying and addressing the causes that led to that behavior, much in the same way that a doctor would diagnose and treat an illness. The possibility of rehabilitation assumes that the person’s bad act was somehow not completely free and there is something that can be rehabilitated. There seem to be reasonable arguments to be made for justifying punishment by appealing both to backward-looking and forward-looking theories. Hybrid theories of punishment try to bring together aspects of more than one of these theories in ways that often appeal better to our commonsense intuitions. Both John Rawls and H. L. A. Hart present this kind of position: Rawls in A Theory of Justice and Hart in Punishment and Responsibility. Rawls argues that it is necessary to justify both the institution of punishment and specific instances of punishment, and claimed that utilitarian principles could justify the institution while retributivist principles could justify the punishments themselves. Hart further develops Rawls’s position.

Why Do We Punish?  5 He argues that it is a mistake to claim that all punishment must aim at the same objective. Instead, punishment should be seen as an institution and for it to address a number of different questions such as: “What justifies the general practice of punishment? To whom may punishment be applied? How severely may we punish?” (Punishment and Responsibility, 3). While Rawls and Hart both recognize that identifying principles relevant to just punishment are more complicated than appealing only to a single backward- or forward-looking principle, neither provides a coherent account of how to achieve the desired blend of appropriate goals. Thom Brooks argues that the common problem for all hybrid theories is that different penal goals pull them in different directions. Actions likely to succeed in, for example, satisfying certain retributive goals likely undermine goals of rehabilitation or deterrence. How we decide when to prioritize these different principles seems to be nothing more than personal preference, preferences that we know get applied inconsistently depending on the background of the person who committed the supposed bad act, who his victims were, and the circumstances. Instead of looking to combine multiple different approaches to punishment, Brooks argues that one, coherent framework can aim at and unite this variety of penal goals. Such a unified theory is not unique to Brooks, but he presents one of the more interesting and thoughtful arguments for this position (Punishment, 123–148). He argues that crimes should be understood as “legal wrongs against rights. Punishment aims at the restoration of our rights and this restoration may take multiple shapes to achieve these aims” (132). Brooks draws his inspiration from the Model Penal Code, which provides a similar approach to how punishment can be used to achieve a variety of penal goals. Brooks claims that the Model Penal Code fails to “justify a legal practice without sufficient consideration of how the individual parts coherently work together in support of the practice aims” (133). In contrast, he believes his unified theory, in which punishment is used to bring together these multiple penal goals to restore the rights of individuals, succeeds in providing a theoretical cohesion that the Model Penal Code lacks. Brooks restricts discussions and applications only to violations of juridical law. He writes, “The only reason why the state may punish is because a person has performed a crime. . . . Punishment is a matter of public justice and of a very different character than, [for example,] the disciplining of children by private individuals” (3). Even though Brooks acknowledges that we use the term ‘punishment’ when talking about a man and his dog, a father and his children, or a boss and his employees, his argument wants to exclude “arbitrary executive decisions made by private individuals outside of a legal system” (2). Punishment should be associated with decisions made inside of a legal system because those decisions involving a loss are non-arbitrary and publicly binding. The most obvious objection is that this claim is false, at least when it comes to these decisions not being arbitrary. Most actual legal systems are

6  Chris W. Surprenant filled with laws that are either arbitrary or designed by legislators to advance private interests instead of the public good. Beyond the laws themselves, judges in these systems frequently hand out punishments in ways that can be described as being arbitrary (and that is being charitable). Perhaps the arbitrary nature of certain laws or judicial decisions would allow us to identify those laws or decisions as unjust, or perhaps we believe that laws are socially constructed generally and that there will always be a hint of what could be described as arbitrariness. This debate between the legal positivists and natural law jurists has gone on for quite a while, and I do not wish to rehash it or weigh in on it here. Instead, I take the important bit of Brooks’s comment about non-arbitrariness to mean that what is unique about punishment inside of a criminal justice system is that it is implemented according to clearly defined rules and procedures. The same is not true when a father punishes his child. In that case, there are no clearly defined rules or procedures for determining which type of punishment he should implement or the severity of that punishment. So while Brooks seems correct in saying that a criminal justice system should aim “at the restoration of our rights and this restoration may take multiple shapes to achieve these aims,” it is not at all clear why punishment is what brings together these multiple goals. For Brooks, punishment is the response—seemingly any response—“involv[ing] a loss” to an individual who has committed a crime (5). Even if we accept this definition, many of these possible reasonable responses he wants to make room for seem to involve no loss at all, at least under any commonsense understanding of the term. Most obvious are responses falling within the realm of rehabilitation. Under certain circumstances bad behavior frequently is thought of as something akin to a disease or illness. If the appropriately just response in these cases is for the perpetrator to enter a rehabilitation program, then it does not take much creative thinking to see how this response fails to “involve a loss” in any significant way. The problem with Brooks’s position is not that punishment should involve a loss—that much seems correct. Rather, he does not tell a complete story for why we punish and why punishment is able to address the seemingly incompatible aims represented by the different theories of punishment that he wants to reconcile. Brooks also misses an opportunity by restricting his discussion of punishment as narrowly as he does. If we are operating from a commonsense understanding of just behavior as that which generally aligns with what is fair, and our challenge is to apply this principle of fairness to punishment inside of a criminal justice system, one of the reasons for this difficulty is that most people either are unfamiliar with the criminal justice system in any meaningful way or see it as a set of rules that will never be applied to them. It is almost impossible to think intelligently about what constitutes fair behavior when either one or both of these factors is in play. The solution to these difficulties may not be as complicated as Brooks seems to think. While there are not a lot of coherent, hybrid theories of

Why Do We Punish?  7 punishment in the history of philosophy, Immanuel Kant presents such a theory that receives little attention from people who are not Kant scholars. His hybrid theory receives little attention because the majority of scholars working on punishment attribute to Kant a view that is strictly retributivist, but that attribution is not entirely accurate. Although Kant’s account of punishment is not without some problems, understanding his position proves especially helpful in answering the question of why we punish and thinking through how we should go about punishing people in practice.

Kant’s Theory of Punishment In the history of philosophy, Kant offers what may be the most interesting hybrid theory of punishment. Although Kant frequently is attributed with a solely retributive approach to punishment, this understanding of his position is inconsistent with the text. While the primary thrusts of his position are retributive, he also identifies necessary deterrent elements that, if absent, seem to make punishment pointless. For Kant scholars, his approach is interesting due primarily to his attempt to reconcile what he believes to be multiple answers to the question of why we punish within his somewhat complex moral and political philosophy. For those of us interested in punishment and legal theory more generally, Kant’s approach is worth considering not because he arrives at a coherent solution, but because it helps us to recognize many (if not all) of the relevant considerations when thinking about just punishment. Kant’s justification of punishment is rooted in his account of human freedom and the external conditions necessary for the possibility of that freedom. He argues that because (1) individuals are under a moral obligation to act autonomously; (2) autonomous action is possible in practice only if an individual’s life, health, liberty, and possessions are secured; and (3) the only mechanism to realize this security is through establishing and maintaining civil society; therefore, (4) individuals “do wrong in the highest degree” by failing to enter or remain in this condition (MM 6:308).3 For Kant, the defining feature of civil society is austeilende Gerechtigkeit or “distributive justice.” While the contemporary use of “distributive justice” often is meant to identify conditions under which incidental inequalities in outcome do not arise, Kant’s condition of austeilende Gerechtigkeit means simply that laws exist “to determine for each what land [or property generally] is mine or yours” (MM 6:267), and are enforced by an appropriately empowered magistrate in a manner that is consistent with the moral equality of persons under the law. To maintain civil society people must see each other as moral equals, and actions that appear to violate this principle of moral equality undermine this civil condition. The appropriate and necessary response to such actions is to respond publicly in a way that aims at restoring this condition of moral equality. That response is punishment. Put differently, when

8  Chris W. Surprenant an individual breaks the law, either juridical or moral, he undermines this condition of equality by making himself “an exemption to the rule” (MM 6:321). But punishing violations of the moral law is tricky. For Kant, an individual violates the moral law when he adopts a principle for action that cannot be willed universally. However, Kant claims it is impossible to work backwards and determine if someone has violated the moral law based on his actions. As a result, punishment for violations of the moral law must come from either God or the individual himself, when he “find[s], upon self-examination, that he is worthless and contemptible in his own eyes” (CPr 5:161; see also TP 8:288). Violations of juridical law are more straightforward. An individual violates juridical law when he either performs a prohibited act or fails to perform a required act. By acting (or failing to act) in this way he exempts himself (covertly, in most cases) from a general rule of behavior that everyone in that community is expected to follow. Similar to violations of the moral law, violations of juridical law undermine the condition of equality that makes civil society possible. An individual who violates the law declares he is not subject to the same rules that apply to other members of the community. But unlike violations of the moral law that are based on principles of action and cannot be rectified publicly, violations of juridical law are based on the actions themselves. Once it comes to light that someone, through his actions, has declared himself to be unequal in this way, the condition of civil society is undermined until equality is restored (at least in the eyes of the public). The act of punishing a lawbreaker aims to bring him back down to the level of the community and restore this condition of equality. Since punishment aims to restore this condition of moral equality, Kant argues that the appropriate kind and degree of punishment must be proportional to the crime committed. Holding to this conclusion leads him to a number of positions on punishment that seem controversial, most notably for their seemingly extreme conclusions. The most frequently cited is his claim that the death penalty must be administered to murderers without exception, going so far to argue that even if people living together agree to disband civil society, before they can do so legitimately they are under a moral obligation to execute any and all of the murderers in prison. For Kant, executing a murderer “does to him what his deeds deserve and blood guilt does not cling to the people for not having insisted upon this punishment; for otherwise the people can be regarded as collaborators in this public violation of justice” (MM 6:333).4 That just punishment must satisfy this principle of proportionality5 often leads scholars to classify Kant’s account of punishment as retributive. But classifying Kant in this way oversimplifies his position. One obvious place where Kant departs from strict retributivism is his frequent comments on the importance of punishment deterring future bad acts.6 He even goes so far to suggest that it is not appropriate to punish someone in cases where

Why Do We Punish?  9 punishment would have no deterrent effect. Here, he cites examples such as a man in a shipwreck who pushes another man off a floating plank so that he may survive (TP 3:300n and MM 6:235–6),7 a mother who kills her baby born out of wedlock (MM 6:336),8 and a military officer who kills another officer in a duel that was initiated after an honor dispute (ibid.).9 What Kant believes connects these acts and other examples of morally blameworthy but not punishable bad behavior is that the appropriate punishment can never be worse than the apparent consequences of inaction. The shipwrecked man at sea faces certain death unless he acts to save his life, killing another man in the process. The mother and military officer face ridicule and shame that Kant believes to be worse than death.10 Even if Kant is wrong about this latter claim, his claims that there are certain bad acts that are not punishable for reasons of deterrence, and that punishment should have a deterrent effect generally, present obvious challenges when trying to reconcile these views with what is generally presented as a strict retributivist justification of punishment. Further complicating Kant’s view of punishment is an entirely separate line of reasoning where he claims punishing an individual for bad behavior is necessary to communicate publicly that the actor is a rational being and responsible for his actions. This communicative use of punishment is different from its retributive use in restoring equality and is important for maintaining the condition that holds civil society together. Civil society is possible as a condition of moral and legal equality because its members are free beings capable of moral decision making. If individuals were not responsible for their behavior in this way, it would not be impossible to enter into the type of agreements that provide the foundation for civil society. Since “[p]unishments happen only to a will that is free but contrary to the law” (TP 8:288), a person is punished because he wills a punishable act (MM 6:335).11 If someone performed a punishable act but then was not punished, the only legitimate reason for not doing so is because he was somehow not responsible for that behavior. In other words, punishment is the mechanism by which we hold people responsible for their bad acts. While it is reasonable not to hold someone responsible under certain conditions (e.g., a driver who kills another person with his car after suffering a heart attack at the wheel), exempting or excusing individuals outside of these unique circumstances undermines civil society. Kant seems to recognize this problem himself. In writing on clemency, he argues that a sovereign’s right to grant clemency is “the slipperiest one he can exercise” and can only be justified when a subject has committed a crime against the sovereign and not against another citizen (MM 6:337). In these latter cases, granting clemency or issuing a lenient punishment would be “the greatest wrong against his subjects” (ibid.). How, then, can Kant claim that punishment is not justified in those instances where it lacks a deterrent effect? Wouldn’t it still be necessary to punish the offenders for communicative or retributive reasons?

10  Chris W. Surprenant A handful of scholars have attempted to reconcile Kant’s seemingly inconsistent views on punishment, Sharon Byrd and Arthur Ripstein most notably. Byrd argues that Kant should be understood as talking about “two [different] facets of punishment, each independent but nevertheless mutually restrictive. Punishment as a threat was intended to deter crime. . . . [But] in its execution . . . the state was limited in its reaction by a retributive theory of justice” (152). Byrd’s solution to resolving the problem presented by both the mother who kills her child born out of wedlock and the person who kills another in a duel rests on her claim that Kant was writing within a context where “honor was more highly valued than life” (200). As a result, it would be appropriate not to punish in either situation because “both present cases for Kant in which executing the death penalty for wrongful killings contradicts the more highly valued sense of honor preserved by the deed” (200). Byrd’s analysis of these two scenarios ends rather abruptly here. But what she seems to be suggesting is that what is important when determining the appropriate punishment is what society believes is the appropriate retributive response for seemingly bad behavior. She believes that Kant would claim that a society that values honor more than life, a position she attributes to Kant, would recognize these unique circumstances involved in these killings and create general rules that did not penalize such behavior. This position seems like a stretch, and there’s very little direct evidence in Kant’s writings to support it. Further, it does not explain away the example of the shipwrecked man on the plank. He does not act for the sake of honor, but to protect his own life. If we believe that Kant is operating from a position where one’s honor is worth more than one’s life, then his position regarding the man on the plank is inconsistent. What he gives us there is a pretty straight argument for the necessity of just punishment having some sort of deterrent effect. More recently, Ripstein has taken a different approach to resolving the apparent inconsistencies between Kant’s retributive and deterrent justifications of punishment. Instead of arguing that retribution or deterrence are two distinct aims of punishment, aims that, perhaps, cannot be reconciled in certain circumstances, Ripstein claims a “Kantian account must analyze punishment as a fundamental aspect of legality, and show how each of deterrence and retribution is partially constitutive of a system of equal freedom under law” (301). His argument focuses on the role of justified coercion as hindering a hindrance to freedom. “Kant’s conception of retribution [is] an expression of this idea, . . . [in that] punishment is nothing more than the supremacy of law; . . . deterrence [falls] under the same principle, . . . [in] that the supremacy of law requires that the prospect of enforcement be capable of guiding conduct” (301–302). Ripstein concludes, “Kant generates each of these aspects of punishment as an a priori feature of public law, rather than as a response to destabilizing features of human nature. . . . The prospective threat and retrospective applications of punishment are thus

Why Do We Punish?  11 not an aim to be pursued and an extrinsic constraint on its pursuit; they are equivalent” (302). While Ripstein identifies his position as Kantian, he is also clear to separate it from the position Kant actually advanced, going so far to call it “[his] own account” of punishment (ibid.), and it, along with other hybrid theories or separate frameworks, can be evaluated independently of Kant’s position. What should we take away from Kant’s discussion of punishment, as well as the subsequent scholarly discussions surrounding Kant’s position and how to reconcile certain aspects of that position? What seems most important is that arriving at an answer to the question, “Why do we punish?” is not as obvious as it may seem. What makes answering this question especially difficult is that it may have multiple answers depending on who is asked and what he believes the aims of punishment to be. The difficulty here is not in identifying the various answers or understanding why people may approach punishment differently in these ways, but in trying to decide on the appropriate aims of relevant public policy measures when the multiple answers to this question pulls us in entirely different directions. While the US has gone in the direction of incarceration and ever-increasing sentence length for those who have been incarcerated, this volume examines why this approach to punishment is unjust in many cases and suggests alternative directions for just punishment. The aim of these discussions is not to provide the final word on any of these issues, but to suggest paths for future work and possible directions for public solutions that aim to address our problem of mass incarceration.

Notes 1 Franz Ziblich, as quoted in John Simerman’s, “Accused New Orleans candy snatcher facing 20 years for pocketing $31 in sweets.” http://theadvocate.com/ news/neworleans/neworleansnews/15365240-79/accused-new-orleans-candysnatcher-facing-20-to-life-for-pocketing-31-in-sweets 2 For a more thorough discussion of the current literature surrounding punishment theory, please see Thom Brooks’s excellent book on this subject, Punishment. 3 That this obligation for Kant is moral, and not merely practical, distinguishes his position from that of someone like Thomas Hobbes, who argued that the reason an individual ought to leave the state of nature and enter civil society is practical and connected to an individual’s desire to preserve his life. 4 MM 6:333. That entire passage reads: “If, however, he has committed murder he must die. Here there is no substitute that will satisfy justice. There is no similarity between life, however wretched it may be, and death, hence no likeness between the crime and the retribution unless death is judicially carried out upon the wrongdoer, although it must still be freed from any mistreatment that could make the humanity in the person suffering it into something abominable.—Even if a civil society were to be dissolved by the consent of all its members (e.g., if a people inhabiting an island decided to separate and disperse throughout the world), the last murderer remaining in prison would first have to be executed, so that each has done to him what his deeds deserve and blood guilt does not cling

12  Chris W. Surprenant to the people for not having insisted upon this punishment; for otherwise the people can be regarded as collaborators in this public violation of justice. This fitting of punishment to the crime, which can occur only by a judge imposing the death sentence in accordance with the strict law of retribution, is shown by the fact that only by this is a sentence of death pronounced on every criminal in proportion to his inner wickedness (even when the crime is not murder but another crime against the state that can be paid for only by death).” 5 Kant’s connection between justice and proportional punishment is outlined clearly in the Metaphysics of Morals: “But what kind and what amount of punishment is it that public justice makes its principle and measure? None other than the principle of equality (in the position of the needle on the scale of justice), to incline no more to one side than to the other. . . [O]nly the law of retribution (ius talionis)—it being understood, of course, that this is applied by a court (not by your private judgment)—can specify definitely the quality and the quantity of punishment; all other principles are fluctuating and unsuited for a sentence of pure and strict justice because extraneous considerations are mixed into them” (MM 6:332, Kant’s emphasis). 6 Another example from Kant where he argues that it would be justified not to implement the appropriately just punishment is when such a large number of individuals would need to be punished for the bad behavior that it would (1) jeopardize the future existence of the state, or (2) offend the sensibilities of the citizenry due to the large number of people who need to be punished. He writes, “If, however, the number of accomplices (correi) to such a deed is so great that the state, in order to have no such criminals in it, could soon find itself without subjects; and if the state still does not want to dissolve, that is, to pass over into the state of nature, which is far worse because there is no external justice at all in it (and if it especially does not want to dull the people’s feeling by the spectacle of a slaughterhouse), then the sovereign must also have it in his power, in this case of necessity (casus necessitatis), to assume the role of judge (to represent him) and pronounce a judgment that decrees for the criminals a sentence other than capital punishment” (MM 6:334). 7 Those passages read: “[B]ut if it is said of someone who, in order to preserve his own life, pushes another survivor of a shipwreck from his plank, that he has a right to do so by his (physical) necessity, that is quite false. For to preserve my life is only a conditional duty (if it can be done without a crime); but not to take the life of another who is committing no offense against me and does not even lead me into the danger of losing my life is an unconditional duty. Yet teachers of general civil right proceed quite consistently in conceding rightful authorization for such extreme measures! For the authorities can connect no punishment with the prohibition, since this punishment would have to be death. But it would be an absurd law to threaten someone with death if he did not voluntarily deliver himself up to death in dangerous circumstances” (TP 8:300n). And: “In other words, there can be no penal law that would assign the death penalty to someone in a shipwreck who, in order to save his own life, shoves another, whose life is equally in danger, off a plank on which he had saved himself. For the punishment threatened by the law could not be greater than the loss of his own life. A penal law of this sort could not have the effect intended, since a threat of an ill [236] that is still uncertain (death by a judicial verdict) cannot outweigh the fear of an ill that is certain (drowning). Hence, the deed of saving one’s life by violence is not to be judged inculpable (inculpable) but only unpunishable (impunibile), and by a strange confusion jurists take this subjective impunity to be objective impunity (conformity with law)” (MM 6:235–6, Kant’s emphasis).

Why Do We Punish?  13 8 That passage reads: “A child that comes into the world apart from marriage is born outside the law (for the law is marriage) and therefore outside the protection of the law. It has, as it were, stolen into the commonwealth (like contraband merchandise), so that the common- wealth can ignore its existence (since it was not right that it should have come to exist in this way), and can therefore also ignore its annihilation; and no decree can remove the mother’s shame when it becomes known that she gave birth without being married” (MM 6:336). 9 That passage reads: [W]hen a junior officer is insulted he sees himself constrained by the public opinion of the other members of his estate to obtain satisfaction for himself and, as in the state of nature, punishment of the offender not by law, taking him before a court, but by a duel, in which he exposes himself to death in order to prove his military courage, upon which the honor of his estate essentially rests. Even if the duel should involve killing his opponent, the killing that occurs in this fight which takes place in public and with the consent of both parties, though reluctantly, cannot strictly be called murder (homocidium dolosum)” (ibid., Kant’s emphasis). 10 Later in the Metaphysics of Morals, Kant provides a slight revision to his position on the appropriate response to these two examples, writing: “There are, however, two crimes deserving of death, with regard to which it still remains doubtful whether legislation is also authorized to impose the death penalty. The feeling of honor leads to both, in one case the honor of one’s sex, in the other military honor, and indeed true honor, which is incumbent as duty on each of these two classes of people. The one crime is a mother’s murder of her child (infonticidium maternale); the other is murdering a follow soldier (commilitonicidium) in a duel.—Legislation cannot remove the disgrace of an illegitimate birth any more than it can wipe away the stain of suspicion of cowardice from a subordinate officer who fails to respond to a humiliating affront with a force of his own rising above fear of death. So it seems that in these two cases people find themselves in the state of nature, and that these acts of killing (homocidium), which would then not have to be called murder (homocidium dolosum), are certainly punishable but cannot be punished with death by the supreme power” (MM 6:335–6). The minor, but important, modification is that while these crimes are not punishable by death, they are “certainly punishable.” This change is odd in part because of his previous comments noting that the only appropriate punishment for murder is death and his claim that in cases where a “right of necessary” can be invoked punishment cannot have a deterrent effect and so it is not justified. Here, what Kant appears to be opening the door to is a third justification for punishment, one that does not connect directly to retributive or deterrent principles. 11 This claim is different from the position that an individual wills to be punished, which Kant denies is the case. At MM 6:335, Kant writes: “No one suffers punishment because he has willed it but because he has willed a punishable action; for it is no punishment if what is done to someone is what he wills, and it is impossible to will to be punished.”

Works Cited Brooks, Thom (2013). Punishment. New York and London: Routledge. Hart, H.L.A. (2008). Punishment and Responsibility. Oxford: Oxford University Press. Kant, Immanuel (1999a). “Critique of Practical Reason.” In The Collected Works of Immanuel Kant: Practical Philosophy. Cambridge: Cambridge University Press.

14  Chris W. Surprenant Kant, Immanuel (1999b). “Metaphysics of Morals.” In The Collected Works of Immanuel Kant: Practical Philosophy. Cambridge: Cambridge University Press. Kant, Immanuel (1999c). “Theory and Practice.” In The Collected Works of Immanuel Kant: Practical Philosophy. Cambridge: Cambridge University Press. Rawls, John (1999). A Theory of Justice. Cambridge, MA: Harvard University Press. Ripstein, Arthur (1999). Force and Freedom: Kant’s Legal and Political Philosophy. Cambridge, MA: Harvard University Press.

1 The Problem of Punishment John Hasnas1

I. Introduction The question of what justifies the imposition of criminal punishment has fascinated moral and legal philosophers for generations. Is it retribution for a wrong done? The deterrence of future wrongdoing? The rehabilitation of the wrongdoer? The expression of society’s outrage over the wrongdoing? Some combination of the four? Over the years, countless articles have been written in the effort to define the proper purpose of criminal punishment. Virtually every textbook on the philosophy of law has a section devoted to this question. The topic is so frequently discussed in the academic setting that it has acquired the designation, “The Problem of Punishment” (see, e.g., Feinberg and Gross 1995; Murphy and Coleman 1990; Adams 1992; Schauer and Sinnott-Armstrong 1996; Arthur and Shaw 1993; Kent 1970; Golding 1975). Indeed, in the ancient days when I was a first-year law student, my first assignment in Criminal Law was to read the section of Kadish and Paulsen’s Criminal Law and its Processes entitled “The Problem of Punishment” (Kadish and Paulsen 1969: 63). Yet, despite the title of the present chapter, the problem of identifying the proper purpose of criminal punishment is not the problem I intend to address. That’s because referring to this as the problem of punishment is a misnomer. Trying to identify the moral purpose of criminal punishment is an intriguing and important philosophical question, under the assumption that criminal punishment is justified in the first place. But as the qualifying phrase suggests, it is not the problem of punishment. The problem of punishment is demonstrating that the state is morally justified in inflicting punishment on its citizens. For if it is not, the question of the morally proper purpose of criminal punishment is moot. This chapter addresses the more fundamental question. I will ask whether criminal punishment can be morally justified in a liberal society. Despite the conventional wisdom on this matter, I will suggest that the answer to this question is not an obvious yes, but rather turns on matters of empirical fact. Further, although I will not attempt to supply an authoritative

16  John Hasnas demonstration of the conclusion, I will suggest that there is good reason to suspect the ultimate answer to the question is no. To do this, I proceed as follows. In Part II, I define the limitations of the inquiry. In Part III, I clarify the nature of criminal punishment and distinguish it from other dispute settlement mechanisms and civil liability. In Part IV, I present an argument that suggests that a system of criminal punishment may not be necessary to maintain an orderly society and, thus, may not be justified in a liberal society. Finally, in Part V, I conclude.

II. Limitations of the Inquiry In this chapter, I am exploring whether the state is morally justified in punishing those who violate the criminal law. In doing so, however, I do not want to become embroiled in controversial and difficult questions about the proper scope of the criminal sanction. For example, the question of whether it is morally proper to punish purely self-regarding action that does not harm others is one of the perennial problems of criminal jurisprudence. However, the justification for punishing victimless crimes or the lack thereof is not material to the present inquiry. Similarly, the question of whether it is acceptable to punish those who act without consciousness of wrongdoing—the problem of strict liability crimes and crimes of negligence—is a contentious one. But once again, it is not material to the subject of this chapter. This chapter is not concerned with whether the criminal sanction as currently applied exceeds the scope of its morally proper boundaries. It is concerned with whether there is moral justification for any criminal punishment at all. Therefore, its focus is restricted to the uncontroversial core applications of the criminal sanction to offenses such as murder, assault, rape, and theft in which there is both harm to others and intentional wrongdoing. If criminal punishment is not justified for these offenses, then criminal punishment is not justified period. A second limitation on this inquiry is that it applies only to a liberal society. Liberalism is usually identified with John Stuart Mill’s assertion that “the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others” (Mill 1975: 10), or as Joel Feinberg expressed it, “[p]aternalistic and moralistic considerations, when introduced as support for penal legislation, have no weight at all” (Feinberg 1984: 15). However, because I wish to avoid the controversy over the justifiability of paternalistic legislation, for purposes of this chapter, I appeal to the minimalist conception of liberalism embodied in what has been called the Fundamental Liberal Principle, which holds that “freedom is normatively basic, and so the onus of justification is on those who would limit freedom, especially through coercive means” (Gaus and Courtland 2014). This principle creates a general presumption in favor of liberty that cannot be overridden merely to improve the moral character of

The Problem of Punishment  17 the citizens or to establish any particular religious or secular conception of the good (see, e.g., Feinberg 1973: 20–2; Husak 1983: 345; Moore 1997: 749–50). A liberal society does not require its citizens to practice a state religion, exhibit patriotism (or internationalism), adopt Puritan (or libertine) attitudes toward sexual activity, or profess Judeo-Christian (or any other culturally specific set of) values. This is often expressed colloquially by saying that the government of a liberal state is not authorized to “legislate morality.” The argument advanced in this chapter will not work in a theocracy. For that matter, it will not work in a society that has officially adopted utilitarianism or egalitarianism or Kantianism as its standard of moral value. It will work only in a liberal society—one in which the state is barred from offering the realization of any particular moral ideal as a justification of the use of its coercive power.

III. The Nature of Criminal Punishment A. What Criminal Punishment Is Punishment is the coercive imposition of a harm upon a party in response to that party’s failure to behave as required by some binding code of conduct. Ordinarily, coercing others is a wrong. What distinguishes punishment from the ordinary application of coercion is that the harm imposed by punishment is deserved. It is the link between the coercion applied and the violation committed by the individual to whom it is applied that renders the coercion morally acceptable. Placing the adjective “criminal” before the word “punishment” identifies who is imposing the harm and what constitutes the offense. Punishment is criminal punishment when the state is imposing the harm for the violation of the state’s rules of criminal law. Hence, criminal punishment is the coercive imposition of a harm by the state on a party who has violated the criminal laws of that state. B. What Criminal Punishment Is Not Criminal punishment is not a mechanism for resolving interpersonal disputes. One way of dealing with situations in which one member of society harms or comes into conflict with another is to allow the parties to resolve the conflict through negotiations. In some cases, such negotiations could be facilitated by third-party mediators; perhaps even state appointed mediators. But as long as the state agents act only as mediators—as long as the state does not visit harm on one of the parties as a response to that party’s violation of a rule of criminal law—such efforts at composition do not involve criminal punishment. Criminal punishment is also not a vehicle for restitution. When one member of society wrongfully harms another, he or she may owe the injured

18  John Hasnas party restitution. The system of civil liability is designed to vindicate this interest. Thus, when one intentionally or negligently harms another, one may be required to restore the harmed party to the position he or she would have been in had the intentional or negligent conduct not occurred. Indeed, when one acts in a way that creates an inordinate risk of harm—when one engages in abnormally dangerous activity—and another suffers harm as a result, one may be required to restore the harmed party even though one has exercised all due care.2 Civil liability is not punitive. The payment required of the defendant is not a penalty imposed in retribution for an act of wrongdoing. It is restitution to compensate the harmed party for the loss he or she suffered at the hand of the defendant. It is true that in our system of civil liability, parties who are harmed by the intentional misconduct of others can recover exemplary damages, which are payments in addition to and beyond the amount required to make the injured party whole. Although such damages are typically referred to as punitive damages, their purpose is not literally to punish the defendant, but to discourage others from engaging in similar conduct. Further, even if exemplary damages are viewed as a punitive element of the civil liability system, they still do not constitute criminal punishment because they are not imposed by the state and the reason for their imposition is not a violation of the criminal code of the relevant jurisdiction. Criminal punishment is wholly distinct from civil liability. It is a harm imposed on individuals who violate the rules of the criminal law that is independent of any finding of civil liability. It is an additional penalty imposed by the state beyond any compensatory and exemplary damage payments that one may be required to make to an injured party.

IV. The Argument A Criminal punishment is inherently coercive. It necessarily involves the state employing coercion against its citizens. Therefore, in a liberal society, it requires justification. To be morally legitimate, criminal punishment must be necessary to attain some end morally more important than individual liberty. In a liberal society, the government may not employ coercion in the effort to improve the moral character of the individual citizens. So, in a liberal society, the mere desire to enforce any particular conception of morally proper action cannot by itself justify the use of coercion inherent in criminal punishment. What other candidates are there for the necessary justification? Perhaps criminal punishment can be justified to prevent harm to others, as suggested by John Stuart Mill; or to protect the individual rights to life, liberty, and

The Problem of Punishment  19 property, as suggested by John Locke; or to minimize the overall number of infringements on individual liberty; or to discourage violence sufficiently to allow peaceful cooperation to flourish; or to promote the general welfare; or . . . Any of these morally worthy goals could justify the state in using the coercion inherent in criminal punishment. But note that merely showing that criminal punishment is effective in promoting these ends will not suffice. One must demonstrate not just that criminal punishment facilitates the attainment of the end, but that it is necessary for it—that the end cannot be achieved without the use of state coercion. Extortion will often be a more effective means of raising money for a worthy cause than soliciting voluntary donations, but that alone does not justify the use of coercion. It would certainly be easier for the state to expropriate private property to build a dam than to bargain for it on an open market or be required to pay just compensation, but that does not entitle a liberal state to do so. Using force to attain one’s ends is usually more efficient than having to persuade others to cooperate, but that cannot be enough to overcome the presumption of liberty. Thus, the question at the heart of “the problem of punishment” is the question of whether criminal punishment is necessary to achieve an important moral end. But because most people believe that the answer is an obvious yes, the question is rarely asked. Most people take it for granted that criminal punishment is required to maintain order in society; that without it, society would devolve into a Hobbesian war of all against all. If so, the philosophically difficult question of which ends are morally significant enough to overcome the presumption of liberty is moot. One need not determine whether the goal to be served is preventing harm to others, protecting individual rights, minimizing the number of infringements on individual liberty, promoting peaceful cooperation, or promoting the general welfare because achieving any of them requires reducing the amount of interpersonal violence in society to some acceptably low level. And if a system of criminal law and punishment is necessary to do this, then such a system is morally justified. This argument, which is compelling if its premises are true, rests crucially on an underlying assumption: that a system of criminal law and punishment is necessary to maintain an orderly society understood as one in which the level of interpersonal violence has been significantly repressed. But what if this is not true? B The format of the current chapter does not permit a full presentation of the empirical and theoretical considerations required to demonstrate that criminal punishment is not necessary to the maintenance of an orderly society. The best that I can do in the present context is to introduce a few suggestive comments designed to cast doubt on the proposition. To that end,

20  John Hasnas please consider the following historical, psychological, and philosophical observations.

1. History Philosophers love to speculate about how human beings move from a state of nature to an ordered society. English history provides empirical evidence of how this occurs in the real world. Early in the fifth century, the Romans, who had ruled Britain for three and half centuries, suddenly left the island, taking their civil government with them. As the literate Romans left, so did their law, and as the legions departed, so did its enforcement. This left the inhabitants living in kinship and tribal groupings exposed to outside aggression and with no authoritatively established mechanisms for dealing with internal disputes; a good approximation of the state of nature. What happened? Telescoping a thousand years of legal history into a few words: tort law evolved. In the absence of any centralized authority, some mechanism was needed for resolving conflicts and discouraging violence. The earliest such mechanism consisted of efforts to compose disputes through negotiations. Rather than fight, parties would bargain with each other to redress injuries, with the members of the relevant kinship groups or communities serving as mediators to aid the process (Goebel 1937: 25–44). Through repetition over time, such compositions gave rise to “extraordinarily detailed schedules of tariffs . . . for various injuries” (Berman 1983: 54), with the result that “[t]he institution of fixed monetary sanctions payable by the kin of wrongdoer to the kin of the victim was a prominent feature of the law of all the peoples of Europe prior to the twelfth century” (Berman 1983: 55). This nascent system of tort law was, in many respects, a very sensible system. The threat of heavy financial burdens upon the wrongdoer and his kin is probably a more effective deterrent of crime than the threat of capital punishment or corporal mutilation, . . . and at least equally as effective as the modern sanction of imprisonment; and it is surely less expensive for society. Moreover, in terms of retributive justice, not only is the wrongdoer made to suffer, but in addition—in contrast to today’s more “civilized” penology—the victim is thereby made whole. (Berman 1983: 55) What we might today recognize as criminal law did not come into being until the 12th century.3 Moreover, the newly recognized offenses against the community that could be punished by the lord or king did not replace the underlying compositional/tort system, but existed side by side with it (Olson 2006: 69). This early criminal law was introduced, not because it was necessary to maintain an orderly society, but to gratify a taste for vengeance.

The Problem of Punishment  21 The distinction between crime and tort was not a difference between two kinds of wrongful acts. In most instances, the same wrong could be prosecuted either as a crime or as a tort. Nor was the distinction a difference between the kinds of persons who could initiate the actions. Victims could initiate actions of both kinds. According to the lawyers, victims who preferred vengeance over compensation prosecuted their wrongdoers for crime. Victims who preferred compensation over vengeance sued their wrongdoers for tort. Royal officers prosecuted wrongdoers for crime on the king’s behalf when victims feared or chose not to do so. (Seipp 1996: 59–60) Further, as counterintuitive as it may seem to contemporary sensibilities, a major purpose of early criminal law was not to punish, but to drive the parties to compose the dispute.4 The threat of punishment was accompanied by various procedural impediments to trial—e.g., benefit of clergy, flight to sanctuary—and opportunities to settle the dispute—e.g., agreements to make amends at a “loveday”—that were designed to encourage settlement rather than the pursuit of a conviction (Olson 2006: 74–5). Indeed, evidence suggests that in the 13th century between 40 and 67 percent of “appeals”—cases brought by criminal charge—were settled (Klerman 2002: 16). Moreover, among those cases that were not settled, “the rate of jury acquittal remained enormously high throughout the period” (Olson 2006: 74) because the medieval jury’s high acquittal rate reflected deeply engrained notions of how social harmony was to be maintained through composition with, rather than ultimate rejection of, the offender. Strikingly, complaints about the acquittal rate were infrequent. Between the fourteenth and sixteenth century in England, not a single jurist nor chronicle writer complained that the jury conviction rate was too low. (Olson: 75–6) In point of fact, criminal law arose not to provide order, but predominantly to serve the financial interests of the Crown. The original difference between tort and crime was who got paid. The action was a tort if the payment went to the party that suffered harm at the hands of the defendant. It was a crime if the payment went to the king. This created a financial incentive for the Crown to continually expand the scope of actions that were considered crimes, which is precisely what the Norman kings of England did. The profits from justice also stemmed from the royal prerogative and constantly increased as the royal justice expanded its jurisdiction. In fact, the expansion of the king’s justice in the period of its early development can be explained satisfactorily only by the royal need for money.

22  John Hasnas Any case tried in the royal court netted the king a sizable fee, and all the profits from those actions that fell under the category of the king’s pleas went into the royal hand. (Lyon 1980: 163)5 This financial incentive also accounts for the introduction of the concept of felony into English law, which resulted in the massive expansion of the scope of the criminal law. Felony originally referred to the feudal crime of betraying or committing treachery against one’s lord. Such an offense, which was not necessary for and played no role in the maintenance of civic order, was an extraordinarily useful mechanism for raising royal revenue. Again royal greed seems to be the best explanation for the expansion of the concept of felony. Any crime called a felony meant that if the appellee was found guilty his possessions escheated to the king. The more crimes called felonies, the greater the income, and so the list of felonies continued to grow throughout the twelfth century. The fortuitous yet conscious combination of the royal prerogative with the concepts of the king’s peace and felony gave royal justice a flexibility that enabled the royal court to spread its jurisdiction over all sorts of places and men and over almost any type of offense. (Lyon 1980: 190) Far from being necessary for the maintenance of social order, the introduction of criminal law tended to undermine it “by effectively destroying the system of community composition” (Laster 1970: 75). Motivated by the desire for revenue, the Crown took steps to ensure that offenders paid fines to the king in preference to restitution to their victims. Thus, it prohibited victims from making settlements that included an agreement not to prosecute, suspended the civil remedies of victims until after the criminal charge was resolved, and prevented victims from receiving stolen goods back until they gave evidence in a criminal proceeding (Laster 1970: 76). As the Crown absorbed more and more of the resources that would otherwise have been available to pay restitution, the incentive to pursue settlement and voluntary composition declined, and was eventually entirely displaced. Without the ability to obtain compensation for their injuries, victims were left with nothing but the desire for vengeance; that is, for the punishment of those who had wronged them.6 Whereas the underlying system of tort law had two important consequences that promoted social harmony and maintained social order: 1) it tended to restore the victim and eliminate his desire for violent revenge, and 2) it benefited [sic] the offender in that he bought back the “peace” and his place in society, . . . the creation of the criminal

The Problem of Punishment  23 law appears to have generated greater social disorder precisely because victims were no longer “restored” to their original level of satisfaction and therefore became more likely to demand severe physical punishment. (Benson 1990: 71) Thus, history suggests that the commonly held belief that criminal punishment was necessary for the preservation of social order has cause and effect reversed. Order was well-maintained by the underlying system of restitution/tort law until the advent of the criminal law with its punitive payments to the Crown rendered that system ineffective. The public’s desire for criminal punishment arose not as a response to disorder, but to the disruptive consequences of the criminal law itself.

2. Psychology One reason for the widespread belief that criminal punishment is necessary to maintain an orderly society is that people assume that there is no other option. When they think about a world without criminal punishment, they imagine a world with no mechanism for discouraging violent, harmful, and dishonest actions. And because such a world calls to mind the chaos of Hobbes’s war of all against all, it is instantly dismissed. But this is not the proper comparison. The options are not criminal punishment or nothing, but criminal punishment or civil liability/restitution. For if the criminal law disappeared, the underlying system of tort law would remain. Therefore, to establish that criminal punishment is necessary for the preservation of an orderly society, one would have to establish first that a system of civil liability alone would not reduce interpersonal violence to an acceptable level, and then that the addition of criminal punishment would. Doing this requires a comparison of the relative effectiveness of civil liability and criminal punishment as a deterrent, which raises a question of human psychology. Civil liability requires one who has wrongfully harmed another to pay financial restitution sufficient to restore the wronged party to the condition he or she would have been in had the wrongful action not occurred. In addition, when the wrongdoing is intentional, civil liability can include exemplary damages that greatly increase that amount the wrongdoers must pay their victims. Criminal punishment can consist in financial fines, physical pain, incarceration, or death. The question then is whether the ability to threaten people with additional financial penalties, pain, imprisonment, or death in addition to the requirement to pay restitution plus exemplary damages is necessary to reduce interpersonal violence to a level that permits a relatively secure life in society. Although people generally assume that the additional punishment is necessary, this is an open empirical question that requires examination.

24  John Hasnas Unfortunately, the present context does not permit a detailed investigation. Here, I can do no more than offer a few observations that cast doubt on the assumption. First, recent psychological research has shown that the brain reacts to the prospect of financial loss in the same way that it reacts to the prospect of physical pain (Seymour et al. 2007; Wellcome Trust 2007), and further, that the fear of financial loss can be as effective a mechanism of “aversive learning” as fear of physical pain (Delgado et al. 2006; Khamsi 2006). This at least suggests that the threat of financial loss may be a more effective deterrent than is generally thought. Indeed, some evidence in support of this suggestion may be found in English legal history. For example, between the 13th and 18th centuries, those accused of serious crimes had to consent to the jury that would try them in order for the trial to proceed. If they did not, they would be imprisoned and subject to peine forte et dure, which consisted of being chained to the floor, fed only bread and water on alternate days, and having ever larger and heavier stones placed on their chests until they either agreed to a jury or died. Because if one died without being convicted of a crime his or her property did not escheat to the Crown, many defendants chose to be pressed to death rather than risk a trial that could result in the impoverishment of their dependents (Baker 2002: 508–9). Second, contemporary psychological evidence casts doubt on the effectiveness of incarceration as a deterrent. Thus, although “[i]t has long been assumed that prison deters criminal offenders, . . . recent research in the forensic sciences has shown that, contrary to popular belief, prison may in fact increase the likelihood that offenders will engage in further crime after they are released” (Aharoni and Fridlund 2012: 599; Cid 2009; Grogger 1991). If in fact “incarceration is not just ineffective but even criminogenic” (Aharoni and Fridlund 2012: 599), then there is at least reason to doubt that it is necessary to preserve an orderly society. Once again, English legal history offers support for this possibility because imprisonment did not develop as a form of punishment under common law. Imprisonment was used to detain those awaiting trial, but was regarded as inappropriate as punishment because “[i]t would force the offender to be idle, making it difficult for him to pay his restitution, and it would be costly to the community” (Benson 1990: 71). In a time where most of life consisted in hard work, remaining idle in prison was not considered highly punitive. Hence, at common law “fines and whipping became the usual forms [of punishment], imprisonment being relatively uncommon for this purpose before the time of George III” (Baker 2002: 513). The assumption that criminal punishment is necessary to maintain an orderly society almost certainly derives from the worry that those without financial resources would not be deterred by the threat of civil liability—that in the absence of the threat of criminal punishment, “judgment-proof”

The Problem of Punishment  25 individuals would have no reason to refrain from criminal activity. However, this worry is based on the improper assumption that a system of civil liability that functions in a world without criminal punishment would be identical to our current system of civil liability, which functions alongside a system of criminal law and punishment. In our current dual system of criminal and civil liability, the threat of criminal punishment supplies the necessary element of deterrence. Hence, the system of civil liability need not and does not contain such an element. However, a system of civil liability that functioned without the deterrence provided by the threat of criminal punishment would be a very different system from the one that presently exists. Such a system would be more like the ancient system of tort law that the rise of criminal law displaced than the present one. The problem is that it is difficult for people to imagine what such a system would look like in contemporary circumstances. So here is where a bit of philosophical speculation can be of service.

3. Philosophy Philosophers love to engage in thought experiments. Apparently, I am no exception. So please indulge me by considering the following hypothetical legal system. Imagine a society with no criminal law and no criminal punishment. In order to not strain credulity too greatly, imagine that the police function precisely as they do at present.7 When citizens are threatened with harm, they may call on the police for protection, and if they have been the victim of a wrongful invasion of any of their legally protected interests, they may call on the police to apprehend those who have wronged them. However, in this society, citizens’ legally protected interests are defined by the rules of tort law as it evolved through the common law process. Again, in order not to strain credulity, let us assume that the substantive rules of tort law in the hypothetical society are identical to those of our current society. As in our current society, citizens would have to take out insurance policies to protect themselves against the possibility that they may be the victims of theft or personal injury by wrongdoers who are never apprehended. However, in cases in which the police do apprehend the wrongdoers, victims may bring private lawsuits against them to recover for the harm that they caused.8 The hypothetical system is similar to our current system of civil liability in that plaintiffs/victims are entitled to recover sufficient compensation to restore them to the position that they would have been in had the tortious conduct not occurred, plus, in the case of intentional wrongdoing, exemplary damages. However, in the hypothetical system, the amount of damages necessary to restore plaintiffs is understood to include both the costs that they bore in apprehending the defendant and the costs of bringing the lawsuit.

26  John Hasnas The hypothetical system is similar to our current system of civil liability in several other respects as well. Plaintiff’s attorneys can be paid on a contingency fee basis in which their compensation is contingent on the success of the lawsuit and is deducted from the judgment. This would ensure that injured parties with limited resources would nevertheless have access to the court system. Juries would be free to set the amount of exemplary damages at whatever level they deem appropriate and would be permitted to consider the defendant’s net worth in doing so. This would ensure that the system could discourage actions that the ordinary members of society find especially reprehensible as well as all intentional wrongdoing by wealthy defendants. Estates of those killed by others would be authorized to bring lawsuits for wrongful death,9 and class action lawsuits could be brought to discourage frauds and other wrongful actions that impose small injuries on large numbers of people. And finally, defendants who have the resources to pay their judgments must do so—either in a lump sum or in installments paid over time that include appropriate interest payments—and this debt cannot be discharged in bankruptcy. Successful plaintiffs are entitled to take out liens on defendants’ personal and real property and to garnish defendants’ wages to satisfy the judgment. The main difference between the hypothetical system and our current liability system concerns the way the hypothetical system treats defendants who do not have sufficient resources to pay their judgments—the so-called “judgment-proof” defendants. In the hypothetical system, successful plaintiffs are entitled to restrain the liberty of defendants in a limited number of legally authorized ways until the judgment has been paid. For example, plaintiffs may elect to sell their judgments to monitoring companies that take on the task of obtaining payment from the defendant. Thus, one who recovers a judgment of $100,000 against an indigent defendant might sell the right to collect the judgment to Acme Monitoring, Inc., for $65,000 in cash.10 Acme Monitoring would then have charge of the defendant and could ensure that he works sufficiently to earn $25,000 per year for four years, which it would collect. Monitoring companies would have the incentive to employ defendants in the most economically efficient way. Thus, those who are neither dangerous nor a flight risk can be lightly monitored and permitted to remain in society while working to pay off their judgment. Those who are a flight risk, but present no danger to others and have marketable skills might be monitored by an electronic ankle bracelet and employed by companies that value their skills. In contrast, violent defendants who pose a danger to others might have to be confined in the equivalent of private prisons and employed at whatever low-skilled labor they are capable of.11 Monitoring companies would release the defendants when they collected the full amount of the judgment. Further, such companies could be subject to whatever form of governmental regulation or inspection is required to guard against abuse of those in their charge.12

The Problem of Punishment  27 Of course, plaintiffs could elect to monitor the defendants themselves if they thought it worthwhile, but our hypothetical system gives them no right to inflict gratuitous harm on defendants and permits them to restrain their liberty only to the extent necessary to collect their judgment. However, plaintiffs would be free to bargain with defendants for alternative arrangements. For example, a vindictive plaintiff could offer to forgive all or a portion of the judgment if the defendant agreed to submit to some form of corporal punishment or humiliation, while a compassionate or merciful plaintiff could forgive all or a portion of the judgment if the defendant agreed to reform or do good works. In the hypothetical system, there is no state prosecution, no state-created criminal law, and no state-imposed punishment. To the extent that the state provides the courts, its role is that of a neutral referee in disputes between private parties. There are no punitive sentences, only restorative judgments; no state parole boards, only victims/plaintiffs with the discretion to release defendants from their debt obligations or not, as they see fit. The system does not create an incentive to vindictively wreak harm on the defendants, but rather to employ them as productively as possible so that they can pay the judgments; i.e., to rehabilitate them. In short, the hypothetical system is not a punitive system. Its purpose is restitution, not punishment. Would such a system deter harmful activity sufficiently to sustain a relatively peaceful society? It certainly would not provide the economically optimal amount of deterrence. There will be many instances in which people who engage in predatory conduct get away with wrongdoing under the hypothetical system. But that is irrelevant because the same is true under a state-run system of criminal punishment. The relevant question is how the hypothetical system compares with a system of criminal punishment. Using the level of deterrence provided by our contemporary system of criminal law and punishment as a baseline, the question is whether the hypothetical system would provide a similar level of deterrence. Perhaps a detailed empirical investigation would provide a definitive answer to this question. But once again, the limited format of this chapter does not permit such an investigation. Hence, I will do no more than make a few observations designed to show that it is not unreasonable to believe that the hypothetical system can do the job at least as well as a criminal punishment system. Skepticism about the efficacy of the hypothetical system can arise from the belief that the threat of financial liability is unlikely to discourage hardened criminals from engaging in serious wrongdoing such as murder, rape, assault, or theft. But one must keep in mind that we are making a comparative assessment, and that even under our present highly punitive system with its threat of imprisonment and death such individuals are often not deterred. Further, in the hypothetical system, plaintiffs victimized by such hardened criminals are likely to receive extraordinarily large exemplary

28  John Hasnas damage awards that would translate into the equivalent of long terms of confinement for defendants who have to work off the debt; and often into the analog of life sentences. In addition, the confinement permitted under the hypothetical system is less likely to be criminogenic because the only way to shorten one’s term of confinement is to become more productive. Skepticism may also arise from the belief that without state supplied public prosecutors, there will be no guarantee that a sufficient number of prosecutions will be brought to provide adequate deterrence. However, once again, one must remember that we are making a comparative assessment. Public prosecutors have limited budgets. This requires them to make decisions about which offenses to expend their limited resources on and which offenders to prosecute. Further, these decisions are often influenced by political considerations that can diverge from the goal of maintaining social order. For example, current efforts to satisfy the political demand for the suppression of drug use, prostitution, and illegal gambling consume such a large share of available prosecutorial resources that relatively small amounts are left for the prosecution of crimes of violence, and even less for nonviolent property offenses. In contrast, under the hypothetical system, there is no limited collective prosecutorial budget. Each party that has suffered a loss at the hands of an offender has the incentive to pursue that party for restitution. Indeed, there is historical evidence that supports the hypothetical system’s ability to deliver the needed deterrence. Private prosecution was the norm for most of English history and adequately supplied the needed deterrence until the late 18th century. With the advent of the industrial revolution, the increasing urbanization of English society produced an increase in criminal activity that strained the ability of crime victims to bring effective prosecutions as individuals. As a result, private prosecution associations formed to allow individuals to share the costs of bringing prosecutions. Beginning in the mid-eighteenth century, private associations and organizations began to be formed across the country. These associations emerged out of agreements neighbors made with one another to prosecute any crime committed against them. Over time, the agreements became more formal. Associations drew up constitutions, membership lists, and rules that committed them to fund prosecutions on behalf of their members. Members paid entry fees and annual subscriptions which were used to help pay for the expenses involved in apprehending and prosecuting a suspect. Most associations had their own solicitor who represented the association in court and provided legal advice to members. The funds of an association were also used to pay informers and to advertise rewards. (Koyama 2014: 287) Thus, even under highly adverse conditions, private parties were able to achieve sufficient economies of scale to supply the deterrence necessary to maintain an orderly society.

The Problem of Punishment  29 It is certainly the case that the hypothetical system will not deter any victimless crimes. There can be no civil lawsuit without a plaintiff, so there will be no deterrence of activities that do not harm others. But it is not clear that the suppression of voluntary drug use and consensual sexual or gambling activities is necessary to the maintenance of an orderly society. In fact, given the amount of violence engendered by the effort to supply black market goods, this may be a reason to believe that the hypothetical system is likely to be more successful than a system of criminal punishment in preserving a peaceful and orderly society.

V. Conclusion In this chapter, I have suggested that, in a liberal society, the problem of punishment is not identifying the purpose of criminal punishment, but determining whether criminal punishment is morally justified in the first place. I have not presented an argument sufficient to establish that criminal punishment is not morally justified, but I have argued that we should not simply assume that it is. If criminal punishment were necessary to maintain an orderly society, then criminal punishment could overcome the liberal presumption in favor of liberty. But whether criminal punishment is necessary to maintain an orderly society is an empirical question. In this chapter, I have introduced reasons to believe that criminal punishment may not be necessary to maintain an orderly society—that alternative mechanisms not involving state coercion can reduce interpersonal violence sufficiently to permit the growth of a prosperous, flourishing society. Of course, even if criminal punishment were not necessary to maintain an orderly society, that alone would not show that criminal punishment is not morally justified. Criminal punishment might be necessary to realize another moral value important enough to overcome the presumption of liberty. But, at least to me, it is not clear what this would be. A system of criminal punishment may be necessary to gratify injured parties’ desire for revenge, or to realize a particular set of moral ideals, or to effectively centralize state power. But these considerations can hardly serve as moral justifications for the use of state coercion inherent in criminal punishment. Because systems of civil liability are not punitive, they are not particularly effective at inflicting suffering on those found liable. Hence, they are not well-designed to wreak vengeance on wrongdoers. However, I do not believe inflicting harm on some to gratify the desire for revenge of others is a genuine moral value that can override the presumption of liberty. I would think that, rather than a defect, it is the glory of a liberal society that the state is not empowered to inflict harm on individuals merely to indulge its citizens’ baser emotions. Similarly, a system of criminal punishment may be necessary to realize the ideals of Christian (or any other system of) morality or to ensure that all parties receive their just deserts as determined by some particular moral

30  John Hasnas perspective. But as discussed in Part II above, a liberal society is one in which the government is prohibited from using the coercion inherent in criminal punishment to “legislate morality.” Finally, history suggests that criminal punishment is necessary for the state to effectively centralize its power over the citizenry. Systems of civil liability are poor mechanisms for social control. They cannot produce restrictions on “victimless crimes” or other measures designed to control citizens’ non-harmful behavior because without an injured plaintiff, no legal action would be brought, and hence, no legal rules would be generated. But the centralization of coercive power cannot be a moral value in itself, and in fact, it is precisely what a liberal society is designed to guard against. The fact that in the absence of a system of criminal punishment it would be more difficult for the state to effectively control the behavior of its citizens cannot count as a reason to have such a system in a liberal society. If criminal punishment cannot be morally justified by the need to realize some state-approved conception of the moral society, and if criminal punishment cannot be morally justified as necessary to maintain an orderly society, and if criminal punishment increases the risk that oppressive measures not necessary to the maintenance of an orderly society will be adopted, then it is likely that criminal punishment is not morally justified. And that is the problem of punishment.

Notes 1 J.D., Ph.D., LL.M.; Professor of Business and Professor of Law (by courtesy), Georgetown University. The author wishes to thank the magnanimous and wonderful people at Duke Law School for hosting him during the research for this article, the participants at the Against Incarceration conference sponsored by the University of New Orleans in September 2016, David Schmidtz, Michael Kates, Govind Persad, David Faraci, Alan Hasnas, and Ann C. Tunstall, for their insightful comments of a draft of this article, and Annette Hasnas of the New School of Northern Virginia and Ava Hasnas of the Oakwood School for giving him firsthand experience with the problem of punishment. 2 This is the doctrine of strict liability in tort. See American Law Institute (2010: §20). 3 See Olson (2006: 68) (“Whereas in the central Middle Ages crimes such as homicide, arson, and theft were private wrongs that entitled the victim or his kin to demand compensation or to engage in a blood feud, beginning in the twelfth century, serious wrongdoing came to be seen as against the entire community and subject to punishment by the lord or king who acted on the community’s behalf.”) 4 See Olson (2006: 73) (“dispute resolution—criminal or otherwise—is and ought to be focused upon concord and reconciliation.”) 5 Lyon further points out that “[i]n the growth of this list [of crimes] we may be certain that although the king’s concern for law and order was a cause, his chief interest was need of money; to increase his income the king only needed to use his prerogative and throw his jurisdiction over another offense” (1980: 189). 6 As Richard Laster explains, When the king or overlord began to take their share of the victim’s compensation, . . . the focus of the proceedings shifted from the victim to the

The Problem of Punishment  31 criminal. The next step was for the king to take the entire compensative payment and thus effectively destroy the process of community composition and raise punishment to the level of satisfaction (1970: 75–6). 7 This article is concerned exclusively with the justifiability of criminal punishment. It asks only whether there is a moral justification for the state to punish citizens who violate its criminal law. It is not concerned with whether the state is justified in supplying police services, which is an entirely different question. Therefore, for purposes of this article, one may assume that the state continues to supply the police who attempt to apprehend criminals. The disappearance of the criminal law does not entail the disappearance of government supplied police. 8 In this respect, our hypothetical legal system functions much like the actual English legal system, in which, until the late nineteenth century, criminal prosecutions were brought by private attorneys hired by the victims. It has long been a common practice in England for crime victims to initiate criminal prosecutions . . . England’s common law considered a crime to be a wrongful act against a particular identified victim, as opposed to a wrongful act against the state. The common belief at the time, therefore, was that those in the best position to prosecute criminal cases were the victims themselves (McCormack 2004: 499–500). 9 Just as automobile insurance policies include provisions to bring suit on behalf of the policy holder against negligent drivers, in the absence of public prosecution, life insurance policies would include provisions to bring wrongful death suits on behalf of the deceased. This would ensure that even those who kill isolated individuals without estates would be subject to liability. 10 There is nothing novel about this proposal. Many financial companies already purchase judgments from successful plaintiffs for lump sum payments. The most well-known of these may be J. G. Wentworth because of its televisions commercials; others such as Novation Settlement Solutions and Peachtree Financial Solutions do the same thing. 11 The discount paid to the plaintiff would reflect the monitoring company’s prospects of collecting the full amount of the judgment. 12 Currently, the federal government and several states employ private contractors to house their inmates in what are misleadingly called private prisons. These are not true private prisons that are subject to the discipline of market competition, but are “privatized” prisons—prisons that the state grants private contractors the exclusive right to operate at taxpayer expense subject to government regulation and oversight. Use of such privatized prisons, several of which have been criticized for abusive practices, is definitely not what is being advocated in this article.   For present purposes, the relevant comparison is between state imposed restrictions of liberty under the criminal justice system—the current system of state punishment, which includes both state operated and privatized prisons—with privately imposed restrictions of liberty under the hypothetical civil liability system—the hypothetical monitoring companies. As a rule, there is no reason to believe that state provision of a service provides a superior product or results in less abuse than private provision of that service subject to government regulation and inspection, as the comparison of public and private schools illustrates.   Indeed, in the context of incarceration, there is good reason to believe that precisely the opposite is the case. The state operated or privatized prisons are paid a fixed amount by taxpayers regardless of how they treat the inmates. The way to maximize profits under this incentive structure is to spend as little money and cut as many corners as possible. Further, state operated systems have a high

32  John Hasnas degree of immunity from tort suits brought by inmates alleging abuse by their employees.   In contrast, privately operated monitoring companies receive no taxpayer funding and can profit only by producing value in excess of their operating costs. This creates the incentive for them to increase the useful skills and productive capacities of their inmates. In addition, as private entities, they have no special immunity from lawsuits brought by those in their charge, and are subject to whatever anti-abuse regulations the government sees fit to impose on the industry.

References Aharoni, Eyal & Fridlund, Alan J. (2012) “Punishment without reason: Isolating retribution in lay punishment of criminal offenders,” Psychology, Public Policy, and Law, Vol 18. No. 4, pp. 599–625. American Law Institute (2010) Restatement of the Law (Third), Torts: Physical and Emotional Harm, Philadelphia, PA: American Law Institute Arthur, John & Shaw, William H. (eds.) (1993) Readings in the Philosophy of Law, 2nd ed., New York, NY: Routledge Adams, David M. (ed.) (1992) Philosophical Problems in the Law, Belmont, CA: Wadsworth Baker, John H. (2002) An Introduction to English Legal History, 4th ed., New York, NY: Oxford University Press Benson, Bruce L. (1990) The Enterprise of Law: Justice Without the State, San Francisco, CA: Pacific Research Institute Berman, Harald (1983) Law and Revolution, Cambridge, MA: Harvard University Press Chen, M. Keith & Shapiro, Jesse M. (2007) “Do Harsher Prison Conditions Reduce Recidivism? A Discontinuity-based Approach,” American Law and Economics Review, Vol. 9, No. 1, pp. 1-29 Cid, José (2009) “Is Imprisonment Criminogenic? A Comparative Study of Recidivism Rates between Prison and Suspended Prison Sanctions,” European Journal of Criminology, Vol. 6, No. 6, pp. 459–80 Delgado, Mauricio R.; Labouliere, Christa D. & Phelps, Elizabeth A. (2006) “Fear of Losing Money? Aversive Conditioning with Secondary Reinforcers,” Social Cognitive and Affective Neuroscience, Vol. 1, No. 3, pp. 250–59 Feinberg, Joel (1973) Social Philosophy, Englewood Cliffs, NJ: Prentice Hall Feinberg, Joel (1984) Moral Limits of the Criminal Law, Vol. 1: Harm to Others, New York, NY: Oxford University Press Feinberg, Joel & Gross, Hyman (eds.) (1995) Philosophy of Law, 5th ed., Belmont, CA: Wadsworth Gaus, Gerald F. & Courtland, Shane D. (2014) "Liberalism," The Stanford Encyclopedia of Philosophy, Zalta E. N. (ed.), from: http://plato.stanford.edu/archives/ spr2015/entries/liberalism Goebel, Julius (1937) Felony and Misdemeanor: A Study in the History of English Criminal Procedure, New York, NY: The Commonwealth Fund Golding, Martin P. (ed.) (1975) Philosophy of Law, Englewood Cliffs, NJ: Prentice Hall Grogger, Jeffrey (1991) “Certainty vs. Severity of Punishment,” Economic Inquiry, Vol. 29, No. 2, pp. 297-309

The Problem of Punishment  33 Husak, Douglas (1983) “The Presumption of Freedom,” Noûs, Vol. 17 No. 3, pp. 345–62 Kadish, Sanford H. & Paulson, Monrad G. (1969) Criminal Law and its Processes, 2nd ed., Boston, MA: Little Brown Kent, Edward A. (ed.) (1970) Law and Philosophy, Newmarket, ON, Canada: Pearson Education Khamsi, R. (2006) “Proof That Losing Money Really Is Scary,” New Scientist Daily News, from: https://www.newscientist.com/article/dn10324-proof-that-losingmoney-really-is-scary/ Klerman, Daniel (2002) “Settlement and the Decline of Private Prosecution in Thirteenth-Century England,” Law and History Review, Vol. 19, No. 1, pp. 1-65 Koyama, Mark (2014) “The Law and Economics of Private Prosecutions in Industrial Revolution England,” Public Choice, Vol. 159, No. 1, pp. 277-98 Laster, Richard E. (1970) “Criminal Restitution: A Survey of its Past History and an Analysis of its Present Usefulness,” University of Richmond Law Review, Vol. 5, No. 1, pp. 71-98 Lyon, Bruce (1980) A Constitutional and Legal History of Medieval England, New York, NY: W. W. Norton McCormack, Michael T. (2004) “The Need for Private Prosecutors: An Analysis of Massachusetts and New Hampshire Law,” Suffolk University Law Review, Vol. 37, No. 3, pp. 497-511 Mill, John Stuart (1975) On Liberty (D. Spitz, ed.), New York, NY: W. W. Norton (or. 1859) Moore, Michael S. (1997) Placing Blame: A General Theory of the Criminal Law, New York, NY: Oxford University Press Murphy, Jeffrie C. & Coleman, Jules L. (eds.) (1990) Philosophy of Law: An Introduction to Jurisprudence, 2nd ed., Boulder, CO: Westview Olson, Trisha (2006) “The Medieval Blood Sanction and the Divine Beneficence of Pain: 1100-1450,” Journal of Law and Religion, Vol. 22, No. 1, pp. 63–129 Schauer, Frederick & Sinnott-Armstrong, Walter (eds.) (1996) Philosophy of Law, New York, NY: Oxford University Press Seipp, David J. (1996) “The Distinction between Crime and Tort in the Early Common Law,” Boston University Law Review, Vol. 76, No. 1, pp. 59–88 Seymour, Ben; Daw, Nathaniel; Dayan, Peter; Singer, Tania & Dolan, Ray (2007) “Differential Encoding of Losses and Gains in the Human Striatum,” Journal of Neuroscience, Vol. 27, No. 18, pp. 4826–31 Wellcome Trust (2007), Why Losing Money May Be More Painful than You Think, from: http://www.eurekalert.org/pub_releases/2007-05/wt-wlm043007.php.

2 Unconscionable Punishment Michael Huemer

1.  Questions about Imprisonment With over two million prisoners—close to a quarter of the entire world’s prison population—America has the dubious distinction of being the world’s leader in imprisoning people. The nation’s incarceration rate has multiplied more than five-fold over the past four decades (Walmsley 2013; Sabol, Cooper, and West 2009; Wikipedia 2016).1 There are many reasons to be concerned about this. To begin with, there is the loss of freedom and the general suffering of the prisoners themselves. Of course, these things are part of the aim of the system; nevertheless, some question whether this is a morally permissible aim (Barnett 1977; Boonin 2008). There are concerns that incarceration renders criminals more dangerous than they were before they entered prison. This occurs because convicts acquire new criminal contacts while in prison, learn new criminal skills from the other prisoners, absorb antisocial values from the prisoner community, and acquire greater resentment toward society and the state due to their experiences in prison. This has led some experts to wonder whether the practice of incarceration actually causes more crime than it prevents (Pritikin 2008; Clear 2007). There is general agreement that incarceration does not rehabilitate criminals and that most will re-offend once released (Slifer 2014; Pritikin 2008: 1055, 1092–3). Incarceration also disrupts social relationships; for example, when a man is sent to prison, his children may in effect be left fatherless, and his spouse left to attempt to provide for them on her own (Clear 2007: 618–21). All of this is accomplished at a cost of 80 billion taxpayer dollars per year in the United States, or more than $30,000 per prisoner (Kyckelhahn 2015: table 1; Henrichson and Delaney 2012: 9). That is in addition to the economic costs of having more than two million people, mostly able-bodied men, removed from the labor market, some of whom would have held jobs had they not been behind bars. Of course, the system’s defenders would point to the potentially large benefits of incarceration: it directly prevents criminals from further victimizing members of the general (non-prison) population for the duration of their sentences. The well-known unpleasantness of imprisonment may deter

Unconscionable Punishment  35 potential criminals from turning to a life of crime, and deter actual criminals from continuing their life of crime. Finally, there is the alleged intrinsic justice achieved through the infliction of suffering on those who deserve to suffer because of their evil deeds. In light of these considerations, the ultimate balance of costs and benefits is not obvious. Perhaps the prison system’s benefits outweigh its costs. Nevertheless, the costs of the system are so serious that it is reasonable to wonder whether, as some maintain, a radically different approach might be better. Perhaps, for example, criminal justice ought to focus mainly on compulsory restitution—forcing offenders to make monetary compensation to their victims—rather than retributive punishment.2 Such an approach would seem to solve or greatly diminish most of the problems with imprisonment. Of course, this approach might have costs of its own, and it might be less effective at achieving the aims brought about by imprisonment. Nevertheless, on its face, it seems eminently worth exploring alternatives to incarceration. With that background in place, I want to raise the issue of a particular cost of the prison system that seems to be most often neglected in theoretical discussions of punishment. This is the widespread abuse that prisoners suffer, at the hands of guards and other prisoners, in most real-world prisons. The prevalence of this abuse renders the majority of prison sentences morally wrong, and speaks in favor of alternatives such as a restitution-based justice system.

2.  The Argument from Unconscionable Punishment In 2008, Kelly Bradley was serving a six-year prison sentence in Florida for burglary and grand theft.3 One morning in May, as Bradley (who suffers from schizophrenia) cowered in a corner of his cell, five officers in riot gear entered the cell and jumped on him. As the officers handcuffed Bradley, one of them dug into Bradley’s right eye socket with a finger and gouged out Bradley’s eye. The officers involved in the incident and their commanding officer attempted to cover up what happened, destroying physical evidence and claiming that they had no idea how Bradley’s eyeball had gotten out of its socket. The incident came to light only because another officer, John Pisciotta, reported the abuse. In the wake of the incident, exactly one officer was prosecuted. Of the six officers who participated in the abuse and cover-up (including the commanding officer), four were later promoted. Pisciotta, on the other hand, was fired. This is a particularly severe case of prison abuse, but not the most severe. In 2015, another inmate, Samuel Harrell (who suffered from bipolar disorder), was beaten to death by a group of as many as 20 guards in a New York prison (Winerip and Schwirtz 2015). In that case, the information about what took place came from other inmates who witnessed the beating.

36  Michael Huemer A year later, no one had been charged in Harrell’s death, and the guards responsible were still employed at the prison (Goodman 2016). How common are such incidents of prison abuse? No one knows, and some of the details of the above cases explain why: in many facilities, corrections officers have a culture of collaboration, such that officers are expected to stick together, support one another’s stories, and hide evidence of abuse. In the Bradley case, officer Pisciotta was subjected to severe retaliation for refusing to go along with the cover-up. Inmates often fear to report abuse by guards or other prisoners, since they live their day-to-day lives entirely at the mercy of the prison guards and surrounded by other prisoners. A 2013 report issued by the US government’s Bureau of Justice Statistics estimated that 4% of state and federal prison inmates and 3.2% of local jail inmates experienced some form of sexual abuse during the course of a year, for a total of about 81,000 incidents (Beck et al. 2013: 8). These estimates are based on anonymous surveys of prisoners; however, it is likely that incidents are dramatically underreported due to the shame felt by victims, as well as the “no snitching” culture in prisons. Over half of the incidents are perpetrated by prison staff, though violent incidents are much more likely to be perpetrated by other inmates. Prison rape has often been ignored by prison officials and the legal system. In some cases, rape is even used as a form of punishment by prison staff, or intentionally permitted by prison staff as a way to reward the rapist for cooperating with prison staff (Human Rights Watch 2001: 143–58; Parenti 1999; Harris 2001). Needless to say, prison rape is an extremely brutal crime, typically resulting in severe pain and injury in addition to the loss of human dignity. Violence and nonconsensual sexual contact are, uncontroversially, forms of abuse. Some experts also consider the increasingly widespread practice of solitary confinement to constitute a form of abuse, or even—as one U.N. observer put it—torture (U.N. News Centre 2011, quoting U.N. Special Rapporteur Juan Méndez). Studies show that solitary confinement causes serious psychological damage to inmates, as well as higher rates of recidivism after prisoners are released (Commission on Safety and Abuse in America’s Prisons 2006: 14–15). That is a very brief indication of the situation in America’s prisons. Now consider a hypothetical. Suppose that a criminal defendant has just been convicted of some felony—say, grand theft. And suppose that the trial judge sentences the defendant to be beaten and raped as punishment. Or to have his eyeball cut out. Or to be beaten to death by government agents. In any of these cases, the punishment would be unconscionable. No morally decent society would countenance such sentences—and not merely because they would be disproportionate for the crime in question; no one should receive such a sentence for any crime, not even if the offender himself committed similarly brutal acts. I am not going to argue for that claim. I assume as obvious that such punishments are unconscionable. That is why no judge in a modern society

Unconscionable Punishment  37 ever would assign such sentences. But here we have an ethical problem: given that it would be unconscionable to sentence a defendant to suffer these sorts of abuses, what is the moral status of handing out prison sentences to defendants, in the knowledge that many of them will in fact suffer prison abuse as a result? It seems to me that this must be considered highly morally problematic. Admittedly, the actual situation is less bad than my hypothetical, in at least two respects. First, in the hypothetical scenario, the judge intends for the convict to suffer unconscionable abuse, whereas in the actual world, judges are merely able to foresee that prisoners will in fact suffer abuse, as an unintended side effect of their confinement. The judge does not actually want the abuse to occur, and if a prisoner escapes abuse, the judge’s plans will not thereby be frustrated. This may be important since some philosophers hold that it is morally worse to bring about a harm intentionally than to merely bring it about knowingly (see Aquinas 1920: 2a2ae, Q. 64, art. 7; Anscombe 2001; Nagel 1986: 179–85). But while there may be a moral difference between intended and merely foreseen harm, and this difference may be important in some cases, it does not seem, in the case of prison abuse, that the difference between intended and foreseen harm is great enough to quiet our moral concerns. Imagine that, to save money, the state lays off all prison doctors and nurses, foreseeing that this will result in sick or injured inmates failing to receive medical care. This is not the state’s purpose in laying off the medical staff; the state’s purpose is merely to save money, and so the state would be perfectly happy if (miraculously) no inmates got sick or injured and thus no one went without needed medical care. The state merely foresees that this is highly unlikely, and so, as an unintended side effect of a money-saving measure, some inmates will be without needed medical care. This, I take it, would be unacceptable, and the fact that the harms to inmates would be merely foreseen rather than intended does not appear to make a great moral difference. Perhaps it would be even worse if the state actually intended inmates to lack needed medical care, but by no stretch is it permissible for the state to (merely) knowingly remove needed medical care. While individuals are in the custody of the state, the state has a moral responsibility to attend to their basic needs. Or imagine that prison officials assign a particular inmate to be confined in a cell with a known rapist. The rapist has been caught beating and raping other inmates many times in the past; his last cellmate, in fact, was recently taken out on a stretcher. The prison staff do not specifically intend for the new cellmate to be beaten and raped; they would be perfectly happy if, surprisingly enough, the rapist did not harm his new cellmate. Nevertheless, they know that this is extremely unlikely and that the rapist’s new cellmate will most likely be savagely beaten and raped. Again, while it would perhaps be even worse if the staff specifically intended for an inmate to be violently abused, by no stretch is the staff’s action in this case permissible. Prison staff

38  Michael Huemer are responsible for the physical safety of inmates in their custody and may not undertake actions that knowingly put inmates at grave risk of bodily injury, even if that injury is not their actual purpose. By the same token, then, it would seem impermissible on its face for a judge to sentence a defendant to prison, or for a legislature to prescribe prison time, in the knowledge that this will result in severe abuse of prisoners, even if this abuse is not their actual purpose. Here is another distinction: in my hypothetical scenario, the judge’s action is almost certain to cause the convict to suffer severe abuse, whereas in the actual world, each inmate only has a modest chance of suffering severe abuse. Surely the probabilities matter; a smaller probability of abuse makes the prison sentence less prima facie wrong, and less difficult to justify. Nevertheless, it seems as though even a significant risk of severe abuse makes a prison sentence seriously morally problematic on its face. Imagine a judge who sentences a group of twenty convicts to be thrown into the lion pit: a hungry lion is to be released, whereupon it will eat one of the convicts. Assuming that death by lion attack is not in general a morally permissible punishment, presumably this sentence—which imposes on each defendant a one in twenty chance of death by lion attack—is also morally unacceptable. Similarly, if prisoners have even a one in twenty chance per year of being beaten or raped in prison, it seems that this should normally be enough to make it wrong to send someone to prison. To summarize the argument thus far: 1. Severe prison abuse would be unconscionable as an official punishment. 2. If so, then it is also prima facie wrong to sentence individuals to live in conditions with a high, ongoing risk of severe abuse. 3. Most actual prisons have a high, ongoing risk of severe abuse for inmates. 4. So it is prima facie wrong to sentence individuals to be incarcerated, in most actual prisons. I do not know whether retributive punishment in general is ethically desirable. If it is, I see no in principle reason why long-term confinement could not sometimes be an acceptable form of punishment. But the actual conditions prevailing in the nation’s prisons seem sufficiently inhumane as to render confinement in them morally impermissible in most cases. (I shall return later to the “most cases” qualifier.) Now, I can imagine a defender of incarceration arguing that, in spite of the high risk of abuse, the state is nevertheless obligated to hold most criminals in custody, since the state has an obligation to ordinary citizens to protect them from crime, and if criminals are not held in custody, there is a high probability of their committing additional crimes. Indeed, one might claim that if the state knowingly releases a criminal, and the criminal commits another crime, the state itself will then be (partly) responsible for that

Unconscionable Punishment  39 crime. Granted, if the criminal suffers abuse in prison, the state might also be held (partly) responsible for that—but, one might think, since it is worse for innocent citizens to suffer from crimes than for convicted criminals to be abused, the correct tradeoff is to hold most criminals in prison. This argument would work best if we believed that prison rehabilitates criminals. If, as is actually the case, prison either has no effect or tends to make criminals more dangerous, then the argument works only if we are prepared to imprison criminals permanently. But to address the argument more directly, I want to explain why, perhaps seemingly perversely, it makes sense for the state to show greater concern about prison abuse than about ordinary crime. The reason is that the state is more responsible for an inmate’s suffering prison abuse than it is for an ordinary citizen’s suffering from a typical crime. Three important factors contribute to this responsibility. First, when an individual suffers prison abuse, the state has intentionally, coercively placed that person into an environment that was known to be extremely dangerous. Second, the state intentionally, coercively prevents prison inmates from undertaking most courses of action that individuals outside of prison may take to protect themselves—including relocating to safer areas, avoiding threatening people or situations, carrying weapons for self-defense, installing security systems, or even hiring security guards. Granted, the state has a general obligation to protect any citizen;4 but if the state takes away a citizen’s most effective (or even the only reasonably effective) self-protection options, then the state incurs a specially strong obligation to provide protection itself. Third and finally, much of the abuse that prison inmates suffer is actually carried out by agents of the state—specifically, correctional officers. None of these three points applies to the case of ordinary citizens suffering from ordinary crimes. Thus, the state has a specially strong obligation to avoid prison abuse, stronger than its obligation to counter ordinary crime.

3.  Prospects for Reform But why curtail the practice of imprisonment? Why not simply reform the prison system to eliminate abuse? Obviously, if we have the option of immediately eliminating abuse, we should take it. But the world is not so simple. Even if one thinks (dubiously) that maintaining imprisonment without abuse would be superior to finding alternative forms of punishment, it still makes sense to ask: taking the existence of widespread prison abuse as given, under what conditions should convicts be sentenced to prison? This is an important question to ask, for three reasons. First, there are systemic reasons why prisoners are prone to abuse; it is not a simple matter of a few bad prison wardens, say. Prisoners are in an inherently vulnerable position (this is compatible with the fact that they are also among society’s most dangerous members). In contrast to the normal

40  Michael Huemer situation on the outside, if a prisoner has a conflict with or feels threatened by another member of the prison community—whether a staff member or another inmate—the prisoner does not have the option of avoiding that person. He cannot at will decide to move out of his cell, or to be guarded by a different set of officers. The prisoner is completely dependent on prison staff in his day-to-day life and is open at any time to being physically coerced by them. Because of this, retaliation for reports of abuse is a particularly serious concern. In addition, the culture of prisons is extremely opposed to the reporting of unlawful conduct, which is referred to as “snitching”—and this culture is not something that the government is likely to be able to change. Even if a prisoner reports abuse, he may simply not be believed, due to the general (perhaps accurate) impression that convicts are untrustworthy as information sources. If the prisoner is believed, prison staff are likely to feel more loyalty to each other than to the inmates (or even to the law) and so may decide to cover up the abuse. If a prisoner appeals to people outside the prison system for help, this too is likely to fall on deaf ears. Convicts serving time in prison are, to put it lightly, among the least popular members of society; as a result, it tends to be difficult to mobilize concern about their plight. Finally, the prison population itself is inherently dangerous, since it consists almost entirely of criminals (apart from the prison staff and falsely convicted inmates), including some of society’s most violent individuals. These are all robust features of the prison system, not easily changeable. As long as these features of the system persist, it is hard to see how one can avoid high risks of prison abuse. Second, and related to the first point: even the most optimistic reformer can hope at best to reduce prison abuse; it will never be eliminated.5 As long as there remains a risk of abuse in prison, that risk will need to be taken into account when deciding whether a particular convict or class of convicts can justly be sentenced to prison. Third and finally, the individuals who make decisions as to whether to send criminal defendants to prison generally are not in a position to decide whether to reform the prison system to reduce abuse. For example, a trial judge must decide whether to sentence the defendant before him to prison time, given the existing state of the prison system; the judge does not have open to him at that moment the option of sending the defendant to prison and simultaneously radically reforming the prison system so as to drastically reduce the rate of abuse. Likewise, a jury in a criminal trial must decide whether to convict a particular defendant, knowing whatever they know about the likelihood that the defendant will receive jail time if convicted. The jury does not also have open to them the option of convicting the defendant and simultaneously drastically reforming the prison system. If the jury convicts, and the judge sentences the defendant to prison, the defendant will enter the prison system as it actually is. If prison reform is to occur, it would have to be undertaken by the legislature, but the judge and jury in a

Unconscionable Punishment  41 particular case, acting merely as such, have no control over the legislature, nor is it rational for them to wishfully assume that radical reform is on the horizon. Thus, knowing of the risk of severe prison abuse, juries at least have a reason to avoid convicting defendants who are likely to receive prison sentences, and judges have a reason to avoid prescribing jail time for defendants; this reason is not neutralized by the hypothetical possibility of a future society in which prison abuse is extremely rare.6 Legislators, too, must take the present, unconscionable prison system as given in certain contexts. When deciding, for example, whether a particular offense should carry a prison sentence as a penalty, lawmakers must take into account the current and likely future conditions in the nation’s prisons; they should not assume a scenario in which prisons are safe and humane. If reform eventually eliminates or drastically reduces prison abuse, it will be rational at that time to revisit the issue; until then, jurors, judges, and legislators have reason to be very reticent about sending people to prison.

4.  Punishment and Deterrence Without Incarceration What alternatives are there to incarceration? In most cases, the proper response to a crime would be to require the offender to make restitution to the victim, usually through monetary payments. Some offenders may (perhaps in addition) be required to perform community service, be placed on probation, be placed under house arrest, or be given other movement restrictions, with their movements being monitored by the state through devices such as ankle monitors.7 53% of all state and federal prisoners in the US are serving time for nonviolent offenses, chiefly property crimes and drug crimes, with drug offenses being the most common reason for admission to the prison system (Fellner 2006; Rothwell 2015).8 Compulsory restitution seems particularly appropriate for property crimes, while drug offenders should receive either no punishment at all (because drug prohibition should be repealed), compulsory drug treatment, or monetary fines.9 Alternatives such as these have obvious advantages over incarceration. First, since the state has proven unable or unwilling to protect convicts in its custody, these sorts of alternatives may be the only way to prevent prison abuse. Second, they tend to be much less expensive than incarceration, which costs an average of $31,000 per inmate per year (Henrichson and Delaney 2012: 9). Third, compulsory restitution is better and more just for crime victims, who usually receive no compensation at all in the current system. Fourth, these alternative approaches avoid placing criminals together where they can teach each other criminal skills, reinforce antisocial values, and form new criminal contacts for purposes of carrying out future crimes. Again, compulsory restitution should be the main response to crime. However, since most criminals are poor, most will be unable to immediately

42  Michael Huemer pay restitution. Offenders who cannot pay restitution immediately may have their wages garnished. Those who are unemployed should be required to take employment, either working for the state or working for private companies that agree to employ convicts. (For reasons of economic efficiency, the latter alternative is preferable, as long as there are sufficiently many businesses willing to employ convicts. In this case, the state might act as an agent matching convicts with employers. To maintain competition and prevent exploitation, convicts should be allowed to switch employers, but not allowed to cease working entirely.) Some believe that criminal behavior deserves to be punished, and thus that justice can only be served by making criminals suffer (Moore 1997; Lippke 2007). Admittedly, on this dimension incarceration does “better” than any of its likely alternatives; incarceration is a highly effective engine of suffering. One might therefore think that retributivist conceptions of justice favor incarceration over milder alternatives. Yet even staunch retributivists do not favor maximizing suffering; rather, retributivists favor punishment that is proportionate to the crime. Excessive punishment violates this imperative as much as insufficient punishment (Lippke 2003; 2007: Ch. 5). The inhumane conditions in the nation’s prisons raise a serious concern that most prison sentences constitute overpunishment. Compulsory restitution, by contrast, raises no such concern. In cases of compulsory restitution, the cost suffered by the criminal is, by definition, proportioned to the harm that the criminal caused to his victim. Retributive justice presumably does not require that wrongdoers suffer a specific type of cost, such as physical pain; there is no obvious reason why the demands of retributive justice cannot be satisfied through the offender’s suffering a monetary cost, which requires the offender, for example, to work more hours or forego benefits that he might have purchased with his money. I am not here taking a stand on retributivism; I do not know whether justice really requires that wrongdoers be made to suffer. What I am claiming is that a restitution-based system is superior either way: if justice demands that wrongdoers be made to suffer, a system of compulsory restitution can satisfy this demand through forcing offenders to suffer monetary costs proportional to the harm their evil deeds cause. This system carries a much lower risk of overpunishment and is thus superior with respect to justice. If, on the other hand, justice does not require that wrongdoers be made to suffer, compulsory restitution is still justified because it enables crime victims to receive what they are owed, and it is superior to incarceration since it also causes less suffering. Another goal often ascribed to the criminal justice system is that of deterrence: by conspicuously harming criminals, the system makes individuals fear to commit crimes. A system of compulsory restitution would provide significant deterrence, because the costs of paying restitution for a crime are typically many times greater than the benefits the criminal expected to receive by committing the crime. The reason for this is that the harm caused

Unconscionable Punishment  43 by a crime is typically many times greater than the benefit the criminal would receive if he got away with the crime. But one might worry that deterrence would be diminished if prison sentences were replaced with compulsory restitution, since prison is much more harmful and frightening than a mere monetary cost. I have two replies to this concern. The first is that, while the suffering of prison inmates might provide a deterrent to crime, especially in view of the severe abuses that prison inmates are vulnerable to, a just and civilized society cannot rely on the threat of severe overpunishment in this way. Imagine, for example, that tax evasion were punishable by public hanging. After a few tax evaders were publicly hanged, we could expect the rate of tax compliance to increase. But this simply is not something that a just or civilized society could do. It is wrong to overpunish people, and especially wrong to subject people to prison abuse, even if doing so frightens would-be criminals. Second, the empirical evidence for the deterrence advantages of imprisonment is weak. Studies have shown that for purposes of deterrence, the probability of apprehension is far more important than the severity of punishment (Nagin 2013: 201–2, 230–3). If potential criminals believe that they are unlikely to get caught, then severe penalties have little effect; if they believe they are almost certain to get caught, then small penalties suffice. Indeed, the probability of being apprehended by the police matters much more than the probability of being convicted if apprehended. Since most crimes are repeat offenses (Cohen and Kyckelhahn 2006: 2), the justice system should focus on increasing the probability of offenders being apprehended if they recidivate, rather than on maintaining or increasing the severity of sentences. For example, suppose that offenders were fitted with permanent ankle monitors and their whereabouts were tracked and recorded on an ongoing basis. This might deter criminal activity more effectively than a prison sentence.

5.  The Need for Incapacitation Until now, I have glossed over one important function of imprisonment that cannot be achieved by alternative sentences. This is the function of incapacitation: by holding criminals captive, the state directly, physically prevents them from committing crimes against the general population for the time that the criminals are imprisoned. There are some criminals who cannot be rehabilitated or deterred, and whose crimes are too terrible for society to risk their recurrence; in these cases, incapacitation is the only realistic approach. Consider, for example, the notorious serial murderer Ted Bundy, who killed over thirty women during the 1970s. He escaped from police custody twice in 1977 and killed at least three more victims before being recaptured for the last time in 1978. He was then held in prison until his ultimate execution in 1989. In his case, there was no alternative to incarceration or

44  Michael Huemer execution. That is an extreme case, but there are in fact many criminals who cannot ever be safely allowed out of custody, not even with a monitoring device, since their risk of killing or severely injuring others is too high. It is for this reason that I cannot agree with those who call for a complete end to prisons. For example, in 2015 the National Lawyers Guild passed a resolution calling for the abolition of all prisons (National Lawyers Guild 2015). They did not explain what would be done with serial killers such as Bundy. Without an answer to that question, abolitionism cannot be regarded as a serious proposal. Thus, my thesis in this chapter is only that most individuals who are actually imprisoned in the United States should not be imprisoned. The use of prisons should be dramatically reduced—but not eliminated. More specifically, all or nearly all nonviolent offenders should receive some alternative (non-custodial) sentence; nevertheless, the prison system must be maintained for the most dangerous, violent criminals. My aim in this section is to explain how this position—in particular, endorsing incarceration for the most dangerous criminals—can be squared with the Argument from Unconscionable Punishment (section 2, above). In discussions of justice, an absolutist stance is often tempting—for example, we may be tempted to insist that it is never permissible to punish an innocent person, no matter the consequences. Taken literally, this doctrine demands the abolition of all punishment, since any system that punishes anyone can be expected to sometimes (at least once!) punish an innocent person. Without going into detail here, I will simply state that I view such absolutist doctrines as untenable.10 For example, it must be permissible, for sufficiently weighty social goals, to incur some risk of punishing the innocent—which entails (in a large society) that it is permissible to maintain a policy or institution that punishes some number of innocent people. This does not mean that anything goes—the state must take reasonable measures to reduce the risk of punishing the innocent and may only punish defendants when the expected social benefits of punishment are very high in comparison to the risks of injustice. Thus, it is often said that it is better to let n guilty people go free than to punish one innocent person, for various values of n (for example, “it is better to let ten guilty men go free than to punish one innocent man”)—everyone seems to agree that the appropriate value of n is greater than one but less than infinity (Volokh 1997). Similarly, though prison abuse would be unconscionable as an official punishment, we cannot maintain the absolutist position that no level of risk of prison abuse is acceptable for the sake of any social benefit. The only reasonable position to take is that the state must make reasonable efforts to minimize the risk, and that the social benefits of imprisonment must be large in comparison with the risks of abuse incurred in maintaining the system. Of course, I am not going to provide a formula for making this evaluation, which is essentially a matter of judgment. I will, however, give three reasons why imprisonment is more morally acceptable in the case of serious, violent criminals than it is for less serious offenders.

Unconscionable Punishment  45 First, the inmates who are most often abused in prison are those who are relatively weak, young, and unaggressive. The most dangerous, violent criminals are the least likely to be abused; there is therefore less cause for concern in sending them to prison. Second, prison abuse is a less serious injustice the more serious the inmate’s own crime is. Of course, no inmate should suffer abuse; nevertheless, it is worse for a nonviolent criminal—say, one guilty only of drug possession—to be subjected to serious abuse than for a violent criminal—say, a murderer—to suffer such abuse. Therefore, the state has stronger reasons for eliminating incarceration for nonviolent offenders than for violent offenders. Third and most importantly, the risks to society of releasing violent offenders are much greater than the risks of releasing nonviolent offenders. Nonviolent criminals are generally guilty of either drug crimes or property crimes. If they recidivate, they are unlikely to cause irreparable damage. A violent criminal, however, may easily cause irreparable damage with a single crime—indeed, damage worse than the sum total of all harm that the criminal himself could expect to suffer from a lifetime in prison. There is one other category of criminal who may be justly imprisoned: those who refuse to make required restitution, or otherwise refuse to comply with court-ordered non-imprisonment sanctions. In this case, the state must either escalate sanctions or allow the criminal to get away with failure to comply. If the state’s policy is to give up on enforcement whenever an offender is sufficiently intransigent in refusing to comply, then compliance will drop dramatically, offenders will thumb their noses en masse at the criminal justice system, and the system will fairly collapse. The state must therefore be prepared to escalate—to increase penalties when a criminal fails to comply with earlier assigned penalties. This escalation may start with non-prison penalties (for example, late fees), but must at some point rise to the level of imprisonment if the offender is sufficiently recalcitrant. Fortunately, this is in practice a minor emendation to the general “no prison for nonviolent offenders” rule, because in point of empirical fact, the threat of prison is extremely effective at inducing compliance with non-prison penalties (Nagin 2013: 227-8). Thus, very few would wind up being imprisoned under this clause. In sum, then, the reasons for incarcerating serious violent criminals are much stronger, and the reasons against incarcerating them much weaker, than is the case for nonviolent criminals. It is therefore reasonable to hold that incarceration is justified only for serious, violent criminals and criminals who refuse to comply with non-prison penalties.

Notes 1 Hereinafter, I use “prison” and “imprisonment” broadly, to cover (detention in) jails, state and federal prisons, and other government detention facilities. 2 A note on “punishment”: some theorists argue that punishment, by definition, must have harm to the offender as a goal (Boonin 2008: 12–17); others reject this

46  Michael Huemer interpretation of the term (Cholbi 2010: 87–8). Herein, I use “punishment” in a broad sense—roughly, to include all costs imposed by the state on an offender for breaking the law. This is purely stipulative. I use “retributive punishment” to refer specifically to punishments motivated by the alleged intrinsic justice found in making those who have acted wrongly suffer. 3 Details of this case are taken from Brown 2015. 4 On this point, I disagree with conventional U.S. legal doctrine, which holds that the state has no general duty to protect individuals who are not in its custody; see Warren v. District of Columbia (444 A.2d. 1, D.C. Ct. of Ap., 1981). 5 In 2003, the U.S. Congress passed the Prison Rape Elimination Act by unanimous consent. Needless to say, however, the act has not come close to eliminating prison rape. Its main effect has been to facilitate better data collection on the problem; see Bruenig 2015. 6 Some hold that juries should decide whether to convict or acquit based solely on the factual evidence. I disagree: juries should decide whether to convict or acquit based upon their judgment as to which result will be more just overall; see Huemer 2013: 168–9; Huemer 2016. 7 For discussion of alternatives to incarceration, see United Nations Office on Drugs and Crime 2007; Families Against Mandatory Minimums 2013. On restitution in particular, see Barnett 1977; Boonin 2008: Ch. 5. 8 Classifications of offenders are based on the most serious offense for which the offender is imprisoned. Statistics for local jails were unavailable. Though 47% of the state and federal prison population were convicted of a violent crime, only a quarter of new admissions are violent criminals (violent criminals tend to receive longer sentences; thus, they occupy a higher portion of the existing prison population than of the new admittees). 9 Portugal has experimented successfully with drug decriminalization; see Hollersen 2013. The best approach would be to repeal the nation’s drug laws, obviating the need for any judicial response to drug users and sellers; see Huemer 2010a. 10 For detailed argument, see Huemer 2010b.

References Anscombe, Elizabeth (2001). “Medalist’s Address: Action, Intention and ‘Double Effect.’ ” In The Doctrine of Double Effect: Philosophers Debate a Controversial Moral Principle, ed. P.A. Woodward. Notre Dame, IN: University of Notre Dame Press. Aquinas, Thomas (1920). The Summa Theologica of St. Thomas Aquinas, 2nd and rev. ed., trans. Fathers of the English Dominican Province. London: Burns Oates & Washbourne Ltd. Barnett, Randy (1977). “Restitution: A New Paradigm of Criminal Justice.” Ethics 87: 279–301. Beck, Allen J., Berzofsky, Marcus, Caspar, Rachel and Krebs, Christopher (2013). “Sexual Victimization in Prisons and Jails Reported by Inmates, 2011–12.” U.S. Department of Justice, Bureau of Justice Statistics. Available at www.bjs.gov/ content/pub/pdf/svpjri1112.pdf, accessed May 17, 2016. Boonin, David (2008). The Problem of Punishment. Cambridge: Cambridge University Press. Brown, Julie K. (2015). “Whistleblower Prison Guard Paid the Price for Reporting Abuse.” Miami Herald, July 18. Available at www.miamiherald.com/news/ special-reports/florida-prisons/article27738046.html, accessed May 17, 2016.

Unconscionable Punishment  47 Bruenig, Elizabeth Stoker (2015). “Why Americans Don’t Care About Prison Rape.” The Nation, March 2. Available at www.thenation.com/article/why-americansdon’t-care-about-prison-rape/, accessed May 18, 2016. Cholbi, Michael (2010). “Compulsory Victim Restitution Is Punishment: A Reply to Boonin.” Public Reason 2: 85–93. Clear, Todd R. (2007). “The Impacts of Incarceration on Public Safety.” Social Research 74: 613–30. Cohen, Thomas and Kyckelhahn, Tracey (2010). “Felony Defendants in Large Urban Counties, 2006.” U.S. Department of Justice, Bureau of Justice Statistics. Available at www.bjs.gov/index.cfm?ty=pbse&sid=27, accessed May 20, 2016. Commission on Safety and Abuse in America’s Prisons (2006). Confronting Confinement. New York: Vera Institute of Justice. Available at www.vera.org/pubs/ confronting-confinement, accessed May 17, 2016. Families Against Mandatory Minimums (2013). “Alternatives to Incarceration in a Nutshell.” Available at famm.org/wp-content/uploads/2013/08/FS-Alternativesin-a-Nutshell-7.8.pdf, accessed May 22, 2016. Fellner, Jamie (2006). “US Addiction to Incarceration Puts 2.3 Million in Prison.” Human Rights Watch. Available at www.hrw.org/print/230976, accessed May 17,  2016. Goodman, Amy (2016). “Family of Samuel Harrell, Killed by ‘Beat Up Squad’ in NY Prison, Holds Hunger Strike for Justice.” Democracy Now. Available at www.democracynow.org/2016/4/26/family_of_samuel_harrell_killed_by, accessed May 17, 2016. Harris, Dan (2001). “Prison Rape Widely Ignored by Authorities.” ABC News, April 16. Available at http://abcnews.go.com/WNT/story?id=131113, accessed May 17, 2016. Henrichson, Christian and Delaney, Ruth (2012). The Price of Prisons: What Incarceration Costs Taxpayers. New York: Vera Institute of Justice. Available at www. vera.org/pubs/special/price-prisons-what-incarceration-costs-taxpayers, accessed May 19, 2016. Hollersen, Wiebke (2013). “Portugal, 12 Years After Decriminalizing Drugs.” Der Spiegel, March 27. Available at www.spiegel.de/international/europe/ evaluating-drug-decriminalization-in-portugal-12-years-later-a-891060.html, accessed May 22, 2016. Huemer, Michael (2010a). “America’s Unjust Drug War.” In The Ethical Life, ed. Russ Shafer-Landau. New York: Oxford University Press. Huemer, Michael (2010b). “Lexical Priority and the Problem of Risk.” Pacific Philosophical Quarterly 91: 332–51. Huemer, Michael (2013). The Problem of Political Authority. New York: Palgrave Macmillan. Huemer, Michael (2016). “The Duty to Disregard the Law,” Criminal Law and Philosophy 9: 423–42. Human Rights Watch (2001). No Escape: Male Rape in U.S. Prisons. New York: Human Rights Watch. Available at www.hrw.org/reports/2001/prison/, accessed May 17, 2016. Kyckelhahn, Tracey (2015). “Justice Expenditure and Employment Extracts, 2012— Preliminary.” U.S. Department of Justice, Bureau of Justice Statistics. Available at www.bjs.gov/index.cfm?ty=pbdetail&iid=5239, accessed May 22, 2016.

48  Michael Huemer Lippke, Richard L. (2003). “Retribution and Incarceration.” Public Affairs Quarterly 17: 29–48. Lippke, Richard L. (2007). Rethinking Imprisonment. Oxford: Oxford University Press. Moore, Michael S. (1997). Placing Blame: A Theory of the Criminal Law. Oxford: Oxford University Press. Nagel, Thomas (1986). The View from Nowhere. Oxford: Oxford University Press. Nagin, Daniel S. (2013). “Deterrence in the Twenty-First Century.” Crime and Justice 42: 199–263. National Lawyers Guild (2015). “Resolution Supporting the Abolition of Prisons.” Available at www.nlg.org/sites/default/files/Resolution Supporting the Abolition of Prisons REV.pdf, accessed May 22, 2016. Parenti, Christian (1999). “Rape as a Disciplinary Tactic.” Salon, August 23. Available at www.salon.com/1999/08/23/prisons_3/, accessed May 17, 2016. Pritikin, Martin H. (2008). “Is Prison Increasing Crime?” Wisconsin Law Review 2008: 1049–108. Rothwell, Jonathan (2015). “Drug Offenders in American Prisons: The Critical Distinction Between Stock and Flow.” The Brookings Institution, November 25. Available at www.brookings.edu/blogs/social-mobility-memos/posts/2015/11/ 25-drug-offenders-stock-flow-prisons-rothwell, accessed May 19, 2016. Sabol, William J., Cooper, Matthew and West, Heather C. (2009). “Prisoners in 2008.” U.S. Department of Justice, Bureau of Justice Statistics. Available at www. bjs.gov/content/pub/pdf/p08.pdf, accessed May 21, 2016. Slifer, Stephanie (2014). “Once a Criminal, Always a Criminal?” CBS News, April 23. Available at www.cbsnews.com/news/once-a-criminal-always-a-criminal/, accessed May 22, 2016. U.N. News Centre (2011). “Solitary Confinement Should Be Banned in Most Cases, UN Expert Says.” Available at www.un.org/apps/news/story.asp?NewsID=40097, accessed May 17, 2016. United Nations Office on Drugs and Crime (2007). Handbook of Basic Principles and Promising Practices on Alternatives to Imprisonment. Vienna: UNODC. Available at www.unodc.org/pdf/criminal_justice/Handbook_of_Basic_Principles_and_Pro mising_Practices_on_Alternatives_to_Imprisonment.pdf, accessed May 20, 2016. Volokh, Alexander (1997). “n Guilty Men.” University of Pennsylvania Law Review 146: 173–216. Walmsley, Roy (2013). World Prison Population List, tenth edition. International Centre for Prison Studies. Available at www.prisonstudies.org/sites/default/files/ resources/downloads/wppl_10.pdf, accessed May 21, 2016. Wikipedia (2016). “United States Incarceration Rate.” Available at https:// en.wikipedia.org/wiki/United_States_incarceration_rate, accessed May 21, 2016. Winerip, Michael and Schwirtz, Michael (2015). “Prison Guard ‘Beat Up Squad’ Is Blamed in New York Inmate’s Death.” New York Times, August 18. Available at www.nytimes.com/2015/08/19/nyregion/fishkill-prison-inmate-died-after-fightwith-officers-records-show.html, accessed May 17, 2016.

3 The Coproduction of Justice Nathan Goodman

The criminal justice system is meant to produce law, order, and public safety. If police and other public employees successfully reduce crime rates and help citizens feel safe, then we might say that they have successfully done their jobs. But these public employees cannot lower crime rates alone. Rather than passively consuming the safety that police hopefully create, citizens either help or hinder the production of that safety. They can lock their doors and install burglar alarms to protect themselves from crime. They can call the police, report crimes, testify in court, serve on juries, and leave crime scenes untouched, therefore helping police enforce laws. They might also impede the police by refusing to call, refusing to testify, or tampering with evidence. Elinor Ostrom and her colleagues coined the term coproduction to describe how citizens contribute to the production of public services they consume, such as police services (Parks et al. 1981).1 Ostrom and her team visited numerous police departments, rode alongside police officers, and conducted in-depth case studies to understand how police organization impacts police performance. Along the way, they learned that coproduction by citizens plays a key role in the success or failure of police. Coproduction enables police departments to incorporate the local knowledge held by citizens, and thus improve their service provision. Some police tactics, such as having neighborhood beat cops who walk the streets and build trust with the community, tend to promote coproduction. Others, such as police militarization, can undermine coproduction. Ostrom and her colleagues found that smaller departments tended to be better at enabling coproduction, while larger and more centralized police departments, while better suited to some tasks, tended to be distant and unresponsive to citizens, thus weakening the coproduction process. Ostrom’s research on coproduction can help illuminate the contemporary phenomenon of mass incarceration. In recent decades, America’s prison population grew rapidly, eventually becoming the largest in the world. Between prisons, jails, and immigration detention centers, approximately 2.3 million people are incarcerated (Wagner and Rabuy 2016). Members of minority groups, particularly poor black men, face especially high incarceration rates

50  Nathan Goodman (Western and Pettit 2004; Alexander 2010). Coproduction offers an analytical lens for exploring possible criminogenic consequences of mass incarceration. While incarceration is intended to reduce crime, some scholars have highlighted how incarceration can have unintended consequences that promote crime (Camp and Gaes 2005; Gaes and Camp 2009; Bench and Allen 2003; Chen and Shapiro 2007; Cid 2009). If mass incarceration impedes the coproduction process, this may create criminogenic consequences. How does mass incarceration impact the coproduction process? Mass incarceration may create a scenario where some individuals feel that they cannot trust the criminal justice system. When marginalized communities are torn apart by mass incarceration, residents may see police as adversaries rather than protectors. As a result, they may adopt norms that discourage calling the police, reporting crimes, or testifying in court. Such anti-snitching norms may impede citizen coproduction. While this empirical claim is difficult to assess, this chapter examines the literature on coproduction, mass incarceration, and anti-snitching norms to establish groundwork for future research on such claims. If mass incarceration impedes coproduction, then it may be counterproductive to its stated aims of reducing crime and improving public safety. The costs of reduced coproduction should be understood and weighed against the deterrence and incapacitation benefits that mass incarceration produces.

Ostrom’s Police Studies In the 1960s and ’70s, Elinor and Vincent Ostrom became involved in debates about metropolitan governments. Many scholars and reformers favored consolidating power and believed that overlapping jurisdictions reflected inefficient and redundant government. This view was supported by theorists who advocated what Vincent Ostrom (1973) termed “bureaucratic administration,” in which the focus is ensuring that a single bureaucratic hierarchy operates smoothly and efficiently. In contrast, the Ostroms favored a theory of “democratic administration” which drew from the work of such classical liberal theorists as Tocqueville, Madison, and Hamilton (ibid.). Rather than assuming that overlapping jurisdictions promote redundancy and inefficiency, the Ostroms argued that they introduce competition, allow power to check power, and create opportunities to incorporate local knowledge and preferences. They modeled metropolitan governance as polycentric, meaning it features “many centers of decision-making that are formally independent of each other” (V. Ostrom, Tiebout, and Warren 1961). Polycentricity can provide an institutional framework for self-governance. This does not necessarily require deferring to local governments. The federal government may be one center of decision making within a polycentric system, and thus produces important checks on local abuses. Federal courts hold local police accountable when they violate civil rights, for example.

The Coproduction of Justice  51 The Ostroms did not simply assert that polycentric governance works better. Rather, Elinor Ostrom and her colleagues empirically studied how consolidated government services compared to more polycentric public services, particularly police services. Ostrom rode in patrol cars with police, surveyed community residents, and generally conducted in-depth fieldwork to understand how policing operated in different jurisdictions and how satisfied community members were with the outcomes of policing. Using fieldwork and case studies allowed Ostrom to understand the specifics of how policing operated on the ground and how it impacted communities, rather than simply learning about aggregate variables such as crime rates that are often produced by police departments themselves (Boettke, Palagashvili, and Lemke 2013). Ostrom and her colleagues employed “natural experiments” to compare service satisfaction in areas that had consolidated to satisfaction in those that had not. One of their first studies compared regions in Indianapolis that had consolidated with regions that maintained their local jurisdictions. Community members were more satisfied with locally provided police services than those in consolidated jurisdictions (Ostrom et al. 1973a, 1973b). Comparative fieldwork in St. Louis, Chicago, Grand Rapids, and Nashville-Davidson County further supported the claim that locally controlled police produced better outcomes and higher citizen satisfaction than consolidated jurisdictions (Ishak 1972; Ostrom and Whitaker 1974; Parks 1979; Rogers and Lipsey 1974). Local police departments functioned better than large and consolidated departments because they better enabled coproduction. The officers themselves were not the only ones producing output and outcomes for the department. Citizens played an active role by engaging in acts of self-help, cooperating with police in investigations, and sharing local knowledge with police departments. Local departments were better equipped to build strong connections and trust with the community, providing an environment that encouraged coproduction. Polycentricity enables coproduction because it allows for self-governance. As Boettke, Palagashvili, and Lemke (2013) explain, “the concept of coproduction is linked to the capability of citizens to engage in self-governance and as such the art and science of association.” Police brutality and mass incarceration disproportionately impact African American communities. Elinor Ostrom and Gordon Whitaker (1974) recognized that police brutality had severely undermined trust in police and were inspired to study community control of policing in several black neighborhoods in Chicago as a solution for police brutality, corruption, and racism. They contrasted community control with police professionalization, which they argued “provides no leverage for blacks to demand improved service” (Ostrom and Whitaker 1974: 209). While the specific conditions of racial inequality in the criminal justice system have changed since the 1970s, Ostrom’s research remains relevant for understanding the harm done by mass incarceration and police brutality, as well as the possibilities for self-governing alternatives to America’s prison state.

52  Nathan Goodman In this chapter, I apply Ostrom’s findings and framework to contemporary problems in criminal justice. I begin by discussing anti-snitching norms, their possible causes, and the determinants of crime reporting rates over time. In the following section I discuss why community policing is not a panacea for encouraging coproduction and community trust, and explore the institutional factors that have led to dysfunction in community policing programs. The next section explores how entrepreneurship can fill gaps left by the criminal justice system. After that I discuss customary law among the Anglo-Saxons as an example of a justice system that provides strong incentives for coproduction, and explore how top-down interventions led to the breakdown of those incentives. I conclude the chapter by discussing avenues for future research applying Elinor Ostrom’s work to problems of criminal justice and coproduction.

Anti-snitching Norms The slogan “Stop Snitching” can be seen emblazoned on t-shirts, hats, and hoodies, particularly in urban African American communities. An admonition against potential informants, the phrase urges participants in criminal enterprises and witnesses more generally to refuse to cooperate with police. Popularized by a DVD of the same name, “Stop Snitching” promotes anti-snitching norms that directly impede coproduction. Anti-snitching norms largely originate as a way to resolve prisoner’s dilemma situations. In the classic game theory example of the prisoner’s dilemma, a prosecutor isolates two suspects. He already has sufficient evidence to convict each suspect on a minor offense, but the prosecutor will be able to convict the suspects on a more serious offense if they talk. By offering each suspect a deal for informing on his partner, the prosecutor creates an incentive to talk. While both suspects would be better off if neither talked, each one would individually be better off informing on their partner. If both partners are silent, both can go free, and if one partner informs, he can avoid the particularly harsh sentence the prosecutor threatens to impose for remaining silent while the partner talks. Criminals often resolve this dilemma with anti-snitching norms. These norms are enforced in various ways. Sometimes the enforcement comes from credible threats of violence against snitches and informants. “Snitches,” it is said, “get stitches.” Snitching also harms one’s reputation, so snitches lose opportunities to engage in illicit transactions even if they do not face assault. This imposes costs that reduce the payoffs of defecting and becoming an informant. The payoffs change, transforming a prisoner’s dilemma game into a different game structure. While threats play a role in enforcement, some of the enforcement may come from the norms and consciences of criminals. Criminals feel loyalty to their associates and have an antagonistic relationship with police, prosecutors, and prison guards. Snitching is a betrayal of that loyalty. Anti-snitching

The Coproduction of Justice  53 norms allow criminals to cooperate with each other without the full risk of exogenous punishment. While anti-snitching norms emerge to prevent criminals from snitching on their co-conspirators, they have spread beyond organized crime to deter witnesses rather than co-conspirators from reporting crimes and cooperating with investigations. For example, Israel “Izzy” Ramirez, a bodyguard for rapper Trevor Smith, better known as Busta Rhymes, was shot and killed in 2006. While an estimated seventy-five people, including Smith, witnessed the shooting, these witnesses refused to cooperate with police (Asbury 2011). This blanket refusal to cooperate with police reflects the spread of anti-snitching norms from deterring snitching by co-conspirators to deterring cooperation by witnesses from the general public. Similar incidents have been reported in Baltimore, Dallas, Philadelphia, and Washington, DC (Masten 2009). What accounts for this spread of anti-snitching norms? Some can be explained by condemnations of snitching in popular culture. But much of it can be explained by antagonistic relations between law enforcement and marginalized communities. Masten (2009) argues that anti-snitching norms partially arise from the current state of Fourth Amendment law. The erosion of Fourth Amendment protections enables police to deceive suspects in the lead up to searches, undermining trust in police. Similarly, Asbury (2011) argues that mass incarceration is responsible for the rise of anti-snitching norms. While incarceration is primarily meant to impose costs upon offenders, within poor black neighborhoods men are incarcerated at high rates that impose collateral costs on other community members. This creates a situation where loyalty to one’s community conflicts with loyalty to the state and the criminal justice system. While in middle and upper class white communities calling the police may signal loyalty to one’s community, in the marginalized communities where anti-snitching norms are most prevalent calling the police may signal disloyalty to the community. Community loyalty contributes to anti-snitching norms in other contexts. For example, Asbury points out that loyalty within police forces creates norms of silence, where police officers refuse to testify against one another or report misconduct. Similar dynamics operate in the military. In impoverished urban black communities that have been ripped apart by mass incarceration, Asbury argues that community loyalty may create anti-snitching norms. Rather than arising from community loyalty, anti-snitching norms may emerge from the broader influence of the underground economy. The underground economy is broader than goods and services that are strictly illegal, and many people work off the books in occupations that would be legal if they complied with taxes and regulations. Venkatesh (2006) documents how these underground economies operate in Chicago’s Maquis Park neighborhood, and how participants often find themselves connected to criminal syndicates even if they resent their impact on the community. This may both

54  Nathan Goodman deter interaction with police and allow gangs and drug syndicates to shape social norms. Mass incarceration may contribute to anti-snitching norms by helping organized crime grow. David Skarbek (2014) argues that prison gangs emerge and gain power in order to govern black markets within prisons. When prison populations are small, individuals can rely on reputation to govern their transactions. If one prisoner defrauds another, their fraudulent behavior may become known by the whole prison, thus preventing them from engaging in future transactions. However, when prison populations become larger, these reputation mechanisms break down because it is impossible for any individual inmate to know all other inmates. Prison gangs arise to provide governance where reputation can no longer provide it. Skarbek documents a rise in the prevalence, number, and power of prison gangs over the same time period when prison populations increased. He explains the mechanisms gangs use to govern inside prisons, and demonstrates how they extend their power to govern gangs outside of prisons. As anti-snitching norms are cultivated by organized crime in order to resolve prisoner’s dilemma problems, public policies that increase the power and reach of organized crime may contribute to the rise of anti-snitching norms. However, the contribution of mass incarceration to anti-snitching norms should not be overstated. While prosecutors and police have been frustrated by uncooperative witnesses influenced by anti-snitching norms, citizens overall have been reporting crimes at higher rates over the past few decades, a period when incarceration rates have increased as well (Cook and MacDonald 2011; Baumer and Lauritsen 2010). The fact that a rise in citizen crime reporting accompanied the rise of mass incarceration provides good reasons to doubt that mass incarceration undermines reporting across the board. This finding holds across demographic groups, with Baumer and Lauritsen finding “a significant increase in police notification since the mid-1980s for crimes of violence against women and men and violence experienced by Latinos, non-Latino Whites, and non-Latino Blacks” (ibid: 165). Moreover, African Americans have largely supported tough on crime policies (Forman 2012; Miller 2015), suggesting that Asbury’s community loyalty explanation may be less accurate than explanations rooted in the underground economy. Rather than deciding based on loyalty, individuals may weigh the benefits they receive from seeing criminals punished against the costs they face if they talk to police. These costs include risks of gang retaliation, further loss of men in the community, and an increase in their individual risk of incarceration if they face police scrutiny or if anti-snitching norms break down. Many factors influence crime victims when they decide whether to report crimes. It is possible that mass incarceration has contributed to anti-snitching norms that undermine reporting, but that the impacts of these norms have been outweighed by other changes over the past several decades. Some policy changes were implemented in the past several decades explicitly to

The Coproduction of Justice  55 encourage coproduction. Many police departments adopted community policing strategies largely to bolster community trust. Federal policies such as the Victims of Crime Act of 1984 and the Crime Victims’ Rights Act of 2004 may have also improved accessibility for crime victims, in the process encouraging them to report crimes. Technological innovations, including the proliferation of cell phones, have made it easier to call the police. Various cultural, social, and political reform efforts have also been made to encourage reporting of specific crimes such as rape, sexual assault, and domestic violence. Soares (2004b) finds that citizens report crimes at higher rates in developed countries. It is possible that economic growth contributed to higher reporting rates. Soares (2004a: 855) also finds that “crime reporting is strongly related to institutional stability, police presence, and perceived corruption.” These elements of institutional quality may also impact crime reporting within the United States. Further research is needed in order to disentangle possible causes and determine whether mass incarceration has any adverse impacts on the likelihood of reporting crimes to police.

Community Policing Is No Panacea Community policing programs are designed to enable citizen coproduction, and the coproduction literature can be interpreted as an argument for community policing. But community policing programs have often failed to improve relationships and trust between police and the community. In many cases, community policing programs are community based in name only and simply put a human face on aggressive policing practices that alienate community members. Boettke, Lemke, and Palagashvili (2016) argue that these problems largely result from community policing programs being implemented within a “nested game” where local police respond to incentives created by federal grants. This leaves local police less accountable to their communities, as they seek to please federal officials rather than locals who might otherwise reduce police revenue by voting with their feet. The federal interventions that have distorted so-called community policing programs are also an important driving force for mass incarceration. Growing federal involvement in criminal law, beginning in the 1950s and 1960s, enabled legislators to create new crimes and punishments. Indeterminate sentencing, which allowed judges and juries discretion in determining sentences, was eventually replaced with legislative decisions about sentencing policy. The right supported this change as a means of avoiding leniency while the left supported this change in order reduce racially biased uses of judicial discretion. The result was a series of mandatory minimum sentences that created grossly disproportionate punishments (National Research Council 2014). Politicians on both sides of the aisle faced incentives to pass new criminal laws in order to prove their toughness and appeal to a variety of special interest groups (Larkin 2013). Indeed, “[t]he two parties embarked on periodic ‘bidding wars’ to ratchet up penalties for drugs and

56  Nathan Goodman other offenses” (National Research Council 2014). The end of indeterminate sentencing ultimately displaced discretion. Rather than resting in the hands of judges and juries, discretion now rested in the hands of prosecutors, who could stack charges in order to threaten very harsh sentences and secure plea bargains. This undermined due process and fueled the growth of America’s prison population (National Research Council 2014). The distortion of community policing programs by federal incentives reflects a broader trend of well-intentioned criminal justice reforms backfiring. Critiques of racial bias in judicial decision making undermined indeterminate sentencing and laid the groundwork for mandatory minimum sentencing (National Research Council 2014). Prisoners’ rights litigation in Florida helped create political conditions that contributed to a dramatic expansion of Florida’s prison population (Schoenfeld 2010). Criminal justice policy, like all policies, is made within complex and unpredictable social systems, and is prone to unintended consequences. After policies produce unintended consequences, how does the system adjust to ameliorate those consequences? Is there a feedback mechanism that creates knowledge of the unintended consequences and incentives to alleviate them? Centralized political systems typically lack such feedback mechanisms. Within free markets feedback is provided by prices and profit and loss. If a firm’s actions do not successfully create value for customers, then their profits will decline, providing an incentive to adjust their behavior. Prices provide individuals and firms with information about the relative scarcity of resources and their most valued uses. The result is a system where dispersed knowledge is utilized, creating a tendency to move resources towards their most valued uses, create value for others, and adjust away from actions that destroy value (Hayek 1945; Lavoie 2015; Mises 1920). The political process has no analogous feedback mechanism. Politicians and bureaucrats are not subject to the discipline of prices or profit and loss. When their actions create unintended consequences, they also typically create interest groups with incentives to lobby against adjustments. For example, the reforms that led to mass incarceration have empowered prosecutors, created jobs for prison guards, and created profit opportunities for firms that operate prisons or provide services within prisons. Similarly, federal interventions that limited the effectiveness of community policing programs also provided resources to police, who generally will be averse to seeing those resources taken away. There is no comparable concentrated interest group opposed to harsh punishments and federal funding of police. The lack of effective feedback mechanisms prevents decision makers from incorporating the relevant knowledge, creating a systemic tendency towards disproportionate punishment (D’Amico 2015). Some feedback mechanisms can exist at the level of local government. The voice of citizens is more powerful on a local level, because the voting population is smaller. Moreover, citizens can vote with their feet by exiting smaller jurisdictions, and this Tiebout migration can decrease the tax base

The Coproduction of Justice  57 of localities that create undesirable policies (Tiebout 1956). This can lead to a competitive or market-preserving federalism that deters rent-seeking and predatory policies (Buchanan 1996; Weingast 1995). Similar exit options can create beneficial incentives in all voluntary associations, though governments and non-profit associations lack the feedback created by prices, profit, and loss. The ability to exit means that local governments, particularly if they do not have access to subsidies by more centralized governments, are likely to be more responsive to citizens. The same applies for voluntary associations, which means that coproduction by non-state actors is likely to be more responsive to the needs of individuals seeking protection from crime. Rather than implementing policies that use the language of community policing, effective criminal justice reform requires incentive-compatible policies that are responsive to the unique needs of diverse citizens and provide incentives for decision makers to correct their errors. Eliminating federal programs that subsidize local police would create incentives for more meaningful responses to local feedback. This approach involves looking beyond particular policies and asking about the constitutional framework policymakers work within (Buchanan and Tullock 1962). Beyond the level of public policy, we should consider how non-state actors can coordinate and incorporate local knowledge in order to create a responsive and incentive-compatible approach to crime prevention.

Government Failures as Entrepreneurial Opportunities When government fails to provide the security, dispute resolution, or justice that individuals seek, this creates incentives for non-state actors to provide whatever governance is lacking. While the state is often defined in terms of its monopoly on the use of force, that monopoly is rarely total or complete. Law and governance may therefore be provided by individuals who are not affiliated with the state when the state fails to provide such governance. Individual preferences for governance vary, so this supplemental provision may exist regardless of whether government has failed. Other scholars in the coproduction literature have noted this tendency for individuals to supplement government production of security when they feel it has been undersupplied. For example, Ann Schneider (1987: 613) notes, “Coproduction also permits persons for whom the government under-supplies a particular good to use private market solutions or voluntary organizations as a means to supplement government provision (Weisbrod 1975).” Sometimes individuals for whom the government has provided insufficient security simply do not feel that the government is devoting sufficient resources to policing their neighborhood or deterring criminals. These individuals may supplement using neighborhood watch programs, private firearm ownership, burglar alarms, anti-theft technologies on cars such as OnStar and LoJack, or a wide range of other voluntary solutions (Ayres and

58  Nathan Goodman Levitt 1998; Cook and MacDonald 2011). Cook and MacDonald (2011) argue that these private responses to crime played an important role in the crime decline since the 1990s that has been overlooked by other scholars studying the crime decline, such as Levitt (2004), Blumstein and Wallman (2000), and Zimring (2007). For some individuals, however, government may provide insufficient security precisely because the form of law enforcement they do provide makes individuals unable to trust the criminal justice system. African Americans face police brutality, racial profiling, and corruption while also feeling that police will not protect them, and therefore “the simultaneous criticism of too much and too little policing may be valid” (Ostrom and Whitaker 1974: 207). This also occurs for individuals who work within black markets, such as drug dealers and sex workers. Because their professions are criminalized, these individuals cannot rely on the criminal justice system to protect them or on the courts to resolve their disputes. In a sense, these individuals see criminal justice as simultaneously oversupplied and undersupplied. The criminal justice that is supplied is not a “good” they consume but rather a “bad” that they attempt to evade (E. Ostrom 1973). And in order to evade this unwanted product of state action, they must avoid consuming the goods that the criminal justice system provides as well. Individuals within black markets often rely upon gangs as mutual protection associations in order to fill this governance gap (Skarbek 2014). Early Bloomington school researchers were aware of this, and Vincent Ostrom noted that gangs largely functioned as mutual protection organizations as early as the 1970s (V. Ostrom 1973: 177). Boettke and Leeson (2009) note that a similar process operated within the Soviet Union and Eastern European countries, as well as other settings that feature black markets. The absence of an effective legal system to protect individual rights “creates a profit opportunity for individuals who can devise institutions of private governance that protect others against the threat of private predation” (ibid: 255). Boettke and Leeson call the individuals who exploit this profit opportunity “institutional entrepreneurs” (ibid). But individuals who work within black markets are not the only people who see justice undersupplied because the justice the state supplied poses a threat to their freedom and well-being. Members of minority groups can also find themselves in this situation. For example, transgender individuals have a variety of negative experiences with police and the prison system. They report high rates of bias and assault from police officers, and are at high risk of sexual violence when they are incarcerated. All of these negative outcomes are more likely for transgender people of color (Grant, Mottet, and Tanis 2011). This leads to a situation where 46% of transgender respondents feel uncomfortable seeking police assistance (Grant, Mottet, and Tanis 2011). In New York City, the Audre Lorde Project has taken this as an opportunity to develop alternative means of providing security, operating a safe neighborhood campaign that trains local businesses to respond

The Coproduction of Justice  59 to conflicts and protect transgender people from violence without calling the police (Spade 2015). This is an example of social entrepreneurship that leads to improved public safety and security. In some sense this coproduction supplements state provided criminal justice, but in another sense it is directly antagonistic to it. The wreckage left by the criminal justice system’s institutional failure provides fertile ground for new experiments in the provision of justice and security. When police, prosecutors, and prisons cannot effectively meet the needs of a population, then that population has an incentive to find alternative means of providing security. They possess knowledge of their circumstances of time and place, their community norms, and the threats they face. This means that the solutions they adopt are likely to be better suited for their unique circumstances than those designed from the top down. Where the solutions adopted are unsuitable, adjustment is more likely because the project is fueled by the resources of private individuals rather than subsidized by taxpayers. These incentives create an effective space for experimentation in the provision of governance, security, and justice. Wherever there is institutional failure, this failure creates profit opportunities that institutional entrepreneurs may exploit by protecting others from private predation.

Customary Law and Incentive-Compatible Justice Bruce Benson (2011) has documented a variety of cases where law and security are produced by non-state actors rather than by the state. One particularly intriguing example is customary law among the Anglo-Saxons. Customary law was provided on a decentralized basis, largely at the town level. Rather than employing prosecutors to charge offenders with crimes against the state, customary law was primarily based upon torts. Those who had been wronged by an offender had the power to sue that offender. This created incentives for participation by victims, as they received restitution. It also limited the scope of offenses covered by the justice system, creating a system where crimes required victims in order to be addressed. Without victimless crimes, black markets were not created, and so there were not particular classes deterred from engaging with the justice system. Without top-down administration, the justice system was responsive to and dependent upon input from members of the community. When law is developed from the bottom up it “requires widespread acceptance.” Reciprocity plays a powerful role in customary legal systems. Respect for others and for community norms, combined with a willingness to defend the legal claims of others, creates mutual support for one’s legal claims. Repeated dealings and the hope that one’s rights will be respected and enforced by the community provides strong incentives for protecting the rights of others. This strengthens the incentives for individuals to participate in coproduction that are already present as a result of the tort-driven

60  Nathan Goodman nature of the system. The particular enforcement mechanism used by the Anglo-Saxons was known as the “hundred.” The hundred was divided into smaller groups of neighbors known as “tithings.” Individuals informed the leader of the hundred, the hundredsmann, of thefts. The hundredsmann spread this information to the tithings, who pursued the offender. Participating in a tithing provided important benefits, and those outside of tithings were effectively ostracized. This created strong incentives to participate in the tithings, which required obeying the law and assisting in the pursuit of offenders. A strong system of community accountability developed, so people felt bound to respect their neighbors’ rights and assist them in pursuing justice (Benson 2011). Customary legal systems are limited in their use of violence, because violence and conflict are costly to initiate. In a state-run system of criminal justice this violence is subsidized. A SWAT team receives a federal grant in order to engage in armed raids, for example (Balko 2013; Boettke, Lemke, and Palagashvili 2016). However, in a customary legal system individuals and voluntary associations must bear the costs of the violence and conflict they participate in. This deters violence, allowing the customary legal system to remain sustainable and avoid excessive force that alienates community members and deters them from participating in coproduction. As the power of the king increased, the Anglo-Saxon system of customary law was displaced by authoritarian law. Legal power shifted from the community to the crown, and this weakened incentives for coproduction. The Norman Conquest led to the rise of a powerful monarchy. These monarchs replaced “the old restitution-based system . . . with a system of fines and capital punishment” (Benson 2011: 47). The monetary payments were increasingly allocated to the crown rather than crime victims, which reduced the incentives of victims to help coproduce. The top-down administration of justice also reduced incentives for reciprocal maintenance of justice, which led to the hundreds and their local court systems declining. “As the business of justice increasingly became the accumulation of royal revenues, voluntary participation in the justice process naturally declined” (ibid: 49). The decline of voluntary participation in the justice system led the Normans to coercively demand participation. The Normans created an institution called the frankpledge, which served a similar function to the tithings that had since atrophied. The frankpledge members “were expected to pursue and capture thieves and perform court duties” (ibid: 49). If they failed in these obligations, the entire frankpledge could be fined. The Normans destroyed a system based on reciprocity and voluntary coproduction. In its place, they used fines and threats. The key distinction between the customary law the Anglo-Saxons employed and the authoritarian law that followed it is that customary law was largely based on restitution rather than punishment. Restitution created significant incentives for coproduction by creating concrete benefits,

The Coproduction of Justice  61 namely monetary compensation, for crime victims who seek justice through the legal system. Incarceration, on the other hand, is punitive in orientation. Crime victims may derive some satisfaction from seeing people who abused them face punishment. They may be pleased that, at least for a time, their abusers are incarcerated and therefore limited in their capacity to harm people outside prison walls. But beyond these feelings, incarceration does not provide concrete benefits for crime victims. Restitution does provide these benefits. The purpose of restitution is to make the victim whole. Incarceration arose in a context of authoritarian law, which tends to transfer resources towards politically powerful groups rather than from offenders to victims. In the case of the Norman conquest, fines were used to gather revenue for the crown. Under Elizabeth, prisons were established in England “to ‘reform’ the idle poor by confining them and forcing them to work at hard labor” (Benson 2011: 71). This imprisonment effectively amounted to a political transfer from workers to landowners. Politically powerful interest groups continue to use the criminal law to extract resources today. Police take property from suspects, often without charges or trial, and auction it off to supplement their budgets (Worrall 2001). Local governments use fines from minor infractions to gather revenue, hurting their poorest residents in the process (United States Department of Justice 2015). Prison guard unions and corrections contractors alike have incentives to seek laws that fill prisons and lead to increases in government spending on incarceration (Volokh 2008). Restitution-based justice systems transfer resources from offenders to victims and create incentives for reciprocal enforcement. Punitive systems transfer resources from suspects, offenders, and taxpayers to special interest groups. A paradigm shift away from punishment and towards restitution is one solution that could create a justice system that serves the public rather than the powerful (Barnett 1977). The Anglo-Saxon example may seem removed from the contemporary context. However, political economists such as Stringham (2015), Leeson (2014), and Benson (2011) have documented examples of justice and governance produced by non-state actors in a variety of historical and contemporary contexts, forming a substantial literature on “analytical anarchism” (Boettke 2005). Regardless of customary law’s contemporary feasibility, the evolution away from customary law towards authoritarian law illuminates how incentives to participate in the justice system were altered by legal reforms. The customary legal system encouraged citizens to participate in the justice system, act on their particular local knowledge, and generally co-produce desirable outcomes. These incentives were altered by a shift from customary law to authoritarian law. Asking whether analogous changes accompanied the rise of mass incarceration may help uncover hidden costs of America’s prison state. Criminal justice reform proposals may also be evaluated through this lens, making us more attentive to how reforms interfere with a complex and historically situated social system.

62  Nathan Goodman

Directions for Further Research: Toward a Bloomington School Analysis of Mass Incarceration Elinor Ostrom’s research on coproduction and municipal policing featured rich empirical work, and diverse empirical research methods are a key feature of the Bloomington School research program that Elinor and Vincent Ostrom founded. They combined this empirical work with a theoretical framework drawn from the best of political economy. As early participants in the public choice revolution, the Ostroms studied the incentives facing individuals in diverse institutional contexts, recognized that self-interest shapes behavior in non-market spaces, and analyzed risks of opportunistic behavior. Drawing on the Austrian School, the Ostroms explored the imperfections of human knowledge and the important role that local knowledge plays in effective decision making. By embracing realistic assumptions regarding both self-interest and the limits of knowledge, the Ostroms developed a framework reminiscent of the contemporary research program in “robust political economy” (Pennington 2011; Leeson and Subrick 2006). To truly understand the relationship between coproduction and mass incarceration, the Bloomington School’s research methods can be effectively applied to the problem. In this chapter, I have synthesized the findings of other researchers and integrated them within the Bloomington School’s Institutional Analysis and Development framework to provide an agenda for future original empirical research. Particularly fertile areas include field research in communities directly impacted by mass incarceration and similar qualitative approaches. Researchers should interview and observe individuals whose friends or family members have been incarcerated and observe interactions in neighborhoods with high incarceration rates or criminalized occupations. Venkatesh (2006) studied the underground economy in Chicago’s Maquis Park neighborhood and found that formal legal rules were often violated, including by authority figures, in ways that allowed illicit social orders to flourish. Goffman (2014) conducted ethnographic research on fugitives and criminals in Philadelphia and their responses to policing and surveillance. A Bloomington School approach would build upon this work and pay particular attention to the provision of governance within the community, the informal rules that govern the community, and how these institutions interact with other institutions in society. The Ostroms analyzed institutions at three levels: “an operational level, a collective choice level, and a constitutional level” (Herzberg 2015). Decisions made by individuals on a daily basis occur at the operational level, while the collective choice level sets policies that constrain these choices. The rules governing collective decision making are set at the constitutional level (Herzberg 2015; Kiser and Ostrom 1982). This analytical framework can illuminate the norms that emerge in communities that face high levels of policing and incarceration. The norms that impact daily interaction within these communities, including anti-snitching

The Coproduction of Justice  63 norms, function on the operational level. Public policies that shape their interaction with the criminal justice system, as well as rules enforced by community organizations (such as non-profits and gangs) are developed at the collective choice level. Finally, the rules that govern the selection of these policies are constitutional level institutions (Buchanan and Tullock 1962). To fully understand the impact of criminal justice policies on marginalized communities, and to appreciate what shapes these policies, researchers should examine these interlocking institutions. Research on interlocking institutions within communities that face high rates of incarceration could be complemented by research on institutional entrepreneurship and coproduction within these communities. Case studies in institutional entrepreneurship should examine the governance gaps that created opportunities for institutional entrepreneurs. They may also analyze the constitutional rules that govern collective decision making by non-state providers of governance. Skarbek’s Social Order of the Underworld (2014) is an excellent example of a study in institutional entrepreneurship and the provision of governance by non-state actors. Future research could continue to examine gangs as mutual protection associations, study projects within LGBT communities such as the Audre Lorde Project’s Safe OUTside the System Collective, or examine the role of churches and other community organizations in promoting social order. By emphasizing coproduction, interlocking institutions, and governance beyond markets and states, this research could continue to build the links between analytical anarchism and the Bloomington School research program. Aligica and Boettke (2009) have already noted substantial affinities between analytical anarchism and the Bloomington School, and applying Ostrom’s insights from the police studies to research non-state provision of governance would further build these connections. These case studies can be used to illuminate a variety of cases where governance emerges within communities, as well as how this governance is shaped by criminal justice policies. This analysis may be supplemented by examining large scale data, including the variation of crime rates among neighborhoods and changes in rates of reporting crimes to police over time, demographic groups, and neighborhoods. Eventually this secondary data analysis could be combined with case studies to develop a theoretical model that explains the changes in reporting of crimes over time. Baumer and Lauritsen (2010) have made substantial progress in studying and explaining the increase in reporting rates in recent decades, but it remains a multicausal phenomenon that may be illuminated through multiple research methods. Case studies that examine individual motivations for reporting and not reporting crime can help identify whether mass incarceration has deterred crime reporting, and if so what other factors have outweighed this phenomenon to increase reporting rates. Research on coproduction might also draw on existing literature on how government spending crowds out private spending on social projects.

64  Nathan Goodman Fishback (2010) finds that private social welfare spending is substantially higher in the United States than in Nordic countries that rely on higher rates of government social spending. It is possible that similar crowding out occurs with the private and public provision of security, and examining this possibility may illuminate how public policy shapes coproduction of criminal justice. Understanding mass incarceration requires recognizing that governance is not solely provided by states. While the state is often conceptualized as a monopoly on force and law, non-state actors provide governance in a variety of ways. More research is needed to learn how mass incarceration has altered incentives for non-state actors to contribute to the production of justice, security, law, and order. Law and order can emerge from the bottom up, and scholars should study which institutional arrangements enable this self-governance and which frustrate its development. Communities directly impacted by mass incarceration are indispensable for answering these questions, because these communities are filled with institutional entrepreneurs who have developed solutions to the governance gaps produced by the drug war, mass incarceration, overcriminalization, and discriminatory policing. By expanding our understanding of institutions, Ostrom’s research on coproduction provides a valuable starting point for critical inquiries regarding criminal justice.

Note 1 Ostrom later applied the coproduction concept to study other public services, including infrastructure production in Brazil and primary education in Nigeria (E. Ostrom 1996).

References Alexander, Michelle (2010). The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York: The New Press. Aligica, Paul and Boettke, Peter (2009). Challenging Institutional Analysis and Development: The Bloomington School. New York: Routledge. Asbury, Bret (2011). “Anti-Snitching Norms and Community Loyalty.” Oregon Law Review 89: 1257–312. Ayres, Ian and Levitt, Steven (1998). “Measuring Positive Externalities from Unobservable Victim Precautions: An Empirical Analysis of Lojack.” Quarterly Journal of Economics 113: 43–77. Balko, Radley (2013). Rise of the Warrior Cop: The Militarization of America’s Police Forces. New York: PublicAffairs. Barnett, Randy (1977). “Restitution: A New Paradigm of Criminal Justice.” Ethics 87: 279–301. Baumer, Eric and Lauritsen, Janet (2010). “Reporting Crime to the Police, 1973–2005: A Multivariate Analysis of Long-Term Trends in the National Crime Survey (NCS) and National Crime Victimization Survey (NCVS).” Criminology 48: 131–85.

The Coproduction of Justice  65 Bench, Lawrence and Allen, Terry (2003). “Investigating the Stigma of Prison Classification: An Experimental Design.” The Prison Journal 83: 367–82. Benson, Bruce (2011). The Enterprise of Law: Justice Without the State. Oakland, CA: Independent Institute. Blumstein, Alfred and Wallman, Joel (2000). The Crime Drop in America. New York: Cambridge University Press. Boettke, Peter (2005). “Anarchism as a Progressive Research Program in Political Economy.” In Anarchy, State, and Public Choice, ed. E. Stringham. Cheltenham: Edward-Elgar Publishing. Boettke, Peter. and Leeson, Peter (2009). “Two-Tiered Entrepreneurship and Economic Development.” International Review of Law and Economics 29: 252–9. Boettke, Peter, Palagashvili, Liya, and Lemke, Jayme (2013). “Riding in Cars with Boys: Elinor Ostrom’s Adventures with the Police.” Journal of Institutional Economics 9: 407–25. Boettke, Peter, Lemke, Jayme, and Palagashvili, Liya (2015). “Re-Evaluating Community Policing in a Polycentric System.” Journal of Institutional Economics 12: 305–25. Buchanan, James (1996). “Federalism and Individual Sovereignty.” Cato Journal 15: 259–68. Buchanan, James and Tullock, Gordon (1962). The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor, MI: University of Michigan Press. Camp, Scott and Gaes, Gerald (2005). “Criminogenic Effects of the Prison Environment on Inmate Behavior: Some Experimental Evidence.” Crime & Delinquency 51: 425–42. Chen, M. Keith. and Shapiro, Jesse (2007). “Do Harsher Prison Conditions Reduce Recidivism? A Discontinuity-Based Approach.” American Law and Economics Review 9: 1–29. Cid, José (2009). “Is Imprisonment Criminogenic? A Comparative Study of Recidivism Rates Between Prison and Suspended Prison Sanctions.” European Journal of Criminology 6: 459–80. Cook, Philip and MacDonald, John (2011). “The Role of Private Action in Controlling Crime.” In Controlling Crime: Strategies and Tradeoffs, ed. Philip Cook, Jens Ludwig, and Justin McCrary. Chicago, IL: University of Chicago Press. D’Amico, Daniel (2015). “Knowledge Problems and Proportionality.” Criminal Justice Ethics 34: 1–25. Fishback, Price (2010). “Social Welfare Expenditures in the United States and the Nordic Countries: 1900–2003.” National Bureau of Economic Research, Working Paper 15982, May. Available at www.nber.org/papers/w15982. Forman Jr., James (2012). “Racial Critiques of Mass Incarceration: Beyond the New Jim Crow.” New York University Law Review 87: 101–46. Gaes, Gerald and Camp, Scott (2009). “Unintended Consequences: Experimental Evidence for the Criminogenic Effect of Prison Security Level Placement on Post-Release Recidivism.” Journal of Experimental Criminology 5: 139–62. Goffman, Alice (2014). On the Run: Fugitive Life in an American City. Chicago, IL: University of Chicago Press. Grant, Jaime, Mottet, Lisa, and Tanis, Justin (2011). “Injustice at Every Turn: A Report of the National Transgender Discrimination Survey.” Available at www.thetaskforce. org/static_html/downloads/reports/reports/ntds_full.pdf, accessed October 1, 2015.

66  Nathan Goodman Hayek, Friedrich (1945). “The Use of Knowledge in Society.” American Economic Review 35: 519–30. Herzberg, Roberta (2015). “Governing Their Commons: Elinor and Vincent Ostrom and the Bloomington School.” Public Choice 163: 95–109. Ishak, Samir (1972). “Consumers’ Perception of Police Performance: Consolidation vs. Deconcentration: The Case of Grand Rapids, Michigan Metropolitan Area.” Ph.D. dissertation, Indiana University. Kiser, Larry and Ostrom, Elinor (1982). “The Three Worlds of Action: A Metatheoretical Synthesis of Institutional Approaches.” In Strategies of Political Inquiry, ed. Elinor Ostrom. Beverly Hills, CA: Sage. Lavoie, Don (2015). Rivalry and Central Planning: The Socialist Calculation Debate Reconsidered. Fairfax, VA: Mercatus Center. Leeson, Peter. and Subrick, J. Robert (2006). “Robust Political Economy.” Review of Austrian Economics 19: 107–11. Levitt, Steven (2004). “Understanding Why Crime Fell in the 1990s: Four Factors that Explain the Decline and Six that Do Not.” Journal of Economic Perspectives 18: 163–90. Masten, Jamie (2009). “ ‘Ain’t No Snitches Ridin’ Wit’ Us’: How Deception in the Fourth Amendment Triggered the Stop Snitching Movement.” Ohio State Law Journal 70: 701–53. Miller, Lisa (2015). “What’s Violence Got to Do with It? Inequality, Punishment, and State Failure in U.S. Politics.” Punishment and Society 17: 184–210. Mises, Ludwig von (1920). “Economic Calculation in the Socialist Commonwealth.” In Collectivist Economic Planning, ed. F.A. Hayek. New York: Augustus M. Kelley. National Research Council (2014). The Growth of Incarceration in the United States: Exploring Causes and Consequences, ed. Jeremy Travis, Bruce Western and Steven Redburn. Washington, DC: The National Academies Press. Ostrom, Elinor (1973). “On the Meaning and Measurement of Output and Efficiency in the Provision of Urban Police Services.” Journal of Criminal Justice 1: 93–111. Ostrom, Elinor (1996). “Crossing the Great Divide: Coproduction, Synergy, and Development.” World Development 24: 1073–87. Ostrom, Elinor, Baugh, William, Gaurasci, Richard, Parks, Roger, and Whitaker, Gordon (1973a). Community Organization and the Provision of Police Services. Beverly Hills, CA: Sage Publications. Ostrom, Elinor, Parks, Richard, and Whitaker, Gordon (1973b). “Do We Really Want to Consolidate Urban Police Forces? A Reappraisal of Some Old Assertions.” Public Administration Review 33: 423–32. Ostrom, Elinor, and Whitaker, Gordon (1974). “Community Control and Governmental Responsiveness.” In Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy Analysis, ed. M.D. McGinnis. Ann Arbor, MI: The University of Michigan Press, 203–31. Ostrom, Vincent (1973). The Intellectual Crisis in American Public Administration. Tuscaloosa, AL: University of Alabama Press. Ostrom, Vincent, Tiebout, Charles, and Warren, Robert (1961). “The Organization of Government in Metropolitan Areas: A Theoretical Inquiry.” American Political Science Review 55: 831–42.

The Coproduction of Justice  67 Parks, Roger (1979). “Assessing the Influence or Organization on Performance: A Study of Police Services in Residential Neighborhoods.” Ph.D. dissertation, Bloomington, IN, Indiana University, Department of Political Science. Parks, Roger, Baker, Paula, Kiser, Larry, Oakerson, Ronald, Ostrom, Elinor, Percy, Stephen, Vandivort, Martha, Whitaker, Gordon, and Wilson, Rick (1981). “Consumers as Coproducers of Public Services: Some Economic and Institutional Considerations.” In Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy Analysis, ed. Michael McGinnis. Ann Arbor, MI: The University of Michigan Press, 381–91. Pennington, Mark (2011). Robust Political Economy: Classical Liberalism and the Future of Public Policy. Northampton, MA: Edward-Elgar Publishing, Inc. Rogers, Bruce and Lipsey, C. McCurdy (1974). “Metropolitan Reform: Citizen Evaluations of Performances in Nashville-Davidson County, Tennessee.” Publius 4: 19–34. Schneider, Anne (1987). “Coproduction of Public and Private Safety: An Analysis of Bystander Intervention, ‘Protective Neighboring,’ and Personal Protection.” Political Research Quarterly 40: 611–30. Schoenfeld, Heather (2010). “Mass Incarceration and the Paradox of Prison Conditions Litigation.” Law & Society Review 44: 731–68. Skarbek, David (2014). The Social Order of the Underworld: How Prison Gangs Govern the American Penal System. New York: Oxford University Press. Soares, Rodrigo (2004a). “Crime Reporting as a Measure of Institutional Development.” Economic Development and Cultural Change 52: 851–71. Soares, Rodrigo (2004b). “Development, Crime, and Punishment: Accounting for the International Differences in Crime Rates.” Journal of Development Economics 73: 155–84. Spade, Dean (2015). Normal Life: Administrative Violence, Critical Trans Politics, and the Limits of Law. Durham, NC: Duke University Press. Stringham, Edward. (2015). Private Governance: Creating Order in Economic and Social Life. New York: Oxford University Press. Tiebout, Charles (1956). “A Pure Theory of Local Expenditures.” Journal of Political Economy 64: 416–24. U.S. Department of Justice. Civil Rights Division (2015). “Investigation of the Ferguson Police Department.” Available at www.justice.gov/sites/default/files/opa/ press-releases/attachments/2015/03/04/ferguson_police_department_report_1.pdf Venkatesh, Sudhir (2006). Off the Books: The Underground Economy of the Urban Poor. Cambridge, MA: Harvard University Press. Wagner, Peter and Rabuy, Bernadette (2016). “Mass Incarceration: The Whole Pie 2016.” Prison Policy Initiative. Available at www.prisonpolicy.org/reports/ pie2016.html, accessed August 29, 2016. Weingast, Barry (1995). “The Economic Role of Political Institutions: MarketPreserving Federalism and Economic Development.” Journal of Law, Economics, and Organization 11: 1–31. Weisbrod, Burton Allen (1975). “Toward a Theory of the Voluntary Non-Profit Sector in a Three-Sector Economy.” In Altruism, Morality, and Economic Theory, ed. E. Phelps. New York: Russell Sage Foundation. Western, Bruce and Pettit, Becky (2004). “Mass Imprisonment and the Life Course: Race and Class Inequality in U.S. Incarceration.” American Sociological Review 69: 151–69.

68  Nathan Goodman Worrall, John (2001). “Addicted to the Drug War: The Role of Civil Asset Forfeiture as a Budgetary Necessity in Contemporary Law Enforcement.” Journal of Criminal Justice 29: 171–87. Zimring, Franklin (2007). The Great American Crime Decline. New York: Oxford University Press.

4 The Certainty of Punishment and the Proportionality of Incarceration Chris Barker

Introduction According to the International Crime Victimization Survey, US victimization rates are similar to those of Canada, Switzerland, and Denmark (Lappi-Seppälä 2015; Farrington, Langan, and Tonry 2004). Yet, unlike these countries, the US has a massive and expensive incarcerated population that costs states such as New York as much as $60,000 per inmate per year (Wagner and Rabuy 2016; Henrichson and Delaney 2012). Crimes that would draw fines in Germany are punished in the US by lengthy jail and prison sentences, often under harsh conditions justified by an Anglo-American commitment to “less eligibility” (Kugler et al. 2013; Sieh 1989; Blecker 2013: 90–5). The number of persons incarcerated in the United States has risen almost every year since the early 1970s, and incarceration as a tool of crime control disparately affects minority groups. For example, African Americans report using drugs at about the same rate as whites, but African Americans are vastly over-represented in drug arrests (38% of all drug offense arrests) and in state prisons (59% of those who are serving time for a drug offense) (Turner 2014). Observations of these trends—and fears that they may become the American norm—have prompted reformers to argue for both piecemeal and global changes in American criminal justice. As it stands, American patterns of incarceration interfere with the principles of liberal self-government. In thinking about how and why to change the criminal justice system, I argue below that reformers can take guidance from 18th century Italian penal reformer Cesare Beccaria. Beccaria’s writings advance social welfare and the greatest happiness, primarily through the prevention of crime (1995: 31). But, for a utilitarian who is focused on crime control, Beccaria spends a surprising amount of effort proposing limits to criminalization and leniency in punishment of the type that we associate with rights-based arguments. For example, the belief that only the guilty should be punished reminds not of utilitarianism, but of Kant’s maxim that someone “must first have been found worthy of punishment before any thought is given to the possible utility which he or his fellow citizens might derive from his punishment”

70  Chris Barker (Kant 1991: 155). Beccaria also protects a liberal and Millian sphere of non-harmful activity in which the state ought not to interfere. While this chapter could dwell on the differences between Kantian deontologists and Beccarian consequentialists, as introductory studies tend to do, I choose to see these two positions as potential allies in the liberalization of criminal justice. Beccaria writes about criminal justice systems that are largely justified on retributivist grounds, and that punish with excessive severity. His own approach is consequentialist—it only looks forward. However, as J. S. Mill observes, one does not need agreement about first principles to have substantial reform or to recruit fellow-travelers. A Kantian retributivist can have more in common with a Beccarian utilitarian than with a defender of severe, retributive punishment, if both the Beccarian and the Kantian insist (as they do) upon the importance of the same secondary principles. Once we leave aside moral metaphysics, in fact, J. S. Mill reminds us that people “are more easily brought to agree in their intermediate principles, vera illa et media axiomata, as Bacon says, than in their first principles” (Mill 1969a: 29). As Mill also observes, a direct appeal to Bentham’s “oracle,” the principle of utility, or to Kant’s principle of right, is only necessary when two secondary principles conflict (Crimmins 2011: 138–9; Mill 1969a: 111). Of course, to say this is not to deny the place of inquiry into first principles. However, the endless confrontation between retributive intuitions and appeals to utility is not the only conversation that should occur, and it may not be the conversation that addresses the worst problems of the American criminal justice system. Here, in a nutshell, is Beccaria’s basic argument. Beccaria argues that harsh punishments can be made much more lenient, and yet remain effective, if they are unerring: small punishments that are swift, public, and certain may be more effective than intense but uncertain future punishments; and sentences of long duration may more effectively deter than brief, intense corporal punishment. After two centuries of experience with the carceral system, can Beccaria’s approach apply to the present? I argue that Beccaria’s emphasis on unerringness in punishment (lenient legal codes and judicial certainty) is misapplied and misunderstood if it leads to harsh punishment. Certainty and predictability of punishment was a lever that Beccaria used to force the door of absolutism; now, the same Beccarian certainty has, not as its intent but as a consequence, the subordination of the vulnerable in American criminal justice. I argue that we can and should use Beccarian leniency to defend a form of proportionality-based strictness in punishment, and I criticize practices that offer false certainty and strictness, usually by limiting judicial discretion and avoiding individualization through mandatory minimum sentences, or by enshrining in law a false individualization by relying on plea bargaining. The overreliance on determinate sentences, prosecutorial discretion, and executive pardoning power, and the blurring of the line between safe and punitive custody and between pre-trial and post-conviction incarceration, are signs of problems, not solutions. These

Certainty and Proportionality  71 workarounds may have practical utility (they save money and relieve congestion within the criminal justice system), and this may reassure the public and play well in Peoria, but they harm offenders, their families, and (in the long run) the public with whom offenders are reintegrated. In sum, preventing harm does not require being harsh, and the latter is unlikely to advance the former.

Leniency and Proportionality For Beccaria, governmental power can only be exercised legitimately and according to right insofar as it is for the “greatest utility” to the “greatest number” (Beccaria 1995: 11). On this conception of political right, the principle of proportionality is justified not, for example, by the Kantian isomorphism between crime and punishment, which licenses severity up to the utmost punishment for the worst crime. Leniency, for Kant, is not a valid principle of criminal justice, if at least we are to honor the humanity of wrongdoers and of their victims. Beccaria disagrees: harsh punishment is justified by the need to pay more heed to deterring serious crimes that cause more harm to society. The “obstacles which repel men from committing crimes ought to be made stronger the more those crimes are against the public good and the more inducements there are for committing them” (19, 20, 77). In “The Classification of Crimes” (Beccaria 1995: Chapter 8), the full typology of crime is laid out, from crimes that are most destructive of the greatest happiness to those that are the least destructive. The list comprises the following categories: 1) crimes against the state (sedition); 2) crimes against persons and their property; and 3) disturbances of the public peace (Beccaria 1995: 24–30). Importantly, acts that do not appear on the continuum of threats/harms to public safety ought not to be punished and should not be criminalized (Beccaria 1995: 20, 25; Kant 1991: 154). This is Beccaria’s contractarian side speaking (Bellamy 1995: xvii–xxi; Shuster 2016: 90–3). That is, it is his commitment to the view that the state’s power is the product of a contract between free, consenting individuals that licenses his emphatic claim that there is a sphere of liberty that people retain even in civil society, so long as their actions do not harm others. Thus far, I have adverted to Beccaria’s first principle and to his justification of punishment. However, Beccaria’s On Crimes and Punishments is a rhetorically complicated essay. The text presupposes an audience that is receptive to the harsh retributivism of the 18th century, and that supports the use of corporal and capital punishment. In Chapter 28—well into the book, and after much preparation for the recalcitrant reader—Beccaria declares that the death penalty is useless and inhumane (1995: 30, 70). He argues that extent of punishment, not intensity, is the chief deterrent (67). However, he has said earlier in the book that shorter and more certain sentences are better deterrents than long, indeterminate, and uncertain

72  Chris Barker sentences (63). Combining the arguments, we find that criminal justice systems should reject capital and corporal punishment, and that they can substitute short-but-certain punishment that is both swift and proportionate (48–9). Thus, if the “purpose [of punishment] is nothing other than to prevent the offender from doing fresh harm to his fellows and to deter others from doing likewise,” we need only combine a few basic elements to achieve crime control: “unswerving judicial severity” and a “lenient code of laws” (Beccaria 1995: 31, 63). The “futile excess” of punishment may be avoided by this combination of narrowing the scope of criminal liability, criminalizing fewer acts and punishing them with less severe sentences, and extending the exact justice of the judicial system so that the offender will inevitably get the predictable, prescribed sentence. Beccaria also wants justice to be procedurally swift and substantively lenient. To summarize, Beccaria’s basic description of criminal punishment is that punishment should be “public, speedy, necessary, the minimum possible in the given circumstances, proportionate to the crime, and determined by the law” (Beccaria 1995: 113). Beccaria’s secondary principles are, in order of exposition, something like the following: necessity, proportionality, unerringness, publicity, celerity, egalitarianism, and leniency (11, 19–21, 29–30, 36, 48–9, 51, 63–5).

Proportionality Proportionality is a key principle for Kant and Beccaria. Kant’s equality principle (“If you insult him, you insult yourself; if you steal from him, you steal from yourself; if you strike him, you strike yourself; if you kill him, you kill yourself”) creates well-known difficulties for punishment: Shall we rape the rapist, and, if not, why not? These problems are not insurmountable. As Nelson T. Potter, Jr. writes, “although one may not be able to find a qualitatively similar punishment, one can find a punishment that brings back upon the criminal his own conduct ‘if not in terms of its letter at least in terms of its spirit’. . .” (Potter 1998: 180; Pincoffs 1966: 9). However, as noted above, enforcement and punishment cannot be lenient if we are to honor the humanity both of wrongdoers and victims (Kant 1991: 155). Beccaria also relies on proportionality and is in some ways stricter than Kant in applying his principle, although he justifies proportionality by favoring leniency (Beccaria 1995: 19). Beccaria wants to assign a severity of offense according to the harm that such an offense does to the state, and, as noted above, he views not intensity but extent/duration as crucial to effective punishment. Both Beccaria and Kant struggle with equating crimes and punishment. Both, for example, hold that incarceration with hard labor is appropriate for some property crimes, but they arrive at their conclusions differently. Beccaria argues that theft without violence (“larceny”) should be punished with fines, and this is (on its surface) the most equal punishment for a pecuniary/property crime. However, the impact of fines on the

Certainty and Proportionality  73 offender’s family and the likelihood of non-payment bring him around to the (Kantian) view that temporary enslavement of the labor and person of an offender may be a proportionate punishment for theft (53). Does Beccaria give us principled guidance as to the limits of criminalization, so that we can say either that a particular act is not one for which one should be criminally liable, or that the probable sentence is too severe? Here, the Beccarian dream of a strict classification of offenses from most to least severe, and a corresponding continuum of punishments that applies to all regardless of social status/privilege, breaks down in practice. A 1991 sentencing commission report explains that “criminal codes often grow in a patchwork fashion, with new crimes added every legislative session. These crimes are placed into an existing hierarchy, usually without a great deal of effort being expended to ascertain the harm relative to other crimes.” Thus, “[t]he decision to rank the seriousness of a crime [was] often an ad hoc event driven by some exceptional set of circumstances” (Rich 2002: 699).

How to Use Beccaria How, then, should we use Beccaria to guide present-day reflections on criminal justice? In Harsh Justice, James Whitman (2003) argues that continental Europe and the United States and England have taken different paths in criminal justice. The former (especially France and Germany) treat all criminal offenders as if they were high-status persons subject to high-status punishments, such as beheading, fortress captivity, and other privileged forms of incarceration (custodia honesta). England and the United States treat criminal offenders as low-status persons subject to low-status punishments such as hanging, branding, and punitive incarceration. Whitman uses Beccaria to represent one justification for the Anglo-American leveling-down, namely the tender-yet-unbending approach that is ultimately motivated by demands for formal equality across all classes, which Beccaria explicitly defends in his chapter on the punishment of the nobility (Beccaria 1995: 51–2, 50–1). The United States and England thus adopt a harsh version of Beccarian egalitarianism and “formal equality” (Whitman 2003: 193), treating with enlightenment rigidity and exactness everyone who commits the same crime, regardless of the moral or legal personality of the offender. Beccaria, for Whitman, therefore became a “dead end” on the continent, where this harsh sort of egalitarianism is rejected (73). Whitman’s account of Beccaria’s influence and reception in the Anglo-American world suggests the following questions: Why can’t proportionality between crime and punishment take the depravity of the offender into consideration without indexing punishment to the “inscrutable malice of the heart” (Beccaria 1995: 23)? Why can’t formal equality include individualization precisely by defending individuals against the power of groups, whether we are concerned with the aristocracy or with the personal dependence of wives and children within a “republic of families” (51–2,

74  Chris Barker 60–1)? Why would it be enlightened to support determinate sentences and reduce judicial discretion, only to watch the criminal justice system become unremittingly and rigidly punitive? Why can’t Beccaria be applied in a manner that respects the leniency principle, which, as he says, is “my topic” (63–5)? Whitman, in his defense, is concerned not with the meaning of a theorist but with the comparative analysis of criminal justice systems (Whitman 2003: 33). He is concerned to explain a system’s “intoxication with degradation” versus a system’s normalization or approximation of outside life on the inside (87). Although my interest in Beccaria is motivated by the same concern, Beccaria, the theorist, can be used to avoid harshness in several key areas identified by Whitman—despite the fact that Whitman sees this harshness as in part the product of Beccarian ideas (Whitman 2003: 33–7, 163). The crucial problem in attempting to apply Beccaria’s principles in the present day is that Beccaria fully supports the use of incarceration and accepts the deterrent effect of long incarceration at hard labor. As observed above, the US criminal justice system is in many ways efficient and humane, but over-incarceration is precisely one of its key features, and reforming the system by relying on Beccaria thus seems as inapposite as Whitman fears it is. What follows in the remainder of this chapter is an attempt at having it both ways: certainty in punishment, but less severity in the conditions under which offenders are incarcerated; certainty of prosecution and the criminal conviction of the guilty, but discretion and individual evaluation in practices such as pressing charges, assessing risk and bail, and imposing sentences. To make my case, I argue that courts and corrections should promote certainty but not severity—not certain and disproportionately harsh punishment, and especially not severe punishment for minor or nonviolent offenses, but certainty about when an act cannot be done, and what will happen to the offender, both during and after punishment, as a result of the actus reus. In light of Beccaria’s complex rhetorical strategy, which replaces capital punishment with incarceration, I accept Beccaria’s logic and further liberalize his argument: why long, harsh, and possibly degrading incarceration if other proportional, swift, and certain alternatives exist? Paradoxically, I conclude that the certainty of bail schedules, mandatory minimums, and determinate sentences is merely a false and overly punitive appearance of certainty that leads to discriminatory outcomes in practice. The right type of certainty looks different. If it cannot be the certainty that an offender will be subject to educative and rehabilitative treatment, based on individualized risk assessment, then it can at least advance the certainty that punishment is bounded by proportionality, procedural fairness, presumption of innocence, and other secondary principles that protect against the abuse of police (Beccaria 1995: 29). Instead, the US criminal justice system mistakes harshness for certainty and proportionality, largely, I hypothesize, because democratic publics have not thought about and do not care about the secondary principle of proportionality, and because offense results in a

Certainty and Proportionality  75 de facto status degradation, so that we no longer worry about what happens to the offender, especially when they are out of sight. Inexperienced with the law and with the effects of punishment, publics endorse a system that does not do what they want it to do, and lacks the restraints that they should want it to have. This outcome is not Beccarian, and it is not an Enlightenment approach: it is a false and exaggerated equality of outcome that, as Whitman would say, treats many or most criminals as serfs and not citizens.

Problem Areas and Suggestions for Reform The basic policy that I suggest, and justify under Beccaria’s logic, is expanded use of “quick dips,” or short, graduated sanctions that are currently used for probation violations. In fact, those familiar with graduated sanctions will recognize the watchwords of this approach as “Beccarian”: graduated rewards and sanctions should be swift, certain, and proportional to be effective (Paternoster 2010). These short, punitive sentences provide an alternative to probation revocation. The approach may not rescale well to deal with violent person crimes. It also abstracts from enforcement questions, because it is typically applied to probation violators who have been caught in a probation violation because they are already under supervision. However, if swift, short, and certain sentences deter probation violators, then we should consider whether this set of secondary principles also satisfies a retributive public. I return to this question in my conclusion. In the following short sections, I highlight some of the areas in which the American criminal justice system fails to meet the standards of unerringness and leniency introduced above, and I suggest some reforms that may provide a more principled way of achieving leniency through unerringness and proportionality. My suggestion is that American justice has become too enamored with the wrong type of certainty, but that there are some fixes that can be instituted to achieve the right kind of leniency. Fines or Short, Graduated Sentences Above, I have noted the Beccarian defense of custodial sentences of long duration. Beccaria also argues that immediate, certain, and public punishments, even if they are small, can be effective. Is it possible to adapt this latter rationale to substitute fines (tariff fines or day fines) or short custodial sentences for long sentences? Instead of long sentences, which, as I argue in the conclusion, neither offenders nor the public think are effective deterrents or proportionate to the crime, perhaps graduated sanctions (quick dips) can be used to place individuals back in jail for minor probation violations—and as part of alternative sentences for misdemeanor and non-person felony offenses. For example, a 2013 Kansas law instituted “quick dips” and other graduated steps in felony probation violation in order to achieve “swift and

76  Chris Barker certain responses to offender non-compliance in the community” (Kansas Sentencing Commission; Kansas Legislative Research Department). As one official explains, before the law passed it was “one size fits all”: violating felony probation would result in the offender’s re-incarceration in prison. The 2013 law, justified by problems of economics and overcrowding but of use on a leniency rationale, forbids probation revocation for a first offense parole violation, except in cases where the offender a) endangers the community or himself, or b) commits a new offense or absconds. Instead, the offender who violates probation is subject to a brief period of commitment in the county jail. These quick dips are 2–3 day sentences that can be imposed any number of times until a maximum of 18 days is reached. In the case of further violation after the first dip, the offender may be sentenced either to 120 days or 180 days. These Tier Two punishments once again offer an alternative to sending the convicted felon back to prison, although any further violation can result in full probation revocation. There is some evidence across the several states experimenting with graduated rewards and sanctions that swift, certain, and short sentences can successfully contribute to “behavioral triage” (Hawken 2014; Hawken and Kleiman 2014). If short sentences are empirically viable alternatives to long, expensive custodial sentences in probation cases, then experiments with graduated sanctions should be extended—at least to misdemeanors, and perhaps to non-person felonies or to violent crimes. Overreliance on Plea and Sentence Bargaining Changes from the false certainty of plea bargaining to the properly individualized treatment of offenders are probably the most pressing reforms needed in the American criminal justice system. A much-cited figure is that more than 95% of felony charges are resolved by pleas, which allow prosecutors to avoid the uncertainty of jury trial. A few examples will illustrate the potential problem. With the exception of the Virgin Islands, 2015 federal data shows that there are no instances where double-digit percentages of federal felony cases went to trial (Statistical Information Packet 2015: 3). In one small-n judicial district, pleas resolved 100% of the 13 cases; in another, large-n district (Arizona), pleas resolved 99.3% of the 6,154 federal cases (Statistical Information Packet 2015: 5). Forced guilty pleas are also an issue. In the theoretical world of the US Attorney’s manual, forced pleas are forbidden: “Charges should not be filed simply to exert leverage to induce a plea, nor should charges be abandoned in an effort to arrive at a bargain that fails to reflect the seriousness of the defendant’s conduct.” However, prosecutors have massive discretion in charge bargaining. Moreover, plea bargaining is a one-sided product of largely unreviewable and secret negotiation (Rakoff 2014; Human Rights Watch 2013). The “charge the max” principle that is found in the US Attorneys’ manual (9–27.300—Selecting Charges—Charging Most Serious Offenses) is

Certainty and Proportionality  77 also a species of false certainty for offenders and for the community. The manual explicitly states that a “faithful and honest” application of the principle of charging the “most serious” offense that will achieve the highest range of punishment under the guidelines is compatible with “individualized assessment.” Thus, just as Beccarian egalitarianism does not mean “treat everyone with equal harshness,” the US Attorney’s manual does not support a simplistic “throw the book” mindset; prosecution should be “fair and effective.” Correctly, public opinion is not considered dispositive; it “deserves the prosecutor’s careful attention” but cannot justify a decision to prosecute or not to prosecute. Prosecutors who are not convinced of the proportionality between punishment and probable sentence, or that this is good use of limited resources, or that it is of sufficient merit to the community to deter/punish, should not prosecute (9–27.250). The guidance of the manual clearly sets out prudential limits on prosecuting offenders, but, in a particularly visible and scandalous case about the misuse of the Sarbanes-Oxley act to prosecute a fisherman, Yates v. US (2014), the standard “charge the maximum” arguably led to absurd consequences, as Justice Scalia observes (28–9). My aim here is not to provide a full analysis of prosecutorial discretion, which plays an important role in criminal justice (e.g., Bibas 2010), but simply to point out that the utilitarian defense of pleas, which reduce workload and costs in a congested system, can be seen as a cheap and artificial means of achieving false certainty. Even or especially when we take certainty seriously, we must acknowledge the importance of other principles that conflict with the cost-effectiveness of plea bargaining: the presumption of innocence, procedural fairness of a trial by one’s peers, proper grading of offenses (which is upset by charge bargaining) and a sentence that is proportionate to the crime (which is often violated by sentence bargaining). One practical workaround to the problem is to change the metric by which prosecutors’ performance is measured (Nugent-Borakove and Budzilowicz 2007). High conviction rates, in themselves, tell the public (and offenders) nothing about whether the good sort of certainty has been achieved. For example, high conviction rates but relatively low rates of prosecution may mean that prosecutors are cherry-picking easier cases (Ramseyer, Rasmusen, and Raghav 2009). This is to say nothing about actual prosecutorial misconduct, which is a problem because there is so rarely a remedy for it. The False Leniency of Pardons and Commutations Recent activity in criminal justice reform, such as rolling back federal and state mandatory minimums, or limiting legislation that impose sentence enhancements for classes of crimes, present attractive changes. However, if certainty is crucial, then relying on pardons and commutations for equity undermines the workings of criminal justice. At most, a varied, unpredictable, and biased

78  Chris Barker system should accept pardons and commutations as no more than transitional means of attaining more just outcomes (Gao 2014). Beccaria argues against the use of pardons and executive clemency because reliance on this tool of equity builds arbitrariness and uncertainty into the criminal justice system. In England, for example, the “Bloody Code” of 18th-century criminal punishment recognized over 200 separate capital offenses. Censorial jurists, such as Bentham, and even those who, like Blackstone, claimed merely to be reporting the law as it was, criticized the severity of criminal punishment as well as the scope of criminal liability (Blackstone 1979: IV, 4–5). Criminal codes that are seen as overly punitive are not properly applied by prosecutors, judges, or juries. They habitually invite impunity: in London and Middlesex in the 1790s, near to the time of Beccaria’s writing of his book, as many as 2/3 of those sentenced for capital offenses were subsequently pardoned, and Tocqueville and Lieber complain at the amount of executive clemency in New York (Sharpe 1990: 50; Beaumont and Tocqueville 1833: 232–7; Whitman 2003: 163–4, 169–70). Beccaria predicts that agents of punishment simply will not have the stomach to apply some sanctions, or to press charges when the probable sentence is too severe. The result is a breakdown of the rule of law, or “pious perjury” among lenient juries (Beccaria 1995: 64; Whitman 2003: 165). In the present day, Whitman cites cases where Christmas pardons in Germany encourage minor lawbreaking. Again in the present day, the sensible Beccarian worry about injecting discretion and arbitrariness into criminal punishment and executive clemency suggests that the criminal justice reform movement should not aim at piecemeal reform through, say, the Federal Office of the Pardon Attorney or through advocacy groups such as the Clemency Initiative. The reason for rejecting an executive-driven approach is not simply to achieve greater reform rather than settling for piecemeal reform; instead, the problem with executive clemency is that it shifts costs and attention from the legislative bodies that are primarily responsible for setting criminal sentencing minimums, and from the US and state sentencing commissions, who, as independent judicial bodies, have the primary responsibility to reduce sentencing disparities, promote transparency, and promote proportionality. Beccaria argues that clemency and pardon should be redundant where punishments are mild and judgment is regular and quick (Beccaria 1995: 111). Executive commutations and pardons seem attractively lenient, but they are evidence of a problem with the system: the more we deviate from the norms of mildness and certainty, the more they are necessary and the less certainty in punishment exists. When a harsh regime becomes so arbitrary that equity is more regular than the laws, executive action is to be preferred to nothing, but it is a trap and a snare to focus attention on the law’s execution. The generally salutary reforming influence that the Department of Justice’s Civil Rights Division has had in Ferguson, MO; Albuquerque, NM; Baltimore, MD; and other problem areas are, again, evidence of problems and not clear-cut solutions.

Certainty and Proportionality  79 “Less Eligibility” vs. Normalization of Conditions One of the places where early modern utilitarians’ commitment to contractarianism makes them sensitive to failures of due process is in the recognition of the difference between punitive and non-punitive incarceration, and between conditions that are less eligible than life on the outside or that approximate life on the outside of jail and prison. Although this may not be the greatest fault of the contemporary criminal justice system in terms of impact, in terms of deviation from principles the confusion over the type and manner of custody is likely the grossest of the American system’s failings. Like continental European countries, many countries in the Americas permit lengthy pre-trial detention. Argentina permits two years of investigative detention, in addition to pre-trial detention; Paraguay permits pre-trial incarceration (PTI) until the completion of the possible sentence, which can be as long as five years (Aborn and Cannon 2016). The United States does not make this mistake, but it still has the world’s largest pre-trial population (3/5 of its jail population), and in many cases confuses safe custody with punitive custody. Although it presents problems for staff trying to deal with pre-trial inmates—as I am told by one official in New Orleans, short-timers in PTI “don’t care” about acting well in the parish prison because they “got nothing to lose”—punitive and safe custody are simply different categories, as Beccaria argues (Beccaria 1995: 73–5). It follows, then, as James Mill argues, that those in safe custody should be able to hold jobs and to use their own money for “all unexceptionable indulgences” (1992: 197–202). After all, they are not being punished! The same point can be made about confusing different types of custody in the present day. About corrections at Guantánamo Bay, for example, a Navy spokesman sums up the difference. Captives who are “detained under the Law of War only as a security measure” have not been convicted and “should not be subjected to a penal environment or comingled with prisoners punitively incarcerated as a consequence of a criminal conviction.” This is quite correct, and the confusion of this distinction, or its abuse, has caused the US Army detention camp at Guantánamo Bay to be derided across the world as a highly punitive, politically motivated, and basically unjust regime (Rosenberg 2011). Given that there are over 11 million jail admittances across the country each year, this is a massively important category error. In fact, the problem of mass incarceration in the United States is, in this sense, a problem with county jails. It can be remedied by better attempts to limit punitive custody to those who have been tried and convicted, and by making using of quantitative risk-assessment tools that allow more nonviolent offenders to remain outside jail before their trial (Wagner and Rabuy 2016; Subramanian et al. 2015). Of course, it is argued that most of those awaiting trial are guilty, or even that they are “always guilty,” as I was told at a rural Kansas jail. However, conviction rates of pre-trial detainees do not bear out this observation:

80  Chris Barker according to a 2007 BJS report, 20% of detainees had their charges dismissed or were acquitted (Cohen and Reaves 2007: 7). Second, guilty of what? Prior to a trial, how can corrections staff know enough about the charges, the mental state, and degree of culpability of inmates to make this claim? This is an example of false certainty’s violating the basic presumption of innocence that should be enshrined in a country with the rule of law. One way of building greater certainty into incarceration is to change conditions in local jails, state prisons, and federal prisons. On the continent and in Scandinavia, prisons tend to follow a “normalcy” or “approximation” principle: life in prison should approximate life on the outside as closely as possible (Whitman 2003: 8, 87; Reiter, Sexton, and Sumner 2016; Benko 2015; Larson 2013). In Germany, the approximation of life outside custody is the legal standard of custody. The normality principle in Finland is understood to mean that the mere fact of incarceration is itself the entire punishment. Like Germany’s, the Norwegian “normality principle” holds that “[d]aily prison life should not be any different than ordinary life, as far as this is possible.” US prisons and jails embrace less eligibility, which is the converse of “normality” (Sieh 1989). Historically, one thinks back to the Philadelphia system’s “unsocial manner of life” where order, obedience, and silence frightened offenders more than corporal punishment (Lieber 1833b: xviii, Lieber 1833a: 289). The deterrent effect of incarcerating offenders under conditions of less eligibility is a traditional and long-standing commitment of American justice. In the present day, a past warden can call life in ADX Florence, a model Supermax prison, a fate “far much worse than death” (Binelli 2015). The question that must be asked is this: To what crime is this “worse-than-death” punishment proportionate? The tension between local control and standardization of conditions of incarceration has been present from the creation of the American prison system. For example, the first warden of Auburn Penitentiary, Elam Lynds, argued for enhanced discretion for guards and especially for wardens in order to permit local control and to advance (as he saw it) real reform and rehabilitation (Beaumont and Tocqueville 2011: 214–8). The obvious problem then as now is arbitrariness: certainty in punishment is undermined when the decision to employ corporal punishment is left to individual guards (445, 454). As a gesture towards the general problem, one can cite the record of inmates’ challenges at Attica. According to the New York Times, only one of the 228 prisoners accused of assaulting guards at Attica has been exonerated through the prison’s internal review process; in contrast, when four officers were criminally indicted in 2011 for beating an inmate, it was the first such indictment for a (non-sexual) assault in state history (Robbins 2015). It is hard to imagine that the proportion between crime and punishment will be maintained when offenders lack exit, transfer, and voice. To speak radically for a moment, inmates should likely have more say in the conditions under which they are held, if we desire to build agency in them. Interpreting and applying the principle “less eligibility” is also discretionary. At El Dorado Correctional Facility in Kansas, I am told that it

Certainty and Proportionality  81 is a guard-by-guard decision whether inmates are punished differently for the crimes they committed prior to being incarcerated, and, in effect, that guards individually decide whether custody is punitive or whether it is safe custody. The security officers maintain order by micromanaging inmates’ liberty—they stop them in the yard for improperly tied shoelaces to create order and a sense of discipline—but the guards also report that they think that incarceration itself is punishment enough. The staff, they say, does not link punishment to a specific crime, and they report that only the inmates do so, and chiefly to punish crimes against children and women. As one guard on Tennessee’s death row observes, “the past counts, but you can’t count it” (Blecker 2013: 100–1). Normalization of Re-entry Certainty and proportionality in punishment requires us to think not only about what happens during an offender’s period of incarceration, but about what happens after it. At the inception of the US penitentiary system, inmates would enter the Philadelphia system under conditions of complete anonymity: they would have a physical bag over their head so that they could not see other inmates; they would not see other inmates during their time in the penitentiary; and their criminal past and the fact of their incarceration would not be made public. The hope was that complete severance from their past criminal associates would help to advance the reform process. Communicative corruption (“contamination”) was, in this system, deemed to be a key cause of crime. Although, empirically, the solitary confinement of the Philadelphia system proved ineffective because of the mental strain that it placed on inmates, the virtue of anonymity has been lost to those incarcerated (or even arrested, much less charged and convicted) in the American system. A Beccarian approach intends to prevent crime (Beccaria 1995: 103). The prison as a “school of crime” is a disaster for crime prevention, and offenders struggle to return to society from prison while experiencing prejudice, the loss of job opportunities, and other disabilities. Without arguing it in this chapter, I suggest that an appearance of proportionality between crime and punishment might satisfy the public and aid offenders’ reintegration into the moral community. Again, this idea of a moral community is not a traditionally utilitarian concern, but we need help in bridging across basic moral intuitions as we develop a criminal justice system with sensible secondary principles.

Conclusion: Small-But-Certain Punishment This chapter revisits early utilitarian arguments for leniency in punishment in order to put recent reformist agitation and public policy in contact with Beccarian thinking. I take Beccaria to be trying to shift opinion towards an initially unpalatable leniency and away from sanguinary punishment.

82  Chris Barker Beccaria argues that the drudgery of a sentence at hard labor has more deterrent value, by virtue of its long duration, than fleeting but intense corporal or capital punishment. However, small interventions, when they are certain and unerring, also have larger deterrent effects than uncertain future punishment. With this in mind, it is possible to draw out the implications of Beccaria’s argument, or, if he would stop short of these implications, to liberalize his approach. Beccaria, for his part, makes the empirically contestable claim that corporal punishment must be used for person felonies so that the rich do not put a price on injuring the weak and poor (Beccaria 1995: 50). It is also an empirical question whether short, graduated punishments are more humane and have greater deterrent force. These claims must be tested, but, at the least, the evidence suggests that “quick dips” are effective deterrents (Hawken 2014; Hawken and Kleiman 2009; American Probation and Parole Association: 6–9). Above, Whitman argued that the fault of the American justice system was its tragic Beccarian reception of formal equality of status, which, coupled with exogenous leveling-down of the status of criminals so they were on par with slaves, makes American justice astonishingly harsh justice for all (2003: 163, 170–81). For Whitman, this has to do with democratic egalitarianism, criminal justice’s insulation from high-status reformers’ experience with and resistance to incarceration, and slavery (Whitman 2003: 176–7; Alexander 2012: 20–58). For Foucault, in contrast, the American penitentiary model actually springs from principles opposed to Beccaria’s, and in particular from a definition of crime that is the opposite of Beccaria’s severity gradation of crimes that bring more or less “harm to society” (Foucault 2015: 67–8, 253–4). Beccaria separated moral wrongdoing and infraction, while the penitentiary system consists in the “moralization of the penal system” under a Quaker theory of evil, and the consequent overpunishment of offenders (87, 107). The embrace of moral reform is a topic beyond this chapter, but intentionally so. Here, I have asked whether we can agree on reforming incarceration without agreeing on what exactly we seek (retribution, or rehabilitation, deterrence, or incapacitation) when we punish. I go back to the birth of the penitentiary, where incarceration was thought to break habits and train minds in a “new victory of mind over matter.” Then, as today, “getting people right” matters. That, as one client of the New Orleans Day Reporting Center tells me, can sometimes be as simple as having staff that “treat us like people” and “treat everybody the same.” This is especially crucial in an environment where, as Francis Lieber observes in the nineteenth century, the lack of pluralistic “struggle of different parties” leaves prisoners particularly vulnerable to poor treatment—and without any way of bargaining with prison administrators or using surrogates to advocate for themselves (Lieber 1833b: viii, xx; Beaumont and Tocqueville 1833: 57–8). This lack of exit and voice in American prisons and jails remains a crucial problem, and makes it that much more important for academics to learn from practitioners and, where possible, from offenders.

Certainty and Proportionality  83 It may be that a liberal state with limited enforcement resources, selective law enforcement patterns, an adversarial judicial system, popular pressures on elected judges and prosecutors, and a corrections system based on “less eligibility” cannot meet Beccaria’s objective of unerringness and proportionality in punishment. Beccaria explicitly rejects the “broken windows” policing approach that would try to control crime by micromanaging vandalism and other nonviolent property crimes (Beccaria 1995: 103; but see Kelling and Wilson 1982; Kelling 2015). It is possible that Beccaria’s argument against 18th century retributivism and nobility-favoring justice has served its purpose and that today we must find a different justification of leniency. Moreover, Beccaria’s argument seems to provide a direct route to the embrace of mandatory minimums and other legal reforms that have prompted massive increases in the incarcerated population since the 1970s (Whitman 2003: 70). Public opinion should not be controlling, but there is evidence that the people recognize that “strict” proportionality between the length of a sentence and a crime is not closely connected with the aim of crime control. In one survey, 91% of respondents agreed that “[i]t does not matter whether a nonviolent offender is in prison for 21 or 24 or 27 months. What really matters is the system does a better job of making sure that when an offender does get out, he is less likely to commit another crime” (PEW 2012). There is also state-level American survey research indicating that the people do not want to see incarceration used in some (non-person felony) cases, and prefer alternatives such as restitution and community service. (Here, the public shares Beccaria’s view of proportionality: the basic sentence for theft should be restitution/fines.) Public preferences concerning the use of custodial sentences also vary by country. According to one result reported in the 1996 International Criminal Victimization Survey, “When asked to decide on a sentence for a burglar convicted for a second time, people in Canada, England and Wales, Northern Ireland, Scotland and the United States chose prison by a wide margin over other sanctions. By contrast, people in Austria, Finland, France, the Netherlands, Sweden and Switzerland much preferred community service.” However, in the US a more recent national large-n survey found that 6 in 10 crime victims “prefer shorter prison sentences and more spending on prevention and rehabilitation to prison sentences that keep people in prison for as long as possible.” Finally, mirroring the “prisons as schools of crime” reasoning, “52 percent of victims also said that prison makes people more likely to commit crimes again” (Ingraham 2016). All this is to say that opinions on principles, secondary principles, and policies vary according to differing moral intuitions, experiences of different policing and correctional techniques, and views on different policies. Anecdotal evidence from offenders suggests that their perceptions of incarceration also vary. For some offenders, jail and prison offer security in the form of “three hots and a cot.” Nevertheless, the public is broadly supportive of the move away from incarceration (Hartney and Marchionna 2009).

84  Chris Barker Although there are no easy answers to the enormous amount of variability in how offenders and the public perceive incarceration, the failings of our current system suggest that leniency is not just defensible, but good. Any overarching goal, such as leniency, should be sought through the proper set of secondary principles, and proportionality and certainty lay claim to being principles agreeable to a variety of persons with competing moral intuitions.

References Aborn, Richard M. and Cannon, Ashley D. (2016). “Prisons: In Jail, But Not Sentenced.” Available at www.americasquarterly.org/aborn-prisons, accessed December 1, 2016. Alexander, Michelle (2012). The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York: The New Press. Alliance for Safety and Justice (n.d.). “Crime Survivors Speak: The First-Ever National Survey of Victims’ Views on Safety and Justice.” Available at www. allianceforsafetyandjustice.org/wp-content/uploads/documents/Crime%20Survi vors%20Speak%20Report.pdf, accessed July 2, 2016. American Probation and Parole Association (n.d.). “Administrative Responses in Probation and Parole Supervision: A Research Memo.” Available at www.appa-net. org/eWeb/Resources/SPSP/Research-Memo.pdf, accessed December 1, 2016. Beaumont, Gustave de and Tocqueville, Alexis de (1833). On the Penitentiary System in the United States and its Application in France; with an Appendix on Penal Colonies, and also Statistical Notes. Philadelphia, PA: Carey, Lea & Blanchard. Beaumont, Gustave de and Tocqueville, Alexis de (2011). Alexis de Tocqueville and Gustave de Beaumont in America: Their Friendship and Their Travels, trans. Arthur Goldhammer. Charlottesville, VA: University of Virginia Press. Beccaria, Cesare (1995). On Crimes and Punishments and Other Writings, trans. Richard Davies. Cambridge: Cambridge University Press. Bellamy, Richard (1995). “Introduction.” In On Crimes and Punishments and Other Writings, ed. Richard Bellamy, trans. Richard Davies. Cambridge: Cambridge University Press. Benko, Jessica (2015). “The Radical Humaneness of Norway’s Halden Prison.” New York Times. Available at www.nytimes.com/2015/03/29/magazine/the-radicalhumaneness-of-norways-halden-prison.html, accessed January 3, 2016. Besserer, Sandra (n.d.). “Criminal Victimization: An International Perspective. Results of the 1996 International Crime Victimization Survey.” Juristat: Canadian Center for Justice Statistics. Available at http://publications.gc.ca/Collection-R/ Statcan/85-002-XIE/0069885-002-XIE.pdf, accessed December 1, 2016. Bibas, Stephanos (2010). “The Need for Prosecutorial Discretion.” Temple Political and Civil Rights Law Review 19: 369–75. Binelli, Mark (2015). “Inside America’s Toughest Federal Prison.” The New York Times Magazine. Available at www.nytimes.com/2015/03/29/magazine/ inside-americas-toughest-federal-prison.html?_r=0, accessed December 1, 2016. Blackstone, William (1979). Commentaries on the Laws of England: A Facsimile of the First Edition of 1765–9, 4 vols. Chicago and London: University of Chicago Press. Blecker, Robert (2013). The Death of Punishment: Searching for Justice Among the Worst of the Worst. New York: St. Martin’s Press.

Certainty and Proportionality  85 Cohen, Thomas H. and Reaves, Brian A. (2007). “Pretrial Release of Felony Defendants in State Courts.” Bureau of Justice Statistics: Special Report. Available at www.bjs.gov/content/pub/pdf/prfdsc.pdf, accessed December 1, 2016. Crimmins, James (2011). “The Principles of Utilitarian Penal Law in Beccaria, Bentham, and J.S. Mill.” In The Philosophy of Punishment and the History of Political Thought, ed. P. Koritansky. Columbia, MO: University of Missouri Press. Department of Justice (2016). “Frequently Asked questions Concerning Executive Clemency.” Available at www.justice.gov/pardon/frequently-asked-questions-concerningexecutive-clemency, accessed July 3, 2016. Farrington, David P., Langan, Patrick A. and Tonry, Michael (eds.) (2004). “Cross-National Studies in Crime and Justice.” Available at www.bjs.gov/ content/pub/pdf/cnscj.pdf, accessed May 3, 2016. Foucault, Michel (2015). The Punitive Society: Lectures at the Collège de France, 1972–3, trans. Graham Burchell. Basingstoke, Hampshire and New York: Palgrave Macmillan. Friedman, Milton and Friedman, Rose (1980). Free to Choose: A Personal Statement. New York: Houghton Mifflin. Gao, George (2014). “Chart of the Week: The Black-White Gap in Incarceration Rates.” Pew Research Center. Available at www.pewresearch.org/fact-tank/2014/07/18/chart-ofthe-week-the-black-white-gap-in-incarceration-rates/, accessed December 1, 2016. Hartney, Christopher and Marchionna, Susan (2009). “Attitudes of US Voters toward Nonserious Offenders and Alternatives to Incarceration.” National Council on Crime and Delinquency, 1–10. Available at www.nccdglobal.org/sites/default/ files/publication_pdf/focus-voter-attitudes.pdf, accessed December 1, 2016. Hawken, Angela (2014). “Swift and Certain Sanctions in Community Supervision.” National Criminal Justice Association. Available at www.ncja.org/sites/default/ files/documents/evidence-behind-swift-webinar-slides.pdf, accessed December 1, 2016. Hawken, Angela and Kleiman, Mark (2009). “Managing Drug Involved Probationers with Swift and Certain Sanctions: Evaluating Hawaii’s HOPE.” National Institute of Justice, 1–66. Available at www.ncjrs.gov/pdffiles1/nij/grants/229023. pdf, accessed December 1, 2016. Hawken, Angela and Kleiman, Mark (2014). “Celerity and Deterrence.” In Deterrence, Choice, and Crime: Contemporary Perspectives, ed. Frank Cullen, Cheryl Lero Jonson and Daniel Nagin. New Brunswick, NJ: Transaction. Henrichson, Christian and Delaney, Ruth (2012). “The Price of Prisons: What Incarceration Costs Taxpayers.” Vera Institute of Justice. Available at www.vera. org/sites/default/files/resources/downloads/Price_of_Prisons_updated_version_072 512.pdf, accessed December 1, 2016. Human Rights Watch (2013). “An Offer You Can’t Refuse: How US Federal Prosecutors Force Drug Defendants to Plead Guilty.” Available at www.hrw.org/ report/2013/12/05/offer-you-cant-refuse/how-us-federal-prosecutors-force-drugdefendants-plead, accessed August 13, 2016. Ingraham, Christopher (2016). “Even Violent Crime Victims Say Our Prisons Are Making Crime Worse.” Washington Post Wonkblog. Available at www.washingtonpost. com/news/wonk/wp/2016/08/05/even-violent-crime-victims-say-our-prisons-aremaking-crime-worse/, accessed August 15, 2016. International Crime Victimization Survey (1996). “Appendix Four.” 178–221. Available at www.unicri.it/services/library_documentation/publications/icvs/statistics/17icvs-app4.pdf, accessed August 14, 2016.

86  Chris Barker Kansas Sentencing Commission. “HB 2170—Frequently Asked Questions.” Available at www.sentencing.ks.gov/docs/default-source/2013-forms/commission-releaseshb-2170-faq.pdf?sfvrsn=0, accessed August 14, 2016. Kant, Immanuel (1991). “Metaphysics of Morals.” In Political Writings, trans. H.B. Nisbet. Cambridge: Cambridge University Press. Kelling, George (2015). “Don’t Blame My ‘Broken Windows’ Theory for Poor Policing.” Politico Magazine. Available at www.politico.com/magazine/story/2015/08/ broken-windows-theory-poor-policing-ferguson-kelling-121268, accessed November1, 2016. Kelling, George L. and Wilson, James Q. (1982). “Broken Windows: The Police and Neighborhood Safety.” The Atlantic. Available at www.theatlantic.com/magazine/archive/1982/03/broken-windows/304465/, accessed November 1, 2016. Kugler, Matthew B., Funk, Friederike, Braun, Judith, Gollwitzer, Mario and Kay, Aaron C. (2013). “Differences in Punitiveness Across Three Cultures: A Test of American Exceptionalism in Justice Attitudes.” Journal of Criminal Law and Criminology 103: 1071–114. Lambe, Joe (2010). “Kansas Reviews Options to Cut Inmate Numbers.” The Wichita Eagle. Available at www.kansas.com/news/local/crime/article1045438. html#storylink=cpy, accessed December 1, 2016. Lappi-Seppälä, Tapio (2015). “Why Some Countries Cope with Lesser Use of Imprisonment? Explaining Differences and Pondering the Remedies.” Criminal Justice Alliance. Available at http://criminaljusticealliance.org/wp-content/uploads/2015/03/ Tapio-Lappi-Sepp%C3%A4l%C3%A4-presentation.pdf, accessed August 27, 2016. Larson, Doran (2013). “Why Scandinavian Prisons Are Superior.” The Atlantic. Available at www.theatlantic.com/international/archive/2013/09/why-scandinavianprisons-are-superior/279949/, accessed October 3, 2016. Lieber, Francis (1833a). “Penitentiary System of Pennsylvania.” In On the Penitentiary System in the United States and Its Application in France; with an Appendix on Penal Colonies, and also Statistical Notes, ed. Gustave de Beaumont and Alexis de Tocqueville. Philadelphia, PA: Carey, Lea & Blanchard. Lieber, Francis (1833b). “Preface and Introduction of the Translator.” In On the Penitentiary System in the United States and Its Application in France; with an Appendix on Penal Colonies, and also Statistical Notes, ed. Gustave de Beaumont and Alexis de Tocqueville. Philadelphia, PA: Carey, Lea & Blanchard. Mill, James (1992). “Prisons and Prison Discipline.” In Political Writings, ed. Terence Ball. Cambridge: Cambridge University Press. Mill, John Stuart (1969a). “Bentham.” In Essays on Ethics, Religion and Society, ed. J. M. Robson. Toronto and London: University of Toronto Press and Routledge & Kegan Paul. Mill, John Stuart (1969b). “Blakey’s History of Moral Science.” In Essays on Ethics, Religion and Society, ed. J. M. Robson. Toronto and London: University of Toronto Press and Routledge & Kegan Paul. Nugent-Borakove, M. Elaine and Budzilowicz, Lisa M. (2007). “Do Lower Conviction Rates Mean Prosecutors’ Offices Are Performing Poorly?” American Prosecutors Research Institute. Available at www.ndaa.org/pdf/do_lower_ conviction_rates_07.pdf, accessed November 27, 2016. Offices of the United States Attorney. “9-27.300—Selecting Charges—Charging Most Serious Offenses.” Available at www.justice.gov/usam/usam-9-27000-principlesfederal-prosecution#9-27.300, accessed August 2, 2016.

Certainty and Proportionality  87 “Oral Arguments.” (2014). Yates v. US. Available at www.supremecourt.gov/oral_argu ments/argument_transcripts/13-7451_4gd5.pdf, accessed November 29, 2016. Paternoster, Raymond (2010). “How Much Do We Really Know About Criminal Deterrence?” Journal of Criminal Law and Criminology 100: 765–823. PEW Charitable Trust (2012). “Public Opinion on Sentencing and Corrections Policy in America.” Available at www.pewtrusts.org/~/media/assets/2012/03/30/ pew_nationalsurveyresearchpaper_final.pdf, accessed November 29, 2016. Pincoffs, Edmund L. (1966). The Rationale of Legal Punishment. New York: Humanities Press. Potter, Nelson T., Jr. (1998). “The Principle of Punishment Is a Categorical Imperative.” In Autonomy and Community: Readings in Contemporary Kantian Social Philosophy, ed. J. Kneller and S. Axinn. Albany, NY: State University of New York Press. Rakoff, Jed (2014). “Why Innocent People Plead Guilty.” The New York Review of Books. Available at www.nybooks.com/articles/2014/11/20/why-innocentpeople-plead-guilty/, accessed March 3, 2015. Ramseyer, Mark, Rasmusen, Eric and Raghav, Manu (2009). “Convictions Versus Conviction Rates: The Prosecutor’s Choice.” The American Law and Economics Review 11: 47–78. Reiter, Keramet, Sexton, Lori and Sumner, Jennifer (2016). “Denmark Doesn’t Treat Its Prisoners Like Prisoners—and It’s Good for Everyone.” Washington Post. Available at https://www.washingtonpost.com/posteverything/wp/2016/02/ 02/denmark-doesnt-treat-its-prisoners-like-prisoners-and-its-good-for-every one/?utm_term=.172f6145f970, accessed April 2, 2016. Rich, William J. (2002). “Prison Conditions and Criminal Sentencing in Kansas: A Public Policy Dialogue.” Kansas Journal of Law and Public Policy 11: 693–721. Robbins, Tom (2015). “A Brutal Beating Wakes Attica’s Ghosts: A Prison, Infamous for Bloodshed, Faces a Reckoning as Guards Go on Trial.” The New York Times. Available at www.nytimes.com/2015/03/01/nyregion/attica-prisoninfamous-for-bloodshed-faces-a-reckoning-as-guards-go-on-trial.html, accessed January 23, 2016. Rosenberg, Carol (2011). “Inside the Convicts’ Cellblock Where War Criminals Stay at Guantánamo.” Miami Herald. Available at www.miamiherald.com/news/ nation-world/world/americas/guantanamo/article1937810.html, accessed April 28, 2017. Sharpe, J.A. (1990). Judicial Punishment in England. London and Boston, MA: Faber and Faber. Shuster, Arthur (2016). Punishment and the History of Political Philosophy: From Classical Republicanism to the Crisis of Modern Criminal Justice. Toronto, Buffalo and London: University of Toronto Press. Sieh, Edward W. (1989). “Less Eligibility: The Upper Limits of Penal Policy.” Criminal Justice Policy Review 3: 159–83. “Statistical Information Packet: Fiscal Year 2015, District of Kansas.” United States Sentencing Commission. Available at www.ussc.gov/sites/default/files/pdf/ research-and-publications/federal-sentencing-statistics/state-district-circuit/2015/ ks15.pdf, accessed July 3, 2016. Subramanian, Ram, Delaney, Ruth, Roberts, Stephen, Fishman, Nancy and McGarry, Peggy (2015). “Incarceration’s Front Door: The Misuse of Jails in America.” Vera Institute of Justice. Available at http://archive.vera.org/sites/default/files/resources/ downloads/incarcerations-front-door-report_02.pdf, accessed December 1, 2016.

88  Chris Barker Turner, Jennifer and Dakwar, Jamil (2014). “Written Submission of the American Civil Liberties Union on Racial Disparities in Sentencing.” American Civil Liberties Union. Available at www.aclu.org/sites/default/files/assets/141027_iachr_ racial_disparities_aclu_submission_0.pdf, accessed December 1, 2016. Wagner, Peter and Rabuy, Bernadette (2016). “Mass Incarceration: The Whole Pie 2016.” Prison Policy Initiative. Available at www.prisonpolicy.org/reports/ pie2016.html, accessed August 18, 2016. Whitman, James (2003). Harsh Justice: Criminal Punishment and the Widening Divide Between America and Europe. Oxford and New York: Oxford University Press.

5 Imprisonment and the Right to Freedom of Movement Robert C. Hughes

In American politics, it is common for liberals (and not uncommon for conservatives) to advocate imprisoning fewer offenders.1 Advocates of penal reform have given many reasons to replace prison with alternative sentences for nonviolent crimes, in particular: among them, that imprisonment of nonviolent offenders is a poor use of public resources, compared either with alternative punishments or with non-punitive ways of preventing crime (King 2005; Stemen 2007); that the experience of prison, at least in the United States, often has the opposite of a rehabilitative effect (Dolovich 2009); and that there are racial disparities in the rates at which nonviolent offenders receive sentences of imprisonment (Blumstein 1993). Though advocates of penal reform have given many reasons our government should not imprison nonviolent offenders, they typically have not questioned government’s general entitlement to impose sentences of imprisonment for a wide range of offenses. Most of the standard arguments for penal reform object to incarceration not in the abstract, but in the context of contingent facts about the availability of alternative sentences, the effects of incarceration, and the ways in which the criminal justice system responds unfairly to class and race. By contrast, some political philosophers have questioned government’s entitlement to impose sentences of imprisonment. These challenges have questioned the extent of government’s entitlement to impose punishment of any form. Some argue, for instance, that punishment is only justified under specific social circumstances which may not obtain in societies like ours (Hanna 2009). Others argue that government has little or no legitimate authority to impose obligations that go beyond natural duties, and that government therefore cannot legitimately punish most mala prohibita (Nozick 1974; Simmons 1995). These broad challenges to government’s authority to punish apply to all forms of punishment, not only to imprisonment. I shall argue that government’s use of imprisonment raises distinctive moral issues. Even if government has broad authority to make and to enforce law, government may not be entitled to use imprisonment as a punishment for all the criminal laws it is entitled to make. Indeed, there may be some serious crimes that it is wrong to punish with imprisonment, even if the

90  Robert C. Hughes conditions of imprisonment are humane and even if no adequate alternative punishments are available. The moral problem with imprisonment is not that it is more severe than alternatives. Imprisonment is scalable in severity. It can be long or brief, and it can be highly restrictive or only moderately so. A reasonable person could regard a fine of half his life savings as a more severe punishment than a one-year sentence to be served in Norway’s Bastoy Prison. Though prisoners on Bastoy Island all have jobs, they can fish and play tennis in their spare time, they have the keys to their rooms, and they have “views of the ocean that are postcard-worthy.” (Sutter 2012) Imprisonment is distinctive in the way that it affects human agency. Even the most lenient forms of imprisonment deprive prisoners of freedom of movement. The loss of freedom of movement makes entire categories of human activity unavailable in a way that a financial penalty, even a large one, does not. The effects of imprisonment on human agency thus differ in kind, not only in degree, from the effects of sanctions such as fines and mandatory community service. Thus, imprisonment requires special justification, over and above the justification of punishment. The only morally acceptable rationales for incarceration as a punishment, I argue, are moral education and collective self-defense. Though retribution may justify punishment of some form, it cannot justify incarceration. Moral education justifies imprisonment only if there is evidence that imprisonment does, in fact, morally educate. If such evidence is not forthcoming, the only possible justification for imprisonment in law enforcement is as a form of collective defense against wrongful attacks. Government may be entitled to imprison people (with due process) to address crimes that attack the necessary conditions for citizens’ effective exercise of agency. These include crimes that attack citizens’ agency directly, such as crimes of violence. These also include crimes that threaten the very existence of the state (e.g., treason), since effective exercise of human agency requires at least a minimally effective government. But there is a wide range of criminal prohibitions that we are justified in having but that the government is not morally entitled to enforce via imprisonment. These include many laws against minor offenses, such as disorderly conduct, but they also include some laws against offenses that are quite serious. Notably, they include prohibitions on white collar crimes such as insider trading that make the market less fair or that do small harms to large numbers of people but that do not seriously undermine anyone’s effectiveness as an agent.2 The core of my argument is that if we fully understand why it is wrong for private individuals to engage in kidnapping, we will see that the moral objection to kidnapping applies more widely. Imprisonment by the state is often (but not always) wrong in the same way as kidnapping. The account of the wrong of kidnapping I offer will build on some of the core claims of Kantian ethics.3 I make no apologies for relying on a popular but controversial theoretical framework. Others have defended this framework, and I will not attempt to motivate it here. Those who are unsympathetic to

Imprisonment, Right to Freedom of Movement  91 Kantian ethical theory likely will not find my argument congenial. Nonetheless, consequentialists may be pleasantly surprised by the conclusion that on the Kantian approach, empirical facts about the effects of imprisonment are relevant to its permissibility.

I. The Value of Freedom of Movement I shall begin by arguing that imprisonment interferes in a distinctive way with the exercise of rational agency. All penalties interfere with some of the things that the penalized person wants or could want. Fines, for instance, reduce people’s ability to do things they want by reducing their financial resources. If imprisonment is distinctively problematic, the problem cannot be simply that imprisonment keeps people from doing some of the things they want. At the same time, it would be absurd to claim that imprisonment keeps people from exercising agency at all. Obviously, prisoners do things. Imprisonment of any form is problematic because it entirely deprives people of one of the central means of human agency. Imprisonment takes many forms. Though many prisons are extremely unpleasant places, there are some prisons (like Bastoy) that people could reasonably find pleasant. Many prisons regulate inmates’ activities tightly, but others do not. Indeed, people can count as prisoners even if authorities disclaim any right to regulate how prisoners spend their time. The only right all prisoners are denied is the right to leave the space to which they are confined. Not all limitations on freedom of movement are normally characterized as imprisonment. I shall use the term “imprisonment” to refer to involuntary confinement that physically separates people from the rest of society. The word “involuntary” here is crucial. People who voluntarily travel on ocean vessels or space ships are confined to spaces that are physically separated from the rest of society, but their voluntary, temporary confinement enhances their agency by helping them to go where they want. Some involuntary limitations on travel, such as house arrest and the use of ankle monitors, do not physically separate people from society. I leave open the question whether these restrictions on movement raise the same ethical concerns as imprisonment. The loss of freedom of movement involved in imprisonment interferes with the exercise of agency in a distinctive way. Locomotion is a primary means of agency for human beings. It is required for any project that requires observing the world personally (i.e., not via others’ reports) beyond the tiny corner of the world an individual can perceive at a given moment. It is also required for any project that involves physically altering the world by one’s own efforts (i.e., not by directing others’ efforts) beyond the tiny corner of the world one can reach without moving the whole of one’s body. Perhaps most importantly, freedom of movement is a prerequisite for effective freedom of association, and thus for the ability to select and to pursue projects that require others’ help. The problem here is not so much with compelled

92  Robert C. Hughes association—though that is a feature of many forms of imprisonment—but with the inability to meet in person with people of one’s own choosing. Even if an imprisoned person has unlimited and unmonitored access to mail, telephone, and electronic communication, and even if visits are allowed without restriction, the prisoner cannot meet in person with anyone who is unable to make the trip, nor can the prisoner attend meetings of associations. Many human projects are either impossible or seriously hampered by the inability to meet in person with other participants. Freedom of movement, then, is one of the basic, all-purpose means of human agency. All-purpose means of human agency are means without which major categories of human activity are foreclosed. Among the other all-purpose means of human agency are the ability to possess and to use physical objects, the ability to solicit the help of others, and the ability to draw on our own skills and talents.4 Though a person who has lost one of these all-purpose means of agency can still live and act, her agency is greatly impaired. Of course, the loss of a specific resource, such as a piece of property or a sum of money, can impair someone’s agency. But the complete loss of access to one of the basic all-purpose means is an impairment of a different kind. Being imprisoned is less like being deprived of a specific sum of money and more like being prohibited from using money at all. One might object that a loss of a specific resource—for instance, a large sum of money—might impair someone’s agency more than the loss of freedom of movement. Admittedly, for some people, losing a large sum of money may interfere with their ability to achieve the ends they currently have more than losing freedom of movement would. Imagine a solitary miser who is uninterested in travel and whose chief pleasures are reading, writing, drawing, and contemplating her own wealth. For her to be incarcerated at a prison with a good library and an adequate supply of paper and pencils would frustrate her present desires less than would the loss of a large part of her financial resources. If she is confined more than briefly, however, she will be greatly limited in her ability to revise her ends.5 She cannot pursue projects that require any form of travel, and she cannot pursue projects that require forms of association that her confinement prevents. By contrast, if she loses a large sum of money, her present desires may be frustrated, but the whole category of projects that involve money does not become off-limits to her. Imprisonment thus interferes with prisoners’ effective agency in a serious way, a way that differs in kind from merely preventing people from getting things that they want.

II. The Duty to Respect Agency and the Wrong of Kidnapping So imprisonment affects agency in distinctive ways. But surely sometimes we may permissibly restrict people’s agency. The question is when. To determine when it is morally acceptable for a government to restrict people’s

Imprisonment, Right to Freedom of Movement  93 agency, I will first address a different question: why is kidnapping typically wrong? I will offer a Kantian answer to this question. If this Kantian account of the wrong of kidnapping is sound, it implies that there is a more extensive duty to respect freedom of movement. My Kantian account of the wrong of kidnapping will build on Barbara Herman’s Kantian account of the duty to refrain from intentional killing (Herman 1989). Consider first cases in which one person kills another merely to satisfy a desire, such as a desire to get an inheritance sooner. Such killings are obviously wrong. What explains the wrongness of such killings? According to the Kantian ethical framework, killing someone merely to satisfy a desire is permissible only if people could rationally endorse the universal acceptance of ethical norms that permitted such an action. Whether people can rationally endorse universal acceptance of norms depends on whether endorsing universal acceptance of these norms is consistent with the commitments that reason itself demands. (It is not relevant whether the acceptance of these norms would satisfy people’s actual desires and preferences.) If I endorsed ethical norms that permitted people to kill whenever they take killing to advance their own interests, then I might thereby endorse someone killing me, since someone could take it to be in their interest to kill me. As a rational agent, if I will anything at all, I am rationally committed to my continued existence as a rational agent. So if I will anything at all, I cannot rationally endorse ethical norms that permit people to destroy my agency, that is, to kill me, whenever doing so would help them to satisfy a mere desire. A parallel argument explains why it is wrong to kidnap someone merely to satisfy a desire. It is obviously wrong to kidnap someone to obtain a ransom or to prevent the victim from closing a deal with a business competitor. What makes such kidnappings wrong, on the Kantian account? These acts are permissible only if one could coherently endorse universal norms that authorized such actions. If I endorsed ethical norms that permitted people to kidnap whenever they took kidnapping to advance their own interests, then I might thereby endorse someone kidnapping me. Can I do so coherently? Kidnapping is unlike killing in that it does not destroy the agency of the victim; it need not involve even a threat of bodily harm. So a commitment to my continued existence as a rational agent is at least superficially consistent with endorsing others kidnapping me. To explain why kidnapping is wrong, on this Kantian framework, it is necessary to explain why people are rationally committed to having freedom of movement. Human beings are rationally and morally committed to preserving their freedom of movement because people are morally required to prepare to fulfill future moral requirements. These include universal moral duties (such as the duty to benefit people in need and the duty to preserve oneself) as well as the obligations people have acquired (such as promissory obligations and the obligations parents have to care for their children). Some things we need to satisfy moral requirements are predictable. To preserve oneself,

94  Robert C. Hughes one needs food; to care for one’s children, one needs food for them. Other needs are unpredictable. It is not always possible to predict what one will have to do to get food for oneself and one’s children. It is often impossible to predict what one will have to do to be a good parent, a good friend, or a good citizen. To be prepared to fulfill unpredictable future moral requirements, we need resources that can be used for many possible purposes. We need broadly applicable skills and talents.6 We need the opportunity to ask others’ help.7 We need the opportunity to obtain property—i.e., physical resources that we are free to use without others’ interference. Because so many possible activities, tasks, and projects require the ability to move from one place to another, we need freedom of movement. Since rationality and morality require us to preserve our access to all-purpose means, every human being is rationally and morally committed to preserving their freedom of movement. This does not entail that one is morally required to maximize one’s available means. It is not clear what would even count as maximizing one’s available means, since increasing the availability of one means almost always involves missing out on another. More time studying language means less time studying mathematics; pursuing greater financial resources through more hours of paid labor involves giving up leisure time available for discretionary purposes. Rather, rationality and morality prohibit entirely giving up one of the all-purpose means of action when there is no compelling reason to do so. One can rationally and morally choose not to take the highest-paying job, but one cannot rationally or morally vow never to accept income or to own property. One can rationally commit never to go to a particular place, but one cannot rationally abandon freedom of movement in general. One should not entirely give up an all-purpose means of action because one needs these means available to be able to respond flexibly to unpredictable future duties and obligations. Because the all-purpose means of action are prerequisites for whole categories of human action, their availability contributes more to one’s ability to fulfill unpredictable future duties than any specific resource. Of course, the loss of a specific resource could interfere with one’s fulfillment of a known obligation. Losing a sum of money earmarked for a particular purpose might prevent one from fulfilling a promise, for instance. But the loss of a specific sum of money does not make whole categories of human action impossible, so a financial loss interferes less than does confinement with one’s ability to fulfill the unpredicted and unpredictable demands of morality. The rational and moral duty to preserve one’s all-purpose means explains the wrongness of most instances of kidnapping. I cannot rationally endorse someone else kidnapping me to fulfill a desire I have no duty to help fulfill. I thus cannot endorse ethical norms that permit people to kidnap others whenever doing so advances their interests. Thus, on the Kantian view of morality, kidnapping others to advance one’s interests is wrong.

Imprisonment, Right to Freedom of Movement  95

III. From Kidnapping to Imprisonment Of course, there are morally salient differences between kidnapping by private citizens and the imprisonment of accused and convicted criminals by the state. Nonetheless, the Kantian argument against kidnapping extends to some (but not all) uses of imprisonment by the state. This section will explain three ways in which the reasoning of the argument against kidnapping can be extended. First, the argument against kidnapping applies to agents of the state. One might think that the principles of just or legitimate state action differ from the ethical principles that apply to individual action. The state is entitled to do many things that individuals are not entitled to do. Why not think that the ethical principles regulating just imprisonment by the state differ from the ethical principles regarding kidnapping by private citizens, just as the ethical principles regulating just seizures of property by the state differ from the ethical principles regarding stealing? If the above arguments for duties to respect individual agency are correct, however, these duties apply to state actors. Nothing about these arguments limits their application to the action of private individuals. For the same reason it is wrong for an individual to kill in order to satisfy her own mere desire, it is wrong for a state actor to kill as a means of satisfying the mere desires of the many. Someone who endorses ethical norms authorizing agents of the state to kill as a means of satisfying public preferences might thereby endorse the destruction of her own agency. Likewise, it is wrong for the state to incarcerate innocent citizens as a means of satisfying public preferences. Someone who endorses ethical norms authorizing agents of the state to imprison citizens for the sake of satisfying public preferences might thereby endorse her own incarceration. Though the primary argument against kidnapping for the sake of preferencesatisfaction applies equally to private individuals and to the state, there are additional constraints on uses of imprisonment by private individuals. For example, the state can provide a form of due process that private citizens cannot. So it may turn out that there are some ends (other than preferencesatisfaction) that justify incarceration by the state but that do not justify private citizens in using force against each other. This raises the question whether the argument against kidnapping can be extended in a second way. Does the argument against confining people to satisfy private desires or public preferences also prohibit confining people to achieve better ends than mere preference-satisfaction? For the moment, consider only confinement of people not convicted or suspected of crimes. The Kantian explanation for the wrongness of kidnapping implies that it is morally acceptable to confine innocent people only if those people could rationally and morally endorse their own confinement. People can rationally and morally endorse their own confinement only if doing so would be consistent with their other duties. Since people have a duty to prepare to fulfill unpredictable future duties, and because freedom of movement is one of the

96  Robert C. Hughes central, all-purpose means of human action, people have a pro tanto duty not to endorse their own confinement. Normally, the fact that an act would bring about a good result does not, by itself, justify violating a pro tanto moral duty. Only a weightier or higher-priority moral duty can justify violating a pro tanto moral duty. It is thus wrong to imprison innocent people unless these people have an important moral duty that their confinement would help to fulfill. Quarantine of people with highly infectious and dangerous diseases is justified, on this account. People have a moral duty to avoid endangering others’ lives. If one has a highly infectious, dangerous disease, going into quarantine would help to fulfill this duty. By contrast, it is wrong knowingly to incarcerate an innocent person as a scapegoat in order to calm a restive population. Calming a restive population is a good thing, but individuals do not have a moral duty to submit to incarceration in order to calm a restive population. The duty to preserve one’s own effective agency and to prepare to fulfill future duties has more weight. On a Kantian picture of morality, a duty to address a grave threat to someone’s agency (e.g., a threat to life) is the only duty that can outweigh the duty to preserve the effectiveness of one’s own agency. Thus, imprisoning the innocent is justified only if it addresses a grave threat to someone’s agency. The third extension of the argument against kidnapping concerns respect for the agency of the guilty. Thus far, I have argued that it is wrong to confine innocent people as a means to any end other than protecting people’s agency. But when government imprisons convicts, thereby depriving them of one of the central means of human agency, typically it imprisons people who are not innocent. When is it permissible to undermine the effective agency of someone who has done a legal or moral wrong in the past? It is intuitively obvious that criminals retain some of their rights—including, at least, their right not to be kidnapped by private persons for reasons unrelated to their crimes (Quinn 1985). One might worry that the earlier Kantian reasoning is unable to explain this. According to that line of reasoning, I cannot endorse ethical norms that condone interference with the agency of the innocent, since I might thereby endorse interference with my own agency. Since my actions are in my own control, however, I can guarantee that I will always be an innocent person. If I endorse ethical norms that permit interference with the agency of the guilty, I would not thereby endorse interference with my own agency. But it would be hubris for me to think I can guarantee that I will never, at any point in the future, commit a wrong. Human beings are flawed. Only a conceited, deluded person could think that he could guarantee he will never do anything morally wrong. It would also be hubris to think one will never break the law. Perhaps someone could know that because of the way her character is constituted, there are specific crimes or types of wrongdoing that she is extremely unlikely ever to commit. But this knowledge would not license endorsing ethical norms that permit undermining the agency of

Imprisonment, Right to Freedom of Movement  97 people who commit those specific crimes or wrongs. The rational acceptability of ethical norms does not depend on knowledge of the peculiarities of one’s own character and situation. A sound system of ethical norms must be acceptable to all flawed human beings, not only to those who have certain flaws and lack others. As flawed rational agents, we are rationally committed to remaining effective agents and to being able to carry out our moral duties and obligations, even after having done a wrong. An agent’s rational commitment to the preservation of the effectiveness of her own agency does not depend on moral desert. All human rational agents are committed to preserving their freedom of movement, since any of us could, in the future, have duties that can only be fulfilled by moving freely. In thinking about what ethical norms are acceptable, then, we must consider the possibility that we will ourselves do wrong, and that we remain committed to the importance of our effective agency if we do. Properly understood, then, the Kantian argument against kidnapping implies that the mere fact that someone has committed a moral or legal wrong in the past has no bearing on the permissibility of confining that person. There is a separate question whether one may confine a person or threaten to confine a person to prevent that person from committing a wrong or a legal offense. This is a special case of the general question when defensive force is justified. It is easy to see why it is permissible to use force that stops a wrongful attack without doing anything to undermine the attacker’s future agency (e.g., blocking a punch). It is harder to see why it is permissible to use defensive force that has long-term effects on an attacker’s agency (e.g., killing or injuring an attacker in self-defense). Appealing to a distinction between intended and merely foreseen harms will not do. When someone kills in self-defense, the death of the attacker is usually the means of stopping the attack; it is not merely a foreseen, unintended side-effect of stopping the attack. Here is one possible explanation why defensive force is permissible even if it undermines the agency of the attacker in a way that goes beyond blocking the attack. When defensive force is proportional to the gravity of a wrongful attack—when impairment of the attacker’s agency is no greater than the impairment the attack would have done to the victim’s agency had it succeeded—defensive force shifts the costs of a moral failure from the victim to the wrongdoer. Flawed rational agents who know that they could become attackers, but who also know that they could be victims, have no objection to ethical norms that permit the costs of their own errors to be shifted onto them. Though everyone is rationally committed to valuing his own agency, no one is rationally or morally entitled to value his own agency more highly than the agency of others. So attackers and potential attackers have no objection to defensive force that is proportional to the gravity of a wrongful attack. Defensive force that undermines the attacker’s agency is justified only in response to an attack that could wrongly interfere with someone’s

98  Robert C. Hughes agency, however. If an action is wrongful in some other way, defensive force cannot be justified as shifting the harm to agency from the would-be victim to the attacker. For example, it is wrong to use force to stop actions that are disrespectful but not injurious.

IV. Justifying Imprisonment I have argued that if we understand the wrong of kidnapping correctly, we will see that there is a broader moral duty to respect agency which prohibits confining people against their will in many circumstances. The only exceptions to this prohibition on involuntary confinement involve the protection of agency. It may be acceptable to confine innocent people against their will to prevent serious harm to someone’s agency (e.g., death from disease). It may be acceptable to use or to threaten confinement as a means of defense against present or future wrongdoing—but only if the wrongdoing would threaten either people’s agency itself or their effective access to one of the all-purpose means of agency. This section will defend some moral limits on the use of incarceration that follow from these principles. To do so, it will examine the extent to which any of three most influential justifications of punishment—retribution, moral education, and deterrence—can overcome the moral presumption against involuntary confinement. The ethical principles concerning restriction of agency do not admit retribution as a justification of imprisonment. As flawed rational agents, we are all rationally committed to the preservation of our agency, and this commitment is not conditional on any assessment of the merits of our past actions. We are also rationally committed to the availability of all-purpose means, such as freedom of movement, to help us satisfy unpredictable future moral duties and obligations. We could not rationally consent to ethical norms that allowed people (including state actors) to undermine or to threaten other people’s agency as a backward-looking remedy for a past wrong. Nor could we rationally consent to ethical norms that allowed people to deprive others of all-purpose means of agency as a backward-looking remedy. Only the hope of protecting or promoting someone’s present or future agency or their access to all-purpose means can justify imprisonment in any form. This does not imply that there is anything wrong with retributive theories of punishment. It only implies that retributive theories of punishment cannot support incarceration as a punishment. One might try to argue that punishing people for their past wrongs is a way of respecting their agency. Kant’s own theory of punishment could be interpreted as taking this position. We respect a thief’s misguided exercise of agency by imposing on the thief the form of insecurity that the thief’s maxim of action would have imposed on everyone (Kant 1796: 474, Ak. 6:333). But it is not plausible to think both that we have a duty to respect criminals’ mistaken exercises of agency by imposing penalties and that this duty outweighs our duty not to prevent criminals from fulfilling future duties.

Imprisonment, Right to Freedom of Movement  99 Moral education theories of punishment, by contrast, may provide broad support for the use of imprisonment as a punishment. Imprisonment may be morally permissible when its use enhances prisoners’ effective rational agency overall, even though it temporarily deprives them of a central means of agency. Assuming that citizens are morally required to refrain from doing all the things that criminal statutes prohibit, citizens who violate criminal statutes display a defect in their rational agency. Imprisonment as a punishment could enhance these citizens’ rational agency if it helps them to understand the moral reasons to refrain from crime and to be motivated by these reasons in the future. Punishment could contribute to moral learning in two ways. In communicative models of moral education through punishment, the fact of punishment itself expresses a message with moral content to the convicted criminal (Hampton 1984; McTaggart 1896; Morris 1981).8 It could communicate, for instance, that we collectively reject the conduct for which the convict is being punished. In a treatment model of moral education, punishment is organized so that the experience of punishment contributes to moral learning. The experience of hard treatment could be intended to give convicts empathy for the suffering of their victims. Or convicts could be put in an environment that is designed to help develop habits of moral behavior. Moral education theories have an important limitation as a justification for imprisonment. A moral education theory of punishment only justifies imprisonment if there is positive evidence that the form of imprisonment used or threatened actually does educate. To justify a limitation of agency that would ordinarily be wrong, it is not enough to speculate that this limitation could lead to an overall enhancement of the restricted person’s agency. It is necessary to have empirical evidence that temporarily limiting a person’s agency is likely to have an educational effect. In assessing the educational effects of imprisonment, one must keep in mind that punishment succeeds at moral education only if the person being punished comes to recognize the moral reasons to behave better. If all that a punished person learns is that a certain behavior leads to aversive consequences, moral education has not taken place, and the agency of the punished person has not been enhanced. So evidence that imprisonment improves convicts’ behavior after release is not sufficient to show that prison is an effective form of moral education, since the improvement could be a result either of moral education or of specific deterrence. If evidence that imprisonment morally educates is not forthcoming, the only possible justification of imprisonment as a punishment would be collective defense. The principle of justified defensive force permits acts that undermine someone’s agency as a means of preventing that person from wrongfully undermining another person’s agency. Justified defensive force is limited to actions that are necessary to prevent wrongful acts and proportional to the harm to agency that the would-be wrongdoer threatens. For the defensive principle to authorize imprisonment as punishment, the

100  Robert C. Hughes explanation of how imprisonment contributes to the defense of agency would have to be indirect. Imprisonment as a punishment does not aim to prevent a convict from committing a specific wrong against a specific potential victim. One possible line of argument is that the threat of imprisonment serves to protect citizens generally against wrongful violence. Warren Quinn’s theory of punishment could be used to produce a justification along these lines for imprisonment of violent offenders (Quinn 1985).9 In many contexts, it is not possible to prevent wrongful violence by directly using force against the attacker, but it is possible to prevent wrongful violence by making a sincere and credible threat to harm the attacker in the future if the attack continues. If the attacker does not have a legitimate complaint against the immediate use of defensive violence, the attacker does not have a legitimate complaint against issuing a sincere threat of future harm that is no more serious than defensive harm that would be justifiable in the present moment. Threatening a term of imprisonment would be threatening a less serious harm than death or serious bodily injury, both of which are permissible defensive responses to attempted murder or mayhem. If it is not wrong for the legislature to make this threat credibly and sincerely to would-be violent criminals throughout the population, it is not wrong for government to carry out the threat against people who commit wrongful violence in defiance of the legislature’s threat. This account may be vulnerable to objections, either as an account of punishment in general or as an account of justified imprisonment. My aim here is not to defend Quinn’s account, but to point to two limitations that result from the account’s reliance on the defensive principle. First, imprisonment may only be used as a punishment for intentional violations of laws that are needed to protect either agency itself (e.g., protections of life and bodily integrity) or the availability of all-purpose means of agency (e.g., basic liberties, the opportunity to have property). These may include laws that are indirectly necessary to protect citizens’ agency. Since some of the all-purpose means of agency, such as property, only exist in a civil condition, imprisonment may be used to enforce laws that are strictly necessary to the operation of a minimally functional state, such as tax laws and laws against bribery, election fraud, and treason. But it would be a mistake to assert that the state’s ability to protect citizens’ effective agency requires general compliance with every justified regulation the state issues. Rational agency can flourish despite widespread disobedience of laws that are inessential to its protection. Collective defense accounts of punishment impose a further limitation on imprisonment in law enforcement: the use or threat of imprisonment to enforce a given law must be proportionate to the threat to agency that violations of this law pose. The threat to imprison people for serious crimes of violence would be proportionate to the threat that such crimes pose to individual victims’ agency. Whether imprisonment may be used to enforce property laws is less straightforward. Theft undermines one of the victim’s

Imprisonment, Right to Freedom of Movement  101 all-purpose means of agency, namely the opportunity to have property. One small theft, or one large theft from a well-off person, does not constitute a great enough interference with the victim’s property for imprisonment to be a proportionate response. Widespread theft, however, would seriously threaten everyone’s ability to control property. Perhaps it could even undermine a society’s economic system in a way that threatens everyone’s ability to sustain themselves. One might argue that because widespread theft, even widespread petty theft, would seriously threaten citizens’ effective agency, the legislature is justified in sincerely threatening everyone with incarceration if they engage in theft. Even if there is a sound argument for imprisonment as a punishment for theft, defense of citizens’ agency cannot justify the use of imprisonment to enforce every justified criminal prohibition. Some justified criminal laws protect only an aspect of one of the all-purpose means of agency. Consider laws against simple trespass. Though widespread violation of laws against simple trespass would undermine an aspect of one of the all-purpose means of agency, the ability to have and to control property, it would not deprive people of effective property rights altogether. Since imprisonment deprives people of freedom of movement entirely, the threat of imprisonment is not a proportionate defensive response to the crime of simple trespass. Other justified criminal laws do not protect people’s agency at all; they protect people’s welfare. For example, laws against disorderly conduct are justified, but since many forms of disorderly conduct do not threaten anyone’s agency—they only interfere with people’s welfare—laws against disorderly conduct should not be enforced using the threat of imprisonment. Though many of the laws that should not be enforced via imprisonment concern minor crimes, some laws should not be enforced via imprisonment even though they concern serious forms of wrongdoing. Consider laws against insider trading. Though there is debate about whether insider trading is wrong, there are reasons to regard it as a form of fraud or deception that is seriously wrong. On the “traditional theory,” insider trading by executives of the firm whose stock is traded violates a fiduciary duty to shareholders. United States v. O’Hagan, 521 US 642, 651 (1997). Another theory holds that insider trading involves a misappropriation of information and thereby violates a duty of trust owed to the source of information. Id. at 652. Still a third theory holds that insider trading involves deception of the investors with whom one is trading. If one has an informational advantage to which one is not entitled, it is deceptive to use this advantage in trading; one should either disclose what one knows before trading or refrain from trading on the information (Strudler and Orts 1999; Strudler 2009). Though insider trading is arguably seriously wrong, and though it does real harm to investors who are not insiders, the harm is typically diffuse. Rarely, if ever, do individual victims suffer losses great enough that the victim is altogether deprived of the effective right to possess financial resources. The harm that widespread insider trading would do is unlikely to undermine the economic

102  Robert C. Hughes system in a way that threatens people’s effective agency. Indeed, some have argued that the practice of insider trading can contribute to economic efficiency by spreading information about the appropriate valuation of securities (Ma and Sun 1998) or by providing an incentive for entrepreneurship (Manne 1966), though others doubt that insider trading promotes efficiency (Werhane 1989). So a collective defense theory of imprisonment cannot justify the threat of imprisonment as a punishment for insider trading. Insider trading may be a serious crime that deserves serious punishment, but moral desert alone cannot justify incarceration.

V. Conclusion Many governments use imprisonment to enforce a wide variety of criminal laws, but government’s entitlement to make criminal laws quite possibly exceeds its entitlement to use imprisonment as an enforcement mechanism. Imprisonment affects people’s agency in distinctive ways that are difficult to justify. The only possible justifications for imprisonment in law enforcement are moral education and collective defense. Collective defense may justify the use of imprisonment to punish violent crimes and other crimes that attack people’s ability to exercise agency effectively. It does not justify the use of imprisonment to enforce all justified criminal laws. Moral education theories of punishment may justify a broader use of imprisonment as punishment, but only if there is positive evidence (not mere speculation or hope) that imprisonment helps people to recognize the moral reasons to do what the law says. The question whether any actual system of incarceration has this effect I must leave to readers’ judgment.10

Notes 1 For discussion of the penal reform movement on the right, see Dagan and Teles 2012. 2 I do not count laws against drug possession and trafficking as examples because there is controversy about whether criminal law should regulate recreational drugs at all. 3 Though I use a Kantian ethical framework, I will not attempt either to interpret or to defend Kant’s opaque discussion of criminal punishment in the Doctrine of Right. 4 Barbara Herman identifies three central means of human agency: skills, things, and the help of others (Herman 1984: 585). I suggest here that movement, including free locomotion, should be added to this list. 5 In Rawlsian terms, loss of freedom limits a person’s ability to revise and pursue a conception of the good. John Rawls added freedom of movement to his list of primary goods, aptly but without explanation, in Political Liberalism (Rawls 2005: 181). 6 This is why Kant believes that it is immoral to neglect the development of one’s talents (Kant 1785: 74–5, Ak. 4:423). 7 This is why Kant believes that it is immoral to have a policy of never helping anyone (Kant 1785: 75, Ak. 4:423)

Imprisonment, Right to Freedom of Movement  103 8 None of these theories is intended to justify imprisonment in particular. Jean Hampton explicitly states that her theory “rejects many forms of incarceration used today.” (Hampton 1984: 228). 9 On Sharon Byrd’s interpretation of Kant’s cryptic remarks about punishment, Kant’s own theory of punishment has a somewhat similar structure (Byrd 1989). 10 I am grateful to Barbara Herman, Herbert Morris, Stephen Munzer, Seana Valentine Shiffrin, Alan Strudler, and the members of the UCLA Ethics Writing Seminar, the UCLA Legal Theory Workshop, the Wharton LGST junior faculty workshop, an audience at IRV 2015, and the workshop on incarceration at the University of New Orleans for helpful comments on previous drafts.

References Blumstein, Alfred (1993). “Racial Disproportionality of U.S. Prison Populations Revisited.” University of Colorado Law Review 64: 743–60. Byrd, B. Sharon (1989). “Kant’s Theory of Punishment: Deterrence in Its Threat, Retribution in Its Execution.” Law and Philosophy 8: 151–200. Dagan, David and Teles, Steven (2012). “The Conservative War on Prisons.” Washington Monthly, November/December. Dolovich, Sharon (2009). “Incarceration American-Style.” Harvard Law and Policy Review 3: 237–59. Hampton, Jean (1984). “The Moral Education Theory of Punishment.” Philosophy & Public Affairs 13: 208–38. Hanna, Nathan (2009). “Liberalism and the General Justifiability of Punishment.” Philosophical Studies 145: 325–49. Herman, Barbara (1984). “Mutual Aid and Respect for Persons.” Ethics 94: 577–602. Herman, Barbara (1989). “Murder and Mayhem.” The Monist 72: 411–31. Kant, Immanuel (1996a). The Doctrine of Right. Trans. Mary Gregor in Practical Philosophy. Cambridge: Cambridge University Press. Kant, Immanuel (1996b). Groundwork of the Metaphysics of Morals. Trans. M.J. Gregor in Practical Philosophy. Cambridge: Cambridge University Press. King, Ryan, Mauer, Marc, and Young, Malcolm (2005). “Incarceration and Crime: A Complex Relationship.” Sentencing Project. Available at www.sentencingpro ject.org/publications/incarceration-and-crime-a-complex-relationship/, accessed December 8, 2016 Ma, Yulong and Sun, Huey-Lian (1998). “Where Should the Line Be Drawn on Insider Trading Ethics?” Journal of Business Ethics 17: 67–75. Manne, Henry (1966). “In Defense of Insider Trading.” Harvard Business Review 44: 113–22. McTaggart, John (1896). “Hegel’s Theory of Punishment.” International Journal of Ethics 6: 479–502. Morris, Herbert (1981). “A Paternalistic Theory of Punishment.” American Philosophical Quarterly 18: 263–71. Nozick, Robert (1974). Anarchy, State, and Utopia. New York: Basic Books. Quinn, Warren (1985). “The Right to Threaten and the Right to Punish.” Philosophy & Public Affairs 14: 327–73. Rawls, John (2005). Political Liberalism. New York: Columbia University Press. Simmons, A. John (1995). On the Edge of Anarchy: Locke, Consent, and the Limits of Society. Princeton, NJ: Princeton University Press.

104  Robert C. Hughes Stemen, Don (2007). “Reconsidering Incarceration: New Directions for Reducing Crime.” Vera Institute of Justice. Available at www.vera.org/publications/reconsi dering-incarceration-new-directions-for-reducing-crime, accessed December 8, 2016 Strudler, Alan (2009). “The Moral Problem in Insider Trading.” In Oxford Handbook of Business Ethics, ed. George Brenkert and Tom Beauchamp. Oxford: Oxford University Press. Strudler, Alan and Orts, Eric (1999). “Moral Principle in the Law of Insider Trading.” Texas Law Review 78: 375–437. Sutter, John (2012). “Welcome to the World’s Nicest Prison.” CNN, May 24. Available at www.cnn.com/2012/05/24/world/europe/norway-prison-bastoy-nicest/, accessed December 10, 2016. United States v. O’Hagan, 521 U.S. 642, 651 (1997) Werhane, Patricia (1989). “The Ethics of Insider Trading.” Journal of Business Ethics 11: 841–5.

6 Are There Expressive Constraints on Incarceration? Bill Wringe

1. Setting the Scene Philosophers’ attempts to justify punishment have focused on a wide range of features that paradigmatic cases of punishment possess. The most obvious such features are, of course, punishment’s (alleged) capacity for incapacitating wrongdoers and deterring future wrongdoing (Bentham 1907 [1780]; Tadros 2011); and its (alleged) ability to exact from an offender a deserved quantum of suffering (Kant 1996 [1797]). Others include its supposed capacity to handicap the endeavors of those who do not regard themselves as bound by the constraints which bind other members of society (Dagger 1993, 2008), the provision of opportunities for offenders to repent and achieve a measure of moral growth (Duff 1986, 2001), and the feature of punishment with which I shall be most concerned in this chapter, its capacity to, as they say, ‘send a message’ (Feinberg 1965; Hampton 1992; Duff 1986, 2001). Those who defend theories of this sort might be expected to have views about the kinds of punishments which their views can and cannot justify. Some of these limitations will be internal to the theory put forward. So, for example, those who defend the institution of punishment on deterrence-based grounds can be expected to argue that their theories only justify punishments up to a level of severity at which trade-offs between deterrent effectiveness and the costs—both direct and indirect—of imposing further punishment moves from positive to negative (Bentham 1907). Other kinds of constraint will be extrinsic to the theory under considerations. For example, someone might hold that there are some forms of treatment which are so cruel and degrading, or so unequal in their effects on those on whom they are imposed, that whatever their deterrent effect or expressive power, we simply should not impose them.1 In this chapter I shall be concerned not just with one particular version of expressivism: the ‘denunciatory’ version of expressivism which I have defended in detail elsewhere (Wringe 2016; Wringe 2017) and in particular with the question of whether this theory generates internal constraints on the kinds of punishment that can justifiably be imposed and if so, what they

106  Bill Wringe are. I shall argue that there are. Indeed, the internal constraints on permissible forms of punishment which the theory generates are quite substantial: they would rule out the harshening and, in many cases, prescribe the lessening of the severity of many kinds of punishment which existing societies inflict on their citizens. Advocates of expressive accounts of punishment often claim that their views provide reasons for reducing the severity of penal sanctions. (See, for example, Duff 2001; Lippke 2007). Their views typically differ from mine in a crucial respect. Their arguments focus on the role that punishment plays or can play in communicating with a convicted offender. Those who have views of this sort are well-placed to argue that forms of punishment which undermine the offender’s status as a recipient of the appropriate form of communication, or their capacity to respond to that communication in an appropriate way, undermine their own putative justification. Duff, for example, regards punishment as being aimed at communicating with an offender in such a way as to elicit remorse and regret—with the eventual goal being an offender’s reintegration into a political community (Duff 2001, as discussed and criticized by Boonin 2008; Hanna 2008; Wringe 2013; Glasgow 2015; Wringe 2016, Wringe 2017). Such remorse and regret would be undermined by treating an offender in ways which undermine their sense of themselves as a moral agent, for regret of the right sort presupposes that one see oneself as such (Cochrane 2017). By contrast, on the view of punishment which I defend, the primary audience of the form of communication which punishment involves is not the convicted offender, but the political society whose laws have been infringed. One might worry that on this view, the severity of the kind of punishment that can be justified will be limited only by the limits—if any exist—of what will communicate effectively with society at large. Indeed the theory might even be thought to require an increase in the level of severity of punishment in societies which, plausibly, already punish some of their members too much. I shall argue that matters are less dire than this pessimistic line of thought suggests. Once we attend to the content, as well as the audience, of the messages which punishment ought to express, we find that there are grounds, internal to the theory, for thinking that there are limits to the severity of the kinds of punishment it can justify. Let’s see why.

2.  Expressivism—The Very Idea Why should we think punishment has an expressive dimension? In a well-known early discussion, Feinberg argues that we can distinguish between a general sense of punishment on which punishment involves “the infliction of hardship by an authority on a person for his prior failing in some respect,” and what he calls a “narrow and more emphatic sense,” which involves a form of symbolic reprobation which can in principle be absent

Expressive Constraints on Incarceration?  107 from what Feinberg calls “mere penalties” such as parking tickets (Feinberg 1965: 397–9). He then argues that legal punishment has a number of significant functions, which go beyond deterrence and retribution, and which would be impossible if it did not have a reprobatory aspect. These include the authoritative disavowal of wrongdoing; symbolic non-acquiescence in criminal acts; vindication of the law and absolution of those who are not involved in wrongdoing (Feinberg 1965: 405ff). Feinberg’s talk of the ‘functions’ of punishment raises many questions. We might ask what is entailed by the claim that a social institution has a certain function, and in virtue of what kinds of facts this is true. Does its truth, if it is true, depend on members of society having certain intentions with respect to punishment? If so, which members of society, and what kind of intentions? Or does Feinberg’s claim simply involve an attribution of a function in a sense which has been analyzed by philosophers of cognitive science, biology and medicine? If so, which of these analyses is the relevant one? Is Feinberg’s claim to be understood as one about the contributions penal institutions make to the functioning of society as a whole (Cummins 1975), or does it purport to explain the persistence of those institutions in ways which depend on the performance of these functions (Wright 1973)? These would be urgent questions if Feinberg were interpreted, as he sometimes has been, as appealing to these functions of punishment as ways in which the institution of punishment might be justified. However, Feinberg might be understood slightly differently—as making what we might call a hermeneutic claim. On this interpretation he might be taken as pointing out both that punishment is typically understood in certain ways, and that it can do so only if it involves a certain kind of symbolic act. This interpretation closes off the possibility of a gap between what punishment appears to us to be, and what it really is. If we are talking about actions which are expressive by their very nature, there is no room for a gap between what they are taken to mean by the right kind of interpreters and what they are typically taken to mean, any more than there is room for a gap between what people take the sense of a commonly used phrase to be and what its sense actually is. This account of Feinberg’s position leaves wide open questions about the justification of punishment: questions which have been pressed by David Boonin among others (Boonin 2008). Boonin argues that as a matter of conceptual necessity punishment involves intentionally harming others (Boonin 2008: 12–17). (This claim might be understood as a particular way of interpreting a very widely accepted claim, namely that punishment involves hard treatment.) He then asks how the fact that a certain kind of behavior involves a kind of reprobatory expression could make permissible forms of behavior which we would not otherwise regard as permissible—namely the inflicting of harm on others. Expressivist conceptions of punishment suggest a way of responding to this challenge. The expressivist should not attempt to argue that the expressive dimension of punishment makes permissible forms of behavior that

108  Bill Wringe would not otherwise be permissible. Instead, they should note that adopting an expressivist conception of punishment puts us in a position to challenge the characterization of punishment that Boonin gives. Rather than accepting that punishment involves the intentional inflicting of harm, the expressivist can claim that the requirement that punishment involve hard treatment is satisfied provided that penal institutions are of a sort which reliably cause suffering to those who are subjected to them (Wringe 2013). Of course there are ways of treating individuals which reliably cause suffering which we would not typically count as forms of punishment (for example, compulsory psychiatric treatment, pre-trial arrest, self-defense and so on). However, the expressivist can explain—in a way that other kinds of theorist cannot—why these are not forms of punishment: rather than appealing to the intentions with which suffering is inflicted, they can point out that these forms of punishment do not have the kind of expressive role characteristic of punishment (Wringe 2013: 866–8).

3.  Punishment’s Audience If this expressivist response to Boonin’s challenge to the legitimacy of punishment is to be made good, the expressivist needs to give a more detailed account of the expressive role of punishment. Without such an account we cannot assess whether it is plausible to think that forms of behavior which reliably cause suffering but which we do not typically think of as involving punishment lack the expressive dimension that is typical of punishment. We need, specifically, an account of what punishment expresses and who it expresses it to. First then: what is the audience that punishment expresses its message to? There seem to be three different kinds of answers one might give. In one view, which we might call ‘pure expressivism,’ the attitudes expressed in punishment need not be expressed to any audience in particular. To express an attitude may simply be to manifest possession of it, in the sense that we use when we talk of a disposition being manifested by its instances. Perhaps a form of behavior can only count as a manifestation of an attitude in this sense if it is in principle capable of being understood as such by some audience or other. But even if this is so, it need not follow that in expressing the attitude, we intend it to be understood by some audience in particular (Glasgow 2015; Bennett 2008). As Joshua Glasgow has noted, one advantage of a pure expressivist view is that it avoids a problem to which views that think of punishment as involving expression to some particular audience seem susceptible: namely the question of whether punishment ceases to be justifiable if the audience at which it is aimed is unreceptive—that is to say, if it is incapable of understanding the message that punishment is intended to send (Glasgow 2015). However, the pure expressivist view seems implausible. It is hard to see how the manifestation of an attitude to no audience could have a value

Expressive Constraints on Incarceration?  109 that the mere possession of that attitude would not. (One possible response might be that the possession of the attitude is in some sense constituted by its manifestation. It’s implausible to think that the attitudes of individuals are constituted by their manifestation—that way, behaviorism lies. Still we might think that punishment expresses the attitudes of some kind of plural subject or collective agent [Gilbert 2006]. It might be less implausible to adopt a behaviorist stance to attitudes of such subjects. Still, this is a big commitment for the pure expressivist to take on.) A further weakness of pure expressivism is that it appears to detach the expressivist position from the main considerations to which Feinberg appeals in making the case that punishment properly understood must have an expressive dimension: it is hard to make sense of the notions of authoritative disavowal of wrongdoing; symbolic non-acquiescence in criminal acts; vindication and absolution of those who are not involved in wrongdoing in terms which do not make reference to an audience to whom something is disavowed; in front of whom someone vindicated; and before whom one is absolved. Pure expressivism also seems to run into trouble when confronted with features of our expressive practice which seem essential to the operation of legal punishment but which is unrelated to pure expression. The point has been made most famously—with reference to the hard treatment of offenders (Feinberg 1965: 419–23). However, it can be as easily made by focusing on something else which seems essential to legal punishment: the fact that it takes place under the eye of the public. By this I mean: the fact that punishments are assigned in publicly accessible courtrooms; that a person is punished and the reasons for which they are punished is a matter of public record; that punishment affects an individual’s civic status in various respects; and so on. Many of these features of punishment increase the hardships which punishment entails. On the view that punishment involves a form of expression which is not directed at any particular audience, they are difficult to explain. (One might think that some of these features of punishment serve the purpose of protecting convicted offenders from abuse. However, it seems highly implausible that they do any such thing; and other features of our penal system make it hard to believe that this can be an aim which is any way prominent.) Antony Duff’s ‘communicative’ version of expressivism (Duff 1986, 2001; see also Falls 1987) might seem to avoid some of the difficulties which pure expressivism faces. For Duff punishment expresses a message to a convicted offender. The point of doing so, on his view, is to encourage (but not coerce) the offender to experience remorse or regret. Duff hopes to explain why punishment needs to involve hard treatment: we might think—though this is an empirical claim, which may have less empirical backing than we might hope for—that either the effective expression of the right kind of message or its expressive uptake will require that it be delivered via the medium of harsh treatment. However, as I have argued elsewhere Duff’s theory runs

110  Bill Wringe into the same difficulties that the pure expressivist faces when accounting for the publicity of punishment (Wringe 2017). No such problem arises for a third version of expressivism, which I call ‘denunciatory expressivism.’ On this view punishment expresses a message to a particular audience. The audience is, however, not the offender, but the members of the political community whose laws the convicted offender has broken. It is natural to take Feinberg to have been defending a view of this sort: when we ask the questions ‘to whom might we be disavowing wrongdoing; displaying symbolic non-acquiescence; vindicating the law; and in front of whom might we be granting absolution,’ the answer might well seem to be ‘society at large.’ Furthermore, on this view, the public nature of punishment ceases to be mysterious: we cannot hope to express a message to society unless we make it available to them to understand. We may also have an answer to the problem Feinberg raises for his own view as to why punishment should involve harsh treatment: the audience to whom punishment expresses a message may be one which will only take the message in the way in which it needs to be taken if it is expressed via the infliction of harsh treatment. And finally, while there might conceivably be societies which are ‘unreceptive’ in ways analogous to the ways that offenders can be unreceptive, it seems reasonable to hope that actually existing societies are not.

4. Does the Denunciatory Account Legitimate Excessive Punishment? However, advocates of a denunciatory account of punishment need to address an important objection. On a denunciatory account, the justifiability of punishment as a whole depends on its capacity to communicate messages about the offenses of convicted offenders to society at large. We might think it followed that the justifiability or otherwise of particular forms of punishment will depend to some extent or other on what the members of that society take the expressive significance of those forms of punishment to be. This gives rise to two kinds of problem. One has to do with relative levels of severity in punishment; the other with absolute levels.2 The problem about relative levels of punishment is that it seems as though in more punitively minded societies denunciatory considerations would justify harsher sentences for a given crime than would be justified in a society where such punitive attitudes were less widespread. Given the plausible empirical hypothesis that levels of punitiveness might vary from society to society, one might suspect the denunciatory theorist will end up arguing that in different societies different levels of punishment for the same crime will count as justified. The problem about absolute levels of punishment is that in some societies understandings of the significance of punishment might be so harsh as to justify forms of punishment which would seem excessive

Expressive Constraints on Incarceration?  111 by any reasonable standard (let alone, for example, the standards of 8th Amendment jurisprudence in the United States.) How serious is the first problem? Perhaps there is nothing especially surprising about the idea that the severity with which certain crimes are punished should vary from place to place, and nothing objectionable—but rather the reverse—about a theory which explained why this might be justifiable. Still we might think that while levels of punitive severity do vary from place to place, we can appropriately criticize the ways in which individuals are punished in one society by comparing their treatment with the treatment of similar offenders elsewhere.3 The denunciatory account runs the risk of ruling this kind of criticism inadmissible. However, the problem about absolute levels of severity seems even more serious. For we might think that there are a range of levels of severity with which any given crime might justifiably be punished, and that societies could justifiably set sentences for a given crime at any of the range of levels within that acceptable range for a variety of public policy-related reasons. Considerations about the ways in which punishments would be understood might then be one factor which could legitimately feature in the determination of appropriate levels of severity in punishment. By contrast, if the denunciatory theory turned out to require levels of punishment which were unacceptable for reasons which were independent of cross-society comparisons, no analogous response would be available. Could a denunciatory theory require punishments of a level that we might otherwise regard as unacceptably harsh? Popular responses to penal policy in various countries might make us pessimistic. There is considerable evidence that in many countries existing penal policies are seen as being too lenient; that proposals to lengthen sentences or to increase their severity are popular with electorates; and that by the same token proposals for sentence reduction tend to be electorally unpopular (Bottoms 1995; Pratt 2007). However, there is also some evidence that support for more severe sentencing policies often reflects mistaken views about what kinds of sentencing policy are in effect, and that the most spectacular instances of overly penal attitudes seem to reflect a considerable degree of misinformation about the levels of punishment prevalent within a given society (Dzur 2012). If so, then instead of providing reasons for increasing the severity of punishments imposed on convicted offenders, the denunciatory account might support a case for more effective programs of civic education in order to ensure that the messages punishment is intended to send are successfully conveyed to the audience the theory says they are directed to.4 Someone might think a denunciatory theorist of punishment cannot face a serious objection here. They might argue as follows. Either there are constraints on the way in which we can legitimately treat people when we punish them that exist for reasons which are independent of a denunciatory account of punishment or there are not. If there are not, then it cannot be true that the denunciatory theory legitimates forms of punishment which we

112  Bill Wringe have independent reason for taking to be unacceptable. But if there are—as most people would no doubt accept—then we already have an explanation of why certain forms of punishment which the denunciatory theory would appear to legitimate cannot in fact be regarded as legitimate. They are illegitimate precisely because they violate these independent constraints. However this line of argument seems unsatisfactory. The denunciatory theory of punishment is supposed to explain how some forms of punishment could be legitimate. Although we need not think that it needs not show that the forms of punishment prevalent in existing societies must be justified, we should nevertheless expect them to show that in such societies some form of punishment is justified. An account which appeals to external constraints on the ways in which we can treat people in order to explain why certain forms of apparently excessive punishment are not legitimate might not meet this constraint. For the minimum level of severity required by the denunciatory account might turn out to be greater than the maximum level of severity permitted by the external constraints.

5. Expressive Limits on Punishment (1) So far I have argued that we have reason to worry that a denunciatory theory of punishment might legitimate unacceptably severe forms of punishment. I have also argued that addressing this worry by appealing to the existence of external constraints on the acceptability of levels of punishment risks leaving us without an account of how any form of punishment could be justified. I shall now argue that there are significant internal constraints on the kinds of punishment which the denunciatory theory could legitimate. To see what these limits are we need to consider the content of the message we should take (justified) punishment to convey. Some expressive theorists of punishment have addressed this issue. Feinberg (1965: 403) writes that punishment involves “a kind of fusing of resentment and reprobation” where resentment is supposed to be a placeholder term for a variety of vengeful attitudes and reprobation involves a stern judgment of disapproval; and Nozick (1981), who finds little room for the idea that punishment incorporates any kind of vengeful attitude, develops the idea of a stern judgment of disapproval in such a way as to accommodate the thought that different degrees of punishment will presumably convey different messages and argues that the messages that punishment conveys will typically be of the form ‘this is how wrong you were.’ There is nonetheless a dearth of arguments for taking the expressive content of punishment to be one thing rather than another. Elsewhere I have argued that a denunciatory theorist of punishment should argue that the message expressed in punishment have the following form: “Offender X has committed an act of type T and in doing so they have acted wrongly” (Wringe 2016). I have also argued that in order to understand this form of communication in a way that does not endow the

Expressive Constraints on Incarceration?  113 state with an inappropriate and unjustifiable kind of moral authority, it is important to take it to be one which takes place between a collective agent and its members, and to take the particular kind of wrongdoing involved to be that of an individual’s reneging on their part in a collective commitment (Wringe 2016: Chapter 3). I shall not recapitulate my arguments for this formulation in detail here.5 Instead I shall argue that the idea that punishment is supposed to convey a content of precisely this sort places limits on the forms of harsh treatment which might be used to express it. I suggest that it gives rise to two non-trivial forms of constraint of precisely this sort. One arises from the fact that the punishment is intended to convey that a certain kind of action is wrong; the second form, the kind of wrongness which is intended to be conveyed. Let us start with the idea that the denunciatory content of punitive hard treatment is supposed to convey that the convicted offender has done something wrong. We might regard it as a condition of having done something wrong—as opposed to, say, unfortunate or destructive—that they be regarded as culpable. Whatever the conditions for culpability are, they must surely include being a responsible moral agent. If so, then forms of treatment which convey to society at large that individuals are not responsible moral agents would seem to be ruled out (or would at least require some form of moral justification). Could any conceivable forms of punishment convey such a message? It is possible to imagine some that might. A form of (supposed) punishment might, for example, require someone convicted of a crime of violence to wear a billboard with the message ‘I am a dangerous animal.’ So the denunciatory account seems to place at least some constraints on the forms that punishment properly so-called might take. But these constraints might seem relatively unimpressive. We might wonder whether the theory rules out any forms of treatment that are carried out with regularity in existing penal systems. We might think that some ways in which offenders are treated are designed to convey to offenders themselves that they are not fully responsible agents. Restrictions on relatively trivial matters of personal choice—such as clothing and hair-style—might count here. So, conceivably, might some of the conditions of confinement in so-called Supermax prisons (cf. Lippke 2004). However, this line of argument does not necessarily establish the conclusion we are searching for. Conveying something about an offender is a different matter from conveying it to society at large: indeed, the distinction I drew earlier between denunciatory and communicative forms of punishment depends on precisely this point. Furthermore, we might think that since many of the features of prison life which appear to convey these messages are not widely known or even widely publicized, they cannot count as intended to convey a message to society at large. We might also worry that insofar as this argument makes substantial claims on the limits that might be placed on justifiable conditions of confinement, it blurs the distinction

114  Bill Wringe between a form of treatment’s reliably having certain kinds of effects and its being intended to communicate something about an offender. The fact that Supermax prisons undermine individuals’ capacities for agency need not entail that imprisoning individuals under Supermax conditions communicates any kind of message that they are not agents. One response to the first of these points might be that the plausibility of a denunciatory theory of punishment should not depend on the public at large having a detailed knowledge of the exact treatment of particular offenders. The existence of widely available representations—in film, fiction, and the news—can be regarded as establishing that certain forms of treatment have a widely understood cultural meaning; and the fact that it is possible for members of the public to learn that offenders have been treated in a certain way might be thought to do enough to make sure that punishment send a particular message about those particular offenders to anyone who is alert to it. However, a different response seems more promising. Suppose we grant that details of the conditions of confinement of convicted offenders are not widely known beyond prison walls. There are, nevertheless, some parts of society to whom they are well known. Prison staff fall into this category; as do law enforcement officials. And so do other offenders. For, as I have already suggested, it is an important feature of punishment that can be justified by denunciatory methods that it convey the message that those who are punished are members of society. What about the worry that the argument put forward so far merely undermines—in a strictly causal sense—the agency of convicted offenders, without necessarily being intended to undermine—in any justificatory sense—a claim to the effect that they are agents? We might simply say that in practice certain kinds of incarcerative practice do both: they inhibit agency, and they inhibit our capacity to see those on whom they are inflicted as agents. Still, an objector might insist that this is no part of what such conditions are intended to convey. If we think, as Nozick did, that the content of penal communication depends on the right kind of Gricean communicative intention, the worry still stands (Nozick 1980; Grice 1957; Wringe 2016). Or we might say that although a sensitive genealogical analysis of the history of these forms of incarcerative practice shows them to bear the trace of intentions of the right (or wrong) sort, we can also imagine practices with a different history, but the same kinds of effects, and that it would be highly counterintuitive to think that a practice of this sort could fail to be objectionable if our actual practice is objectionable. Could there be other ways in which punishments which undermine an offender’s capacity for agency might be regarded as objectionable from the denunciatory point of view? Perhaps the existence of conditions of confinement which undermine a prisoner’s capacity for agency make it harder for the intended audience to grasp the message that punishment is supposed to convey. One weakness in this line of argument is that it seems to be

Expressive Constraints on Incarceration?  115 undermined by concessions already made. If we have accepted that a form of punishment may be justified by denunciatory means even if knowledge of its application is not especially widespread, then it is unclear why a justification of this sort should be undermined by the fact that the form of punishment chosen is not an especially effective means of conveying the relevant kind of message. Furthermore we might think that the ways in which inappropriate treatment of convicted offenders might undermine the effectiveness of the message that punishment is supposed to convey could be remedied by other measures. A more promising suggestion is that forms of punishment which undermine the agential capacities of convicted offenders are pragmatically self-defeating. In general, pragmatic self-defeat occurs when the way in which a message is conveyed to its intended audience undermines the credibility of that message. Consider someone who asserts, in perfect English, that they speak no English at all. From a logical point of view, there is nothing amiss with the content of their assertion: there is nothing in the existence of someone who speaks no English that contravenes the laws of logic. Nevertheless, that particular message cannot be conveyed in that particular way. Other examples can be given: consider someone asserting that they make no assertions at all; and so on. The denunciatory theorist might argue analogously, that treating someone in ways which undermine their capacities for agency involves pragmatic self-defeat of precisely this sort. One might worry that this argument proves too much. It might appear to rule out the possibility of any form of confinement being used as a form of punishment. For one might think that any form of confinement impinges on an individual’s capacity for agency in a negative manner. So we might think that the denunciatory theorist entails that, that no form of confinement can be an acceptable form of punishment. Some might welcome this corollary; but others will no doubt regard it as a reductio of the denunciatory view. However, I do not think it need be either. These cases of pragmatic self-defeat which I used to introduce the view suggest that pragmatic self-defeat is an all-or-nothing affair. However, it need not be. Consider someone who says in fluent, well-accented English, that they speak almost no English. The truth of what they say is not incompatible with their having said it. Nevertheless, their utterance seems closely analogous to the more obvious cases of pragmatic self-defeat we have already considered. We might understand agency undermining form of punitive practice in a similar way.

6. Expressive Limits on Punishment (2) The arguments of Section 5 suggested that the content of penal communication might place significant constraints on the kinds of punishment that might be acceptable. Now I wish to explore the idea, broached in Section 4, that further constraints stem from the fact that offenders are part of the audience for penal communications.

116  Bill Wringe We might think that the fact that the intended audience of penal communication includes the offender precludes the possibility of using expressive means which exclude the offender from membership of that community. Suppose this were correct. What, if anything, might it rule out? One obvious possibility is exile. Some might regard this as a historical curiosity (but see McGinnis, this volume). Nevertheless it seems significant that exile is a form of treatment which is ruled out as a form of punishment on conceptual, and not merely pragmatic, grounds. (Another example which shares some of the features of exile might be excommunication. We might regard this as a marginal case in a discussion of state punishment; but in the light of Duff’s suggestion that we understand punishment as involving a kind of “secular penance” it seems worth noting [Duff 2001:106–30].) We might think that the disenfranchisement of discharged felons is also ruled out. For it might seem that in disenfranchising members of a democratic society, we are both excluding them from full membership of society, and communicating to society at large that they are so disqualified. Some might argue that felon disenfranchisement is not a punishment, strictly so called, but rather a civil disqualification. However, it seems plausible that on the communicative view of punishment defended in Section 2, post-release disenfranchisement should be counted as a form punishment. It is less clear that the same can be said of the disenfranchisement of convicted criminals during a limited period of punitive imprisonment. For loss of the franchise during a period of temporary imprisonment would at best be expressive of temporary exclusion from a political community. So those who wish to argue that prisoners should have the right to vote should do so on other grounds. The requirement that punishment be addressed to an offender qua member of a political community might rule out considerably more than this. We might think that the requirement rules out very long periods of incarceration, including, but perhaps not limited to, lifelong incarceration. Those who are imprisoned for the rest of their lives might be regarded as being permanently excluded from the political community; and those who are imprisoned for very long terms might be regarded as effectively precluded from effective future membership of a political community, given what we know about the marginalizing effects of long-term confinement. We might also regard many of the standard conditions of confinement imposed on convicted offenders as excluding them from the political community. I have in mind here in particular restrictions on prisoners’ freedom to associate and to undertake paid work. Someone might suggest that any form of imprisonment worthy of the name must incorporate these features; but work by Richard Lippke (among others) suggests that we should regard these as merely contingent—and undesirable—features of imprisonment (Lippke 2007). (We might also note that to the extent that these cease to be features of imprisonment, the case for thinking that expressive constraints on imprisonment rule out long-term imprisonment will be weakened.)

Expressive Constraints on Incarceration?  117 Someone might wonder whether these conditions can be regarded as central to membership in a political community. They might do so on the basis of a relatively narrow understanding of the domain of the political, as encompassing only those kinds of activities which are covered in the ‘political’ sections of national newspapers: for example, voting in elections, participating in electoral campaigns, and so on. Still we might think that this draws the limits of the political too narrowly. A broader conception might start from Aristotle’s conception of human beings as ‘political animals,’ where the political might be understood as encompassing those activities which are characteristic of human beings living in self-governing groups. What is characteristic of such beings is—at least as Aristotle seems to conceive of it—a capacity for engaging in self-directed activity with a collective dimension (Schmid 2016). If we understand membership in a political community in this way then the range of potential expressive constraints on incarceration will be considerably expanded. But we might think that their strength will be decreased. Consider the mooted argument against felon disenfranchisement. Expanding the scope of the political seems to leave open the possibility that someone who was excluded from the vote might still be a member of the political community in a broader sense. This might be an argument against the broader view. But it need not be. It does suggest that a case against felon disenfranchisement might need to be made on other kinds of grounds. But a number of authors have explored just such possibilities (Cochrane 2017; Bennett 2008).

7.  Some Objections Considered One might wonder to what extent the view I have outlined here represents a distinctively denunciatory version of an expressive approach to the idea of limits on incarceration. Insofar as arguments for those limits stem from a conception of an offender as a member of a political community, they are likely to be capable of being endorsed by someone who understands the expressive dimension of punishment along communicative, rather than denunciatory, grounds. On communicative theories, punishment is understood as a communication addressed to an offender. Considerations about what makes for effective communication to such an offender might be appealed to by someone who thinks the primary purpose of punishment is communicative rather than denunciatory. This need not be an objection to my position. I set out to show that denunciatory accounts need not be vulnerable to the charge of justifying excessive penal harshness. I did not set out to show that they were unique in providing grounds for avoiding such harshness. Indeed, anyone who is concerned with reducing the harm done by incarceration ought to hope that they are not. They ought to hope that arguments in favor of harshness can

118  Bill Wringe be justified on a variety of different grounds, corresponding to different possible justifications of punishment. Nevertheless, there does seem to be something distinctive about the denunciatory view that I have set out here. It seems to require a certain kind of transparency in the operation of our penal institutions. On a (purely) communicative view we might be content with penal practices that took place out of the public eye. Provided that they served the goals of effective communication with offenders well, and provided that publicity was not required for effective oversight to prevent abuse of penal institutions, the average law-abiding citizen might be justified in being– more or less indifferent to what goes on in penal institutions. On the denunciatory account such indifference seems harder to justify. This seems like a consideration—albeit a fairly minor one—in favor of preferring a denunciatory version of an expressivist account of how expressive constraints on incarceration might be justified to a communicative one. A view which builds a requirement of transparency into the basis of our account of the kinds of punishment that can be justified seems attractive. But this advantage brings with it a corresponding drawback. Any account of legal punishment needs to provide some kind of account of the justifiability of punishing those who are not members of the political community whose rules they have transgressed. For almost everyone thinks that the punishment of visiting foreigners and of non-resident aliens can be justified. Denunciatory theorists of punishment have fewer problems in accounting for this familiar fact than many of their rivals. For there is no reason why the punishment of those who are not members of a given community should not play the same denunciatory role as the punishment of members. However, the denunciatory theorist does seem to face a problem when we consider the existence of expressive constraints on acceptable forms of punishment. For it seems intuitively appealing to suppose that if such constraints exist, they would apply in the same kinds of way to anyone the state punishes: and in particular to foreigners and non-resident aliens as much as to anyone else. But the denunciatory account that I have put forward here seems unable to explain why this should be so. The constraints were explained on the supposition that those who were punished were part of the intended audience of penal communication; that they were, in other words, part of the political community. It is worth considering two different kinds of non-member. One will be slightly surprising. Consider corporate entities. They are not themselves members of a political community (though their members typically will be). It doesn’t seem at all implausible to think that constraints on the punishment of individual community members might not be applicable to the punishment of corporations.6 For example, it might be permissible to respond to corporate wrongdoing by undermining or destroying the agency of the corporate

Expressive Constraints on Incarceration?  119 agent, provided this did not constitute unjustified punishment of, or in any other way infringe the rights of, any individuals (see Wringe 2012, 2016). The punishment of individuals who are not members of a given society presents a more difficult case. The denunciatory theorist might regard resident aliens (and perhaps even visitors) as having a kind of temporary membership of a political society (continued at least as long as they are imprisoned). Expanding the domain of the political, along lines I have suggested above, may make this strategy attractive. Alternatively, we might appeal to considerations of proportionality to rule out the possibility of non-citizens being punished more severely than citizens for the very same offenses. One final objection might arise with respect to those who might be regarded by their actions as having put themselves outside the political community. Aristotle says of those that live outside a political community that they must be either a monster or a god.7 We may not have any gods living among us, but some might think we have some monsters. I am not so sure. But there does seem to be one thing we can say. This is that even if some people do put themselves outside the political community, considerations of proportionality will still dictate that, if they commit the same crimes as non-monsters, their punishments need to be constrained in the same way. Only if we think that there are monstrous crimes, and not merely monstrous criminals, will we reach a case where the arguments of this paper do not apply.

Notes As did the authors of the American constitution, for example. 1 2 Talk of ‘levels’ of severity may suggest, misleadingly, that there’s only one dimension along which punishments may vary in respect of their severity. I doubt there is. However, all that’s required for my argument is that some sentences can be compared with other sentences in their degree of harshness. There’s little reason to be skeptical of this less ambitious claim. 3 Theories which incorporate a (general) deterrence-based element may face a similar problem here. For on such theories the justifiable level of punishment will depend on considerations such as how easily other offenders in a given society are deterred from committing similar crimes; how widespread knowledge of the levels of punishment to which particular individuals are subjected really is; and so on. 4 We might also take heart from Dzur’s (2012) interesting work on jury participation, which suggests that the prevalence of punitive attitudes among society as a whole might be reduced by increasing the degree to which members of the public participate in the judicial system as jury members and so on, 5 But see Wringe 2016, Chapter 3 6 Note that there will be those who think that it makes no sense to think of collective agents being punished, since they cannot suffer (Rich 2015). However, I do not agree that punishment is an essential element of suffering for reasons given in Wringe 2013. For further defense of this view about the relationship between punishment and suffering see Poama 2016; Coverdale 2013. 7 Aristotle 1981: 1253a1.

120  Bill Wringe

References Aristotle (1981). Politics, trans. J. Saunders Trevor. London: Penguin. Bennett, Christopher. (2008). The Apology Ritual. Cambridge: Cambridge University Press. Bentham, Jeremy. (1907). An Introduction to the Principles of Morals and Legislation. Oxford: Clarendon Press. Boonin, David. (2008). The Problem of Punishment. Cambridge: Cambridge University Press. Bottoms, Anthony E. (1995). “The Philosophy and Politics of Punishment and Sentencing.” In The Politics of Sentencing Reform, ed. C. Clarkson and R. Morgan. Oxford: Oxford University Press. Cochrane Alasdair. (2017). “Prison on Appeal: The Idea of Communicative Incarceration.” Criminal Law and Philosophy 11: 295–312 Coverdale, Helen. (2013). “Punishing with Care: Treating Offenders as Equal Persons in Criminal Punishment.” Unpublished Ph.D. Thesis, London School of Economics. Available at http://ww.etheses.lse.ac.uk/1080/. Cummins, Robert (1975). “Functional Analysis.” Journal of Philosophy 72: 741–64. Dagger, Richard (1993). “Playing Fair with Punishment.” Ethics 103: 473–88. Dagger, Richard (2008). “Punishment as Fair Play.” Res Publica 14: 259–75. Duff, R.Antony(1986). Trials and Punishments. Cambridge: Cambridge University Press. Duff, R.Antony (2001). Punishment, Communication and Community. Cambridge: Cambridge University Press. Dzur, Albert (2012). Punishment, Participatory Democracy, and the Jury. Oxford: Oxford University Press. Falls, Margaret M. (1987). “Retribution, Reciprocity, and Respect for Persons.” Law and Philosophy 6: 25–51. Feinberg, Joel (1965). “The Expressive Function of Punishment.” The Monist 49: 397–423, reprinted in Doing and Deserving, Princeton, NJ: Princeton University Press. Gilbert, Margaret (2006). A Theory of Political Obligation: Membership Commitment and the Bonds of Society. Oxford: Oxford University Press. Glasgow, Joshua (2015). “The Expressive Theory of Punishment Defended.” Law and Philosophy 34: 601–31. Grice, H. Paul (1957). “Meaning.” Philosophical Review 66: 377–88. Hampton, Jean (1992). “Correcting Harms Versus Righting Wrongs: The Goal of Retribution.” UCLA Law Review 39: 1659–702. Hanna, Nathan (2008). “Say What? A Critique of Expressive Retributivism.” Law and Philosophy 27: 123–50. Kant, Immanuel (1996). The Metaphysics of Morals, trans. M. McGregor and R. Sullivan. Cambridge: Cambridge University Press. Lippke, Richard (2004). “Against Supermax.” Journal of Applied Philosophy 21: 109–24. Lippke, Richard (2007). Rethinking Imprisonment. Oxford: Oxford University Press. Nozick, Robert (1981). Philosophical Explanations. Boston, MA: Harvard University Press. Poama, Andrei (2015). “Punishment Without Pain: Outline for a Non-Afflictive Definition of Legal Punishment.” Philosophy and Public Issues 5: 97–134.

Expressive Constraints on Incarceration?  121 Pratt, John. (2007). Penal Populism. London: Routledge. Rich, Sylvia (2016). “Corporate Criminals and Punishment Theory.” Canadian Journal of Law & Jurisprudence 29: 97–118. Schmid, Hans-Bernard (2016). “Being Well Together: Aristotle on Joint Activity, Plural Self-Awareness, and Common Sense.” Proceedings of the 37th International Wittgenstein Symposium, Berlin, De Gruyter. 295–312 Tadros, Victor (2011). The Ends of Harm: The Moral Foundations of Criminal Law. Oxford: Oxford University Press. Wright, Larry (1973). “Functions.” Philosophical Review 82:139–68. Wringe, Bill (2012). “Collective Agents and Communicative Theories of Punishment.” Journal of Social Philosophy 43: 436–56. Wringe, Bill (2013). “Must Punishment Be Intended to Cause Suffering?” Ethical Theory and Moral Practice 16: 863–77. Wringe, Bill (2016). An Expressive Theory of Punishment. London: Palgrave MacMillan. Wringe, Bill (2017). “Rethinking Expressive Theories of Punishment.” Philosophical Studies 174: 681–708.

7 Punishment, Restitution, and Incarceration David Boonin

Legal punishment involves treating those who break the law in ways that it would be wrong to treat those who do not. Even if we assume that those who break the law are responsible for their actions and that the laws that they break are just and reasonable, this practice raises a moral problem. How can the fact that a person is responsible for having broken a just and reasonable law render it morally permissible for the state to treat that person in ways that it would otherwise be impermissible to treat them? How can the line between those who break such laws and those who do not be morally relevant in the way that the practice of punishment requires it to be? This is the problem of punishment. In my 2008 book, The Problem of Punishment,1 I argued that there is no satisfactory solution to this problem and that it is morally impermissible for the state to punish people for breaking the law. Call this position abolitionism about punishment. Does abolitionism about punishment justify abolitionism about incarceration? That depends on a few things: what we mean by “punishment,” what we mean by “incarceration,” and what we favor as an alternative to punishment. In the Problem of Punishment, I argued that abolitionism about punishment does not entail abolitionism about compulsory victim restitution and that compulsory victim restitution could serve as a satisfactory alternative to the practice of punishment. In this chapter, I will consider what an abolitionist about punishment who favors compulsory victim restitution as an alternative should say about the practice of incarceration. I will begin by developing a definition of punishment and explaining why the case against punishment does not commit one to opposing compulsory victim restitution. I will then consider whether confining a person against their will could be justified as a form of restitution rather than as a form of punishment. I will argue that, as a general matter, such confinement is not justified by a system of compulsory victim restitution and that as a result an abolitionist about punishment who favors compulsory victim restitution should oppose the practice of incarceration. But I will also argue that in a limited set of circumstances, confining a person against their will might indeed be justified as form of restitution even if punishment is taken to be morally impermissible. This concession may initially seem incompatible with abolitionism about

Punishment, Restitution, Incarceration  123 incarceration. But I will conclude by suggesting that it is compatible with abolitionism about incarceration in the only sense of the term in which such abolitionism could prove to be defensible.

1. Punishment When we talk about punishment, what, exactly, are we talking about? A general answer to this question is simple: we are talking about such practices as the state’s imposition of monetary fines, forced imprisonment, physical suffering, and, in extreme cases, death. A more specific answer is more complicated, but it is also necessary. We can’t investigate the implications of abolishing punishment if we don’t know what, precisely, we are abolishing. So we need a definition of punishment. But not just any definition will do. We need a good one. What would make a definition of punishment a good one? First, a good definition of punishment must be accurate. It must provide us with a set of necessary and sufficient conditions that clearly demarcates cases of punishment from cases of something else. The results produced by this demarcation must cohere sufficiently well with what we mean by punishment when we argue about it and must do so over a sufficiently wide range of cases. Second, a good definition of punishment must be illuminating. A definition of punishment that successfully distinguishes between cases of punishment and cases of something else only because it contains various stipulations that are thrown in solely in order to produce the desired results and that have no further independent motivation is unacceptably ad hoc. We need a definition that gives us the right answers about particular cases and that explains why they are the right answers. Finally, a good definition of punishment must be neutral with respect to the question of whether or not punishment is morally permissible. It must not build into the meaning of punishment a feature that settles the debate about punishment’s merits by mere definitional fiat. In short, we need a definition of legal punishment that respects and reflects our beliefs both about what cases count as cases of punishment and about why they do, and that leaves open the question of what, if anything, renders it morally permissible for the state to punish people. A definition that satisfies these three requirements can be obtained by testing various conditions against our intuitive reactions to clear, paradigmatic instances of punishment: cases such as fines, imprisonment, torture, and death. So we should begin by asking what these various practices have in common. One thing these practices have in common is that they are bad for the person who is on the receiving end of them. If an offender were to receive a monetary prize for their offense, or a paid vacation, a relaxing massage or life-extending therapy, we would not be inclined to say that they had been punished for their transgression. And so a natural starting point in generating a definition of punishment is to say that punishment harms the person who is punished. Call this the harm requirement.

124  David Boonin The harm requirement enables us to correctly distinguish cases of punishment from cases of reward but by itself it is not enough. There are other acts that involve inflicting the same kinds of harms that are inflicted in cases of punishment but that are not themselves cases of punishment. Suppose, for example, that Larry marries Laura and is charged a fee for processing his marriage license while Moe marries both Betty and Veronica and is charged a fine for violating anti-polygamy laws. In both cases, someone does an act and is harmed by the state as a result of having done it and the kind of harm that is incurred is the same in both cases: financial harm. But while Moe is clearly punished by being forced to give up some of his money, Larry is clearly not. One reason for this arises from the distinction between intentionally harming someone and foreseeably harming them. When the state charges a fee for processing a marriage license, it understands that the cost imposes a harm on those who are getting married, but harming those who get married is not its intention. Its intention is merely to recover the costs that are involved in processing the relevant paperwork. When the state punishes someone, on the other hand, it inflicts various sorts of harmful treatment on them in order to harm them. It is not merely that in sentencing a prisoner to hard labor, for example, we foresee that the prisoner will suffer. Rather, a prisoner who is sentenced to hard labor is sentenced to hard labor so that they will suffer. If a given form of labor turned out to be too pleasant and enjoyable, they would be sentenced to some other form of labor for precisely that reason. This is not to insist, it is important to emphasize, that the suffering of the offender must be intended for its own sake. That would be to reduce punishment to sadism. Rather, it is to maintain that the punisher intends to harm the recipient of the punishment, and does not merely foresee it, even if this harm is in turn intended for the sake of some further end. When a parent punishes their child in a non-legal context by spanking them, for example, the pain inflicted on the child is not simply a foreseen side effect of the spanking, as it might be in the case of the pain caused by removing a splinter. If a splinter came out painlessly, the parent would not reinsert it in order to pull it out again in a more painful manner, but if the first spank was too mild to cause any pain, the parent would spank again, and harder. But while the parent who spanks their child thus clearly intends the pain and does not merely foresee it, they cause the pain not as an end in itself, but rather for the sake of some further end, such as educating the child or deterring them from committing similar infractions in the future. This is what makes what the punishing parent does similar to what the state does when it punishes offenders. As a result of these considerations, we must accept what I will call the intentional harm requirement: in order for an act to be an act of punishment, it must be done with the intent of causing harm to the person being punished. So punishment involves, at the least, intentional harm. But it also involves more than this. If I walk up to you and deliberately punch you in the nose

Punishment, Restitution, Incarceration  125 in order to make you suffer, then I harm you, and do so intentionally, but I do not punish you. This is because my harming you is not being done in response to some transgression of yours. So an additional requirement is this: in order to be an act of punishment, an act must involve intentionally harming someone because they previously did a prohibited act. And since we are concerned here with legal punishment in particular, we can be more specific: in order to be an act of legal punishment, an act must involve intentionally harming someone because they previously did a legally prohibited act, where saying that they did a legally prohibited act means that they are responsible for having done the act and that they had no valid legal excuse for doing so. Call this the retributive requirement. The retributive requirement is needed in order to distinguish cases of punishment from cases of mere gratuitous injury. Two additional requirements are needed in order to arrive at a definition of punishment that is sufficiently accurate, illuminating, and neutral. First, return for a moment to the case of Larry and Moe. Moe is punished since a fine is a case of punishment, and Larry is not punished because a fee is not a case of punishment. Part of the difference between a fine and a fee is successfully illuminated by attending to the distinction between intending harm and merely foreseeably bringing it about, and part of it is successfully illuminated by appealing to the fact that a fee is charged for doing a legal act while a fine is charged for doing an illegal one. But focusing exclusively on these two differences threatens to obscure a further relevant difference between the cases. When the state charges Larry a fee in order to process his wedding license, it is in no way expressing disapproval of Larry’s decision to get married. But when the state imposes a fine on Moe for violating its anti-polygamy laws, part of what it is doing is expressing its disapproval of Moe’s behavior. A fully illuminating definition of punishment will therefore need to embody a further requirement: in order for an act to count as an act of punishment of an offender for an offense, the act must express official disapproval of the offender for having committed the offense. Call this the reprobative requirement. Finally, since we are concerned here with legal punishment in particular, we need a definition that distinguishes the subject of our discussion from other kinds of punishment such as parental or divine punishment. And it is clear that an act of punishment cannot be an act of legal punishment, in particular, unless it is an act that is carried out by an authorized agent of the state acting in his or her official capacity. Call this the authorization requirement. Legal punishment, then, is authorized intentional reprobative retributive harm. Let’s now assume that we have a sufficiently clear understanding of what legal punishment is. Given this understanding, what is the problem? The answer is simple. First, the practice of punishment involves drawing a line between two different sets of people and treating the members of one group very differently from the members of the other. In general, it is wrong to

126  David Boonin treat two groups of people so differently unless there is a morally relevant difference between them. So part of the problem is explaining why the difference between those that the practice of punishment targets and those that it does not is a morally relevant one. Second, the practice of punishment involves not merely treating the members of one group differently from the members of the other, but causing harm to the members of one group and not to the members of the other. The fact that an act will cause harm to someone is clearly morally relevant, and so part of the problem is explaining why the difference between offender and non-offender is important enough to justify acts that will harm the offender. Finally, and perhaps most importantly, the practice of punishment involves not merely acts that predictably cause harm to offenders, but acts that are carried out precisely in order to harm them. Since it seems considerably more difficult to justify intentionally harming someone than it is to justify merely foreseeably harming them, the problem of punishment is even greater than it might at first appear to be: we must explain not simply why the line between offenders and non-offenders is morally relevant at all but must explain, in particular, how it can be important enough to justify not merely harming those on one side of the line, but intentionally harming them. It is one thing to justify the claim that it is morally permissible for the state to act in various ways while foreseeing that acting in those ways will cause harm to some of its citizens (e.g., changing the speed limit, modifying air pollution standards, imposing new regulations, raising taxes, imposing zoning restrictions, requiring people to serve on juries, conscripting soldiers, all of which foreseeably cause harm to a significant number of people). It is quite another to justify the claim that it is morally permissible for the state to act in various ways in order that some of its citizens be made to suffer. Yet this is precisely what must be justified in order to justify the practice of punishment. The practice of punishment, in short, involves the state treating some of its citizens in ways that it would clearly be wrong to treat others. The problem is to explain how this can be a morally permissible thing to do. There have, of course, been many attempts to meet this burden, and I argued, in The Problem of Punishment, that none of them are successful. Let’s assume, at least for the purposes of this chapter, that I was right about that.

2.  Punishment and Compulsory Victim Restitution I argued in the previous section that legal punishment should be defined as authorized intentional reprobative retributive harm. In arguing for this definition, I claimed that it does the best job of accurately distinguishing between punishment and other practices and that it does so in an illuminating and neutral manner. In developing this claim, I appealed to a variety of examples, but I deliberately refrained from appealing to one particular kind of case, that of compulsory victim restitution. I did this precisely so that the

Punishment, Restitution, Incarceration  127 definition of punishment that I arrived at would not have been influenced ahead of time by preconceptions about the relationship between punishment and restitution. Having arrived at what strikes me as a satisfactory definition of punishment, however, we must now ask: given this definition, what should we say about the practice of compulsory victim restitution? Is it a form of punishment or is it something else? If it is a form of punishment, then abolitionism about punishment straightforwardly entails abolitionism about compulsory victim restitution. But if it is not a form of punishment, then considerations that ground a case against punishment need not ground a case against compulsory victim restitution, and it is open to the abolitionist about punishment to recommend that compulsory victim restitution take its place. Compulsory victim restitution involves the state’s forcing an offender to compensate their victims for the harms that they are responsible for having wrongfully caused. Here is an example. Larry vandalizes Moe’s car by painting obscene words on it and breaking the windshield. Larry is caught and found guilty. The judge orders Larry to compensate Moe for the harm he has inflicted on Moe. She forces Larry to remove the spray paint from Moe’s car, to pay the costs of replacing the windshield, to pay the costs involved in Moe’s renting a car while his is in the shop, and to compensate him for the inconvenience and emotional distress Larry caused him. This seems clearly to be a case of compulsory victim restitution. Is it also a form of punishment? Clearly the act satisfies the authorization requirement. The judge is a legally authorized official acting in her official capacity. Clearly it satisfies the retributive requirement. She orders Larry to do these things because of the fact that Larry illegally vandalized Moe’s car. It seems equally clear that the act satisfies the reprobative requirement. The judge’s act, at the very least, expresses the view that Moe is entitled to have his car returned to its original condition, which entails that Larry was not entitled to render it in the condition his act rendered it in, which in turn entails that Larry did something that he was not entitled to do. And finally, it seems clear enough that the judge’s act harms Larry. Larry is made worse off by having to give Moe some of his money and by having to spend some of his time cleaning up Moe’s car. In all of these respects, the judge’s act of forcing Larry to make restitution to Moe satisfies the definition of punishment we have arrived at. Indeed, this set of similarities between punishment and restitution has seemed to some theorists to be sufficient to conclude that compulsory victim restitution is a form of punishment.2 But this conclusion is premature. For one further question remains: in imposing this burden on Larry, does the judge act with the intention of causing harm to Larry? As we saw in the previous section, distinguishing between harm and intentional harm is necessary in order to fully account for the difference between, for example, fees and fines. Part of what makes a fine punitive and a fee non-punitive is that fines are levied in order to inflict

128  David Boonin harm on those who are fined and fees are not. If the answer to the question of whether the judge acts with the intention of causing harm to Larry is yes, then compulsory victim restitution is more like a fine than a fee and is a form of punishment. If the answer is no, then compulsory victim restitution is more like a fee than a fine and is not a form of punishment. The answer to the question is this: there is no one answer that covers every case in which a judge compels an offender to make restitution to their victim. In some cases, the judge’s intent may well be to impose a cost on the offender. She may say, for example, that she wants Larry to suffer the drudgery involved in cleaning the paint off Moe’s car so that Larry will come to see how wrongful his act really was. In this sort of case, part of the judge’s motivation for imposing the cost on Larry is, indeed, punitive. Following Barnett, we can refer to cases of this sort as cases of punitive restitution (Barnett 1977: 219–20). But this need not be true of all cases. In some cases, the judge’s reasoning may simply be that Larry must pay Moe the money simply because Larry owes it to Moe. In this case, the judge foresees that paying Moe will impose a cost on Larry, in the same way that she might foresee that enforcing the terms of a contract will impose costs on one of the parties to the contract, but this fact plays no role in her decision. Following Barnett, we can refer to cases of this sort as cases of pure restitution (Barnett 1977: 220). So is compulsory victim restitution a form of punishment? Sometimes it is and sometimes it isn’t. It is when it involves punitive restitution but it isn’t when it involves pure restitution. Let’s focus, then, on a system in which pure restitution replaces the practice of punishment. In such a system, the law would compel offenders to make compensation to their victims in the same spirit in which it enforces contractual agreements: recognizing that harm will result but not aiming at the harm. To be made to honor the terms of a contract that you have consented to is clearly not to be punished and so, on a system that treats compensation in the same way, to be made to compensate a victim for the harms that you have wrongfully caused them is not to be punished either. In both cases, the state compels you to do something that will harm you, but in neither case does it do what it does in order to harm you. In each case, it simply enforces the payment of a debt that you have incurred. Since a system of pure restitution would not involve the intentional infliction of harm, and since the intentional harm requirement was a crucial component of the definition of punishment we arrived at in the previous section, a system of pure restitution would not be a system of punishment according to that definition. To establish that a system of pure restitution would not be a system of punishment is not yet to establish that abolitionism about punishment does not commit one to abolitionism about compulsory victim restitution. It shows that punishment abolitionism does not entail restitution abolitionism as a matter of mere definition. But it might still be the case that punishment abolitionism entails restitution abolitionism as a substantive matter.

Punishment, Restitution, Incarceration  129 If the case against legal punishment requires making a case against the existence of the state, for example, then since compulsory victim restitution also depends on the existence of the state, the case against punishment will entail a case against compulsory victim restitution as well. The abolitionist about punishment who appeals to such an argument, therefore, will not be in a position to appeal to compulsory victim restitution as an alternative to punishment. But while abolitionists about punishment could appeal to an argument against the very existence of the state, they need not do so. This is because of two of the features of legal punishment that distinguish it from compulsory victim restitution and that help to generate the problem of punishment in the first place. First, as I emphasized at the end of the previous section, punishment involves not just the state’s harming some of its citizens, but the state’s intentionally harming some of its citizens. There are many contexts in which it might prove permissible for the state to do an act that results in harm to someone even though it would be impermissible for it to do an act in order to intentionally harm them. And so given this important feature of punishment, an abolitionist about punishment can agree that it is permissible for the state to levy taxes, impose environmental regulations, zoning restrictions, labor standards, and so on, while nonetheless maintaining that it is impermissible for the state to punish offenders for breaking the law since punishment involves intentional harm and these other practices do not. Indeed, since it is difficult if not impossible to think of any commonly accepted state activity other than punishment that involves the state’s deliberately targeting some of its own citizens and then deliberately acting in certain ways in order to harm them, it is open to the abolitionist about punishment whose critique focuses on the intentional nature of the harm involved in punishment to agree that every commonly accepted state activity other than punishment is morally justified. The second relevant difference between a system of punishment and a system of pure restitution arises from the retributive nature of punishment, the fact that we punish people as a response to their having violated a legal prohibition. The practice of punishment, that is, involves not just the state’s intentionally harming some of its own citizens but its treating the line between those who break the law and those who do not as justifying treating those on one side of the line in ways that it would never treat people on the other. This feature, too, is absent in a system of pure restitution. While the practice of punishment involves the state’s treating people who break the law in ways that we would not permit it to treat people who do not break the law, the practice of compulsory victim restitution does not. The state compels non-offenders to make restitution to others all the time. Whenever one party successfully sues another party for damages in a civil lawsuit, for example, someone who is guilty of no violation of the criminal law is nonetheless compelled to compensate someone else for damages that they are found culpable for having caused. And so it is also open to the abolitionist about punishment

130  David Boonin to make use of this fact in constructing an objection to punishment that does not entail an objection to compulsory victim restitution. As a result of the intentional harm requirement and the retributive requirement, then, the practice of compelling an offender to make restitution to their victims does not raise the distinctively difficult moral problem that the practice of punishment does, and the claim that the state has no right to punish people for breaking the law therefore provides no reason to deny that the state has the right to compel offenders to make restitution to their victims. All of this matters, of course, only if a system of pure restitution can reasonably be viewed as a viable alternative to a system of punishment. It might seem clear that it cannot. And, if it cannot, then it cannot be used to ground a view on which the practice of incarceration should be abolished. Space does not permit a complete discussion of the objections that can be raised against a system of pure restitution, so after making a few comments about the basic idea behind such a system, I will focus here on one objection that seems most salient to illuminating the connection between the abolition of punishment and the abolition of incarceration.3 The idea behind a system of pure restitution can be put as follows: when offenders break the law, they cause wrongful harms to their victims. When people cause wrongful harms to their victims, their doing so generates a debt on which they owe their victims compensation sufficient to restore them to the level of well-being that they rightfully enjoyed prior to their having been wrongfully harmed. People who commit such offenses therefore owe such compensation to their victims. When people break the law, it is thus morally appropriate for the state to compel them to make such compensation. When an offender is forced to compensate their victims for the harms that they have wrongfully caused them, the offender is harmed as a result in various ways. But the state does not extract compensation from the offender in order to cause them harm, either as an end in itself or as a means to some further end. Note that the theory behind a system of pure restitution does not by itself specify what forms restitution should take. Ideally, restitution would involve restoring the victim to a condition that is identical to the one they enjoyed prior to the offender’s wrongful act. To the extent that this is impossible in any particular circumstance, the theory maintains that the offender must restore the victim to a condition that is of equivalent value to that of the condition they enjoyed prior to the offender’s wrongful act. And, to the extent that this, too, is impossible in any particular circumstance, the theory maintains that the offender must restore the victim to a condition that is as close in value to that of their original condition as possible. But how this is to be done is, in all of these kinds of cases, a question that the theory itself does not answer. Just as a theory of punishment need not itself tell us how, specifically, we should inflict on an offender the suffering that they deserve, so the theory of pure restitution does not tell us how, specifically, we should have the offender make restitution to their victim.

Punishment, Restitution, Incarceration  131 This lack of specificity about the form that restitution should take can seem to pose a problem for the following reason: without an explicit provision to the contrary, victim restitution is often taken to be exclusively limited to monetary compensation, and a system based exclusively on monetary compensation can seem inadequate as a replacement for a system based largely on punishing people for breaking the law. This can perhaps be seen most clearly in the case of what is arguably the most common and fundamental objection to a system of pure restitution: the claim that it is too individualistic. “The idea of restitution is actually quite simple,” Barnett writes in defending a system of pure restitution. “It views crime as an offense by one individual against the rights of another” (Barnett 1977: 219). “Where we once saw an offense against society,” he adds, “we now see an offense against an individual victim. In a way, it is a common sense view of crime. The armed robber did not rob society; he robbed the victim” (Barnett 1977: 219, emphasis in original). This may be a common sense view of what is involved when a person violates the law. But, according to its critics, it is also a deeply mistaken view. An offense committed against a single individual does more than wrongfully harm that particular individual, the objection maintains; it also has important repercussions for the rest of society. If a gunman robs my neighbor as he is about to enter his home, for example, then the gunman has wrongfully harmed my neighbor. But by robbing my neighbor, he may well also wrongfully cause various harms to me. I may suffer from anxiety and lost sleep as a result. I may feel forced to incur the added expense of installing and maintaining a security system in my house, or of buying a gun. My property value may go down. My home insurance rates may go up. I may incur various opportunity costs, forfeiting whatever I would have enjoyed with the time and money that I have instead been forced to devote to responding to the offense against my neighbor. And these negative consequences that the armed robbery of my neighbor causes to me may in turn impose still further costs on other people, including people who don’t even know that my neighbor was robbed (e.g., people whom I would have benefited with my money had I not bought the security system instead). In all of these ways, and in many others as well, an offense against a single individual can also seem to be an offense against the rest of society. Yet since, according to the objection, a system of pure restitution requires only that offenders provide monetary compensation to their individual victims, such a system seems incapable of satisfactorily recognizing and addressing this fact. And so, it is said, the proposal that such a system replace the practice of punishment must be rejected. Requiring restitution may be good enough when it comes to dealing with the victim of the offense, but punishment is still necessary when it comes to dealing with the effects of a crime on the rest of society. As one critic of a system of pure restitution without punishment has put it, “the [pure] restitution theory understates the importance and the complexity of the network of relationships that is

132  David Boonin disrupted by crime, relationships too complex to be repaired through payment of compensation” (Hoekema 1991: 343).4 If compensation is limited to monetary payments to the offender’s immediate victim, this objection seems difficult to overcome. But there is nothing about the idea behind a system of pure restitution that requires compensation to be limited in this way. There are two distinct ways in which the scope of compensation might be expanded, and once this is recognized, we are in a position to see how a system of pure restitution might well serve as a satisfactory alternative to a system based primarily on punishment. First, there is no reason for a proponent of a system of pure restitution to deny that, at least in a good number of cases, individuals other than the offender’s immediate victim are also wrongfully harmed by the offender’s act. Rather than denying or ignoring this fact, a proponent of such a system can simply treat these additional people as additional victims, albeit secondary victims rather than direct or primary victims. What I will call the secondary victims response to the harm to society objection maintains that in such cases the offender must make restitution to these people as well. In short, the secondary victims response simply permits the rest of society to be treated as a further victim and then to be treated in just the same way that any other victim would be treated by a system of pure restitution.5 If punishment is not needed in virtue of the offender’s effect on their direct or primary victim, then it is not needed in virtue of their effect on the rest of society either. How could an offender be made to provide restitution to their secondary victims? In an ideal world, the state would simply perform all of the relevant calculations, determine precisely how much harm an armed robber who robbed a particular person had wrongfully caused to the other members of their community, and then assign a precise dollar amount to measure each instance of harm. Perhaps it would determine that the robber caused more secondary harm to the victim’s immediate neighbors, with the level of harm then gradually dissipating as people were further and further removed from the victim. In this case, the offender would owe some money to each nearby neighbor and then some, but less, to others further removed. Or perhaps it would be determined that the harm had been distributed in some other pattern, in which case the requisite monetary payments would be as well. So an initial attempt to sustain the secondary victims response would simply maintain that the offender should be compelled to give just enough money to every secondary victim to compensate each of them for the harm that they wrongfully caused them by their unlawful act. This answer follows naturally enough from the basic idea of restitution and would show how a system of pure restitution can, in fact, accommodate the legitimate claims that secondary victims may often have without resorting to the imposition of punishment. This version of the secondary victims response is also, of course, completely impractical. But this fact does nothing to show that the response

Punishment, Restitution, Incarceration  133 itself is mistaken. After all, it may often be equally impractical to punish a criminal by subjecting them to precisely the amount of suffering that they are thought to merit, but this does not in itself establish that the practice of punishment is unjustified. It simply establishes, in either case, that in the real world we often cannot do precisely what we should do, or precisely what we are entitled to do. And here, as in other relevantly similar cases, if the law cannot do precisely what it should or may do, this does not mean that it should or may do nothing at all. It simply means that it should do the best that it can to approximate doing what it should or may do. Consider, for example, the widely publicized law suits against the major tobacco companies. Assume for the sake of the example that the companies in question wrongfully harmed millions of smokers and should have to pay damages to millions of smokers as a result. Ideally, the state would identify every person who has been wrongfully harmed by tobacco smoke, either directly or indirectly, determine the precise dollar amount of the harm that each person suffered, and use this as a basis for determining the precise amount of compensation owed to each of the millions of victims involved. Given that it would be permissible for the state to do this if it could, but given also that it cannot, it is permissible for the state to attempt to approximate this result in a reasonable manner. It might, for example, take a lump sum payment from the companies and use it to improve the health care system in ways that would be most likely to most benefit those who were most harmed by the wrongful behavior of the tobacco companies. Similarly, then, given that it would be permissible for the state to compel the armed robber to pay every member of the community precisely the amount of compensation he owed to each of them, but given also that doing so would be practically impossible, it is permissible for the state to attempt to approximate this result in a reasonable manner. It might, for example, compel the robber to pay a lump sum to the city which would use the money to help pay for an extra patrol officer, or two, or three, depending on how much was necessary to restore the community to its previous level of well-being, and which would deploy the new officers in whatever manner would be most likely to bring the greatest benefits to those who had suffered the greatest harms as a result of the robber’s wrongful actions. If the forms of restitution that may permissibly be extracted from an offender are limited strictly to monetary payments, there may be little more to add in response to the harm to society objection. But there is nothing about the idea behind a system of pure restitution that commits one to this limitation. Barnett himself describes the theory of pure restitution in general as holding that the offender must “mak[e] good the loss he has caused” (Barnett 1977: 219), and this characterization does not, in itself, specify the form in which restitution can be made in order for the debt to be erased. Barnett himself, it is true, seems to argue as if the only form of payment that can permissibly be extracted from the offender is a monetary payment. All of the specific examples of restitution that Barnett provides, for example,

134  David Boonin involve the offender making monetary payments to the victim, either from their wages or sale of property (Barnett 1977: 219) or from garnishment on their future income (Barnett 1977: 220). And in his discussion of objections to the theory of pure restitution, Barnett at times seems to equate restitution in general with financial restitution in particular. He considers the objection that some offenses cause harms that cannot be “expressed in monetary terms” as an objection to the theory itself, for example (Barnett 1977: 223), and presents as an objection to the theory the claim that “monetary sanctions are insufficient deterrents to crime” again equating restitution in general with monetary compensation in particular (Barnett 1977: 226). But this is a mistake, and a common one.6 While giving money to a primary victim or to secondary victims is one way in which an offender might help to restore victims to their previous level of well-being, it is not the only way. And once this fact is recognized, the secondary victims response to the harm to society objection can be developed even further. Suppose, for example, that a particularly talented burglar has been particularly successful. He has finally been apprehended and has been made to compensate his victims for the harms he has wrongfully caused them. He has returned all the stolen goods or made payments sufficient to cover the costs in replacing them, has compensated people for whatever relevant emotional or psychological distress was involved, and has paid for each of them to have a new burglar alarm installed as a means of restoring them to their previous level of security. But suppose that all of this is still not sufficient to restore his victims to their previous level of well-being. Suppose, in particular, that even with their new alarm systems, organized community watches and extra police officers patrolling the neighborhood, his victims have still been made less secure by his actions than they were before merely in virtue of the fact that he is so skilled at evasion. In cases such as this, a system of pure restitution may be unable to do anything more if it limits itself to monetary forms of restitution. But this is not a problem for the proponent of such a system. Rather, it is a reason for the system to refuse to be constrained by such limits. There are many other ways that the burglar in this case could more fully restore his victims to their previous level of security. He might, for example, be compelled to wear a monitoring device secured to his ankle by which his location could be confirmed by the police at all times.7 He might be subjected to a form of intensive supervision, such as those that currently accompany the granting of probation in some cases and that often include the imposition of a curfew.8 In other sorts of cases, an offender might be made to take anger management courses, to undergo therapy, to give up drinking, to stay away from certain areas, or certain people, or people under a certain age, and so on.9 If one or more of these impositions are necessary in order for an offender’s victims to be fully restored to the level of security that they rightfully enjoyed prior to his wrongful acts, then he owes it to his victims to undergo them and the impositions could be fully justified as a form of pure

Punishment, Restitution, Incarceration  135 restitution. In these kinds of ways, then, and many more, a system of pure restitution could go beyond compelling an offender to pay money to his immediate victim as a way of compelling him to repair the various social harms that he is also responsible for having caused.

3.  Compulsory Victim Restitution and Incarceration I have argued so far that once the difference between punishment and restitution is fully understood, we can see that an abolitionist about punishment is not committed to abolitionism about compulsory victim restitution and that once the variety of options available to a system of pure restitution are fully understood, we can see that a proponent of compulsory victim restitution need not succumb to the common objection that restitution without punishment neglects legitimate social claims beyond those of the immediate victim. If a system of pure restitution can deal with other potential objections in a similarly effective manner, there might therefore seem to be a relatively straightforward path to the conclusion that an abolitionist about punishment should be an abolitionist about incarceration: incarceration is a form of punishment, punishment cannot be justified as a form of pure restitution, so incarceration cannot be justified as a form of pure restitution; an abolitionist about punishment should endorse pure restitution as an alternative to punishment, and so an abolitionist about punishment should be an abolitionist about incarceration. But things are not so simple. For having reached this point, we must now ask two final questions: if a system of pure restitution can include such non-monetary responses to criminal activity as intensive supervision, ankle monitors, mandatory anger management classes, and restraining orders, is there anything to stop it from including involuntary confinement as well? And if involuntary confinement can be justified as a form of pure restitution, doesn’t that mean that the abolitionist about punishment who endorses compulsory victim restitution can’t be an abolitionist about incarceration after all? I’m inclined to think that the answer to the first question is no: there is no principled reason to preclude the possibility that involuntary confinement could be justified as a form of pure restitution. If the particularly talented burglar is sufficiently talented not just at evading intensive forms of supervision but at disabling ankle monitors, for example, it might be that nothing short of locking him up will suffice to restore his primary and secondary victims to the level of security and well-being that they were previously enjoying. Under such circumstances, involuntary confinement might well be justified as a form of pure restitution. A critic might object to this claim by maintaining that to accommodate the legitimate grievances of the community in this way is simply to abandon the practice of restitution altogether and to retreat to the practice of punishment. What, after all, is locking an offender up under such circumstances if

136  David Boonin not punishing him for breaking the law? McCarthy, for example, characterizes home detention and even intensive supervision as “humane, but punitive” on the grounds that “they deliberately impose suffering through the deprivation of liberty” (McCarthy 1987: 2). And Schmidt and Curtis write, approvingly, that “[h]ome incarceration meets the public demand for punishment” (Schmidt and Curtis 1987: 142). Indeed, some writers have taken what I am presenting here as an extension of a system of pure restitution and tried to use it to ground a novel and defensible theory of punishment. On what Holmgren defends as a “restitutive theory of punishment,” for example, punishment is justified precisely as a form of restitution: “The persons who commit crimes deprive members of the community of the security they are entitled to, and part of that security is restored to them when those criminals are punished” (Holmgren 1983: 45, 41).10 But viewing involuntary confinement under these circumstances as a form of punishment is a mistake, and an important one. Not every legally imposed restriction on a person’s freedom of movement is a form of punishment. It is a form of punishment only if it is done with the aim of making an offender suffer for his offense.11 McCarthy, for example, plainly acknowledges this requirement when she appeals to the claim that home detentions “deliberately impose suffering,”12 but there is nothing intrinsic to involuntary detention or intensive supervision that requires that they be done in order to impose suffering rather than in order to bring about some other result with the harms to the person being detained or supervised being merely foreseen rather than intended. Consider, for example, a defendant awaiting trial who is involuntarily confined because they are considered a risk to flee. The restriction on their freedom of movement is every bit as harmful to them as that of an offender serving a prison term is to them. But pre-trial detention, even though it involves precisely the same restriction on freedom of movement, is not a form of punishment precisely because it is not done with the intent of inflicting the harm that it brings about. The same can be said of a wide variety of other practices including curfews, quarantines, travel restrictions, restraining orders, conditions of probation or of making bail while awaiting trial, and so on. In the case of the exceptionally talented burglar in a system of pure restitution, if the offender is locked up this is done with the intent of restoring his victims to their previous level of security. Since this, rather than the harm to the offender that the detention brings about, is the state’s aim, his involuntary confinement does not count as an instance of punishment. And so there is no reason that involuntary confinement must be ruled out by a system of pure restitution. Indeed, if it is necessary in order to restore people to their original level of well-being in certain cases, it may well occasionally prove to be a necessary part of such a system.13 A critic might also object that the difference between detention as punishment and detention as restitution is purely verbal and that without any substantive differences, there is no reason for an abolitionist about the former to be open to endorsing the permissibility of the latter. But this, too, is

Punishment, Restitution, Incarceration  137 a mistake. The difference between involuntary confinement of an offender on punitive grounds and involuntary confinement of an offender on restitutive grounds is of practical as well as theoretical significance. Suppose, for example, that it was determined that a particular offender was going to be sentenced to a form of involuntary confinement, and the question was whether they should be detained in their home and monitored electronically, or confined to a prison. The literature on home detention strongly suggests that, at least in many kinds of cases, confining an offender at home decreases their chances for committing further offenses and costs less money relative to confining them to a prison.14 If the detention is to be justified as a form of restoring the community to its previous level of well-being, these considerations would tell in favor of home detention. But it is surely much more safe, comfortable, and pleasant to be confined in one’s own home than to be sent away to a prison, and so if the incarceration were to be justified as a form of punishment, these considerations would tell in favor of prison. In addition, when offenders are involuntarily confined on punitive grounds, they are routinely deprived of goods such as cigarettes, television, exercise equipment, and a long list of other things that might make life in prison less unpleasant for them. If the goal of confining the offender is to make them suffer, these deprivations will often be justified. But if the goal of confining the offender is simply to ensure that their community is restored to the level of security it enjoyed prior to their offense, then there will be no justification for making their life any less pleasant than is required by their being confined. Compensatory confinement differs from punitive confinement both in theory and in practice, then, and as a result an abolitionist about punishment should remain open to the possibility of endorsing involuntary confinement as a possible form of pure restitution. If involuntary confinement can be justified as a form of pure restitution, though, doesn’t that mean that the abolitionist about punishment who endorses compulsory victim restitution can’t be an abolitionist about incarceration after all? That depends on what we mean by incarceration. On perhaps the most common understanding of the term, “incarceration” is simply a synonym for “imprisonment.” And on the assumption that a prison is a building in which people are confined as a form of punishment for the crimes they have committed, it follows that an abolitionist about punishment must indeed be an abolitionist about incarceration. Incarceration, on this understanding, is simply a form of punishment and so is included in the scope of the punishment abolitionist’s view by definition. On another common understanding of the term, though, “incarceration” has a much broader meaning. On this interpretation, to incarcerate someone simply means to confine or detain them, to keep them in custody or captivity, regardless of whether the reason for doing so is punitive in nature. On this broader sense of the term, the kind of involuntary confinement that I have claimed can be justified as a form of pure restitution would count as a form of incarceration. And so in this sense of the term, an abolitionist about punishment

138  David Boonin cannot be an abolitionist about incarceration, at least if their opposition to punishment does not, as I think it should not, extend to opposition to compulsory victim restitution. There is no reason to think that one of these definitions of incarceration is better or more reasonable than the other as a general matter, but there is a reason to prefer one to the other in the particular context of a serious debate about whether the practice of incarceration should be abolished. The broad definition of incarceration includes a number of practices that virtually everyone will agree are perfectly legitimate. Virtually everyone, for example, agrees that under some circumstances it can be justifiable for a patient who has been diagnosed with a highly contagious disease to be quarantined. On the broad understanding of the term, quarantine amounts to a form of incarceration. The claim that some people may permissibly be confined to a psychiatric hospital because they pose a threat to themselves or others is also largely uncontroversial, and on the broad interpretation of the term this, too, involves a form of incarceration. And most people agree that it can be permissible to detain someone awaiting trial in cases where they pose a clear flight risk. On the broad understanding of the word, this practice involves a form of incarceration as well. The only sense of the term in which the practice of incarceration is genuinely controversial, then, is the narrow sense, and in the narrow sense of the term, the abolitionist about punishment is indeed an abolitionist about incarceration. So the answer to the question of whether the punishment abolitionist should be an incarceration abolitionist seems to be not so much “yes and no” as “in the relevant sense, yes.” This isn’t quite the final answer, though, because there is a third way of understanding the term “incarceration,” one that lies somewhere in between the narrow and broad senses. Like the narrow sense of the term, this intermediate sense begins by limiting the meaning of incarceration to cases in which a person is confined to a prison in particular. But unlike the narrow interpretation, this intermediate interpretation does not assume that a prison must be, by definition, a place where offenders are sent to be punished. It requires only that a prison be a place where people are confined because they have broken the law. This leaves open the possibility that a prison might best be understood as a penal facility, a correctional facility, or simply a detention facility. And this, in turn, leaves it open that the kind of involuntary confinement that I have claimed can be justified as a form of pure restitution could turn out to count as a form of incarceration on a more limited understanding of that term after all. It therefore makes sense to conclude by considering whether, in this third and final sense of the term “incarceration,” the abolitionist about punishment should be an abolitionist about incarceration. Here the answer depends on what we mean by an abolitionist. In one obvious sense, someone who is an abolitionist about a particular practice wants to abolish every single instance of the practice. Abolitionism about

Punishment, Restitution, Incarceration  139 slavery means that there should be no slaves at all, for example, not just that there should only be a relatively small number of them. When this sense of abolitionism is combined with the intermediate sense of incarceration, the result is that an abolitionist about punishment who endorses compulsory victim restitution as an alternative to punishment cannot be an abolitionist about incarceration. They cannot insist that people who break the law should never, under any circumstances, be detained against their will because they should concede that doing so may sometimes be justified as a form of pure restitution. But in another important sense of the term, an abolitionist about a particular form of behavior can be someone who wants to end a practice on which the behavior in question is treated as routine without insisting that the form of behavior be ended under all circumstances. Mammograms are widely viewed as a routine part of women’s health care, for example, and someone who wants to abolish the practice of treating mammograms in this way can still agree that mammograms are appropriate under particular kinds of circumstances. Similarly, incarceration is currently a routine, indeed the routine, practice when it comes to violations of the criminal law. On a standard definition of the term, for example, what it is for a crime to constitute a “felony” just is for it to be an offense of the sort that warrants more than a year in prison. An abolitionist about incarceration in this second sense of the term is someone who wants to abolish the system on which incarceration is established as a routine response to violations of the criminal law, and this need not commit them to denying that there may be particular cases in which involuntary confinement of an offender may be warranted. I have argued here that involuntary confinement may sometimes prove to be justified as a form of pure restitution in response to a violation of the law, but only in cases where nothing short of this suffices to restore a community to the level of security that it is entitled to. A system of pure restitution without punishment would treat incarceration as a last resort, not as standard operating procedure. And so in this second sense of abolitionism, an abolitionist about punishment who endorses compulsory victim restitution as an alternative to punishment can and should indeed be an abolitionist about incarceration. Which sense of abolitionism is relevant to the debate over incarceration here probably depends on which form of abolitionism seems more reasonable. Some people may believe that confining an offender against their will is wrong regardless of the circumstances and regardless of the rationale. For such people, no form of opposition to incarceration that stops short of opposing it in all cases merits the term “abolitionism.” But most people, I suspect, including most people who consider themselves to be against incarceration, do not really believe that there are no circumstances at all under which it may be permissible for an offender to be involuntarily confined as a result of their criminal behavior. What they really believe is that the system on which a prison sentence is the standard response to criminal activity, the

140  David Boonin system on which a felony is by its very nature something that justifies putting someone in prison for over a year, is unjustified and should be replaced. For such people, being an abolitionist about incarceration means opposing the system of incarceration. And in this sense, the sense that strikes me as the most salient sense of the relevant terms, the abolitionist about punishment who endorses compulsory victim restitution as an alternative to punishment should indeed be an abolitionist about incarceration. Whether one should be an abolitionist about punishment in the first place, of course, remains beyond the scope of this chapter, as does the question of whether an abolitionist about punishment should endorse compulsory victim restitution as an alternative to punishment. Readers interested in these further questions are encouraged to consult my book, The Problem of Punishment. But if the central argument in that book is sustained, then the practice on which incarceration is the routine response to criminal activity should indeed be abolished.

Notes 1 Cambridge: Cambridge University Press, 2008. The discussion in this essay is largely drawn from that book. 2 Martin, for example, characterizes “compensation, to be paid to victims of crime” as one of the three main “modes of punishment”: “[w]hen a violator is forced to pay compensation—as a feature of the verdict against him—such compensation counts as punishment” (1987: 73, 74; for a related argument see also Martin 1990). Similarly, Abel and Marsh argue that compulsory victim restitution is a form of punishment “because it involves unpleasant consequences for an individual who has interfered with our pursuit or realization of individual and social ends” (1984: 18–19). Similar views are endorsed by Dagger (1991: 36–7), Baylis (1968: 45–6), Hoekema (1991: 336–40), and, at least implicitly, Loewy (2000: 15–16). 3 For a more complete defense of a system of pure restitution against a wide variety of objections that have been raised against it, see Boonin (2008: ch 5). 4 See also Hoekema (1991: 340). This harm to society objection has been pressed by a number of other critics of the theory of pure restitution, including Miller (1978: 359), Tunick (1992: 158ff.), and Kleinberg (1980: 277). 5 It might be objected that the secondary victims response depends on a certain view of responsibility and that depending on it thus prevents the defender of the theory of pure restitution from maintaining neutrality about the nature of responsibility. On some views, for example, it might turn out that an offender is only responsible for the harms that he intends to inflict on his immediate victim, and on such a view the theory could not, in fact, justify the extraction of restitution to secondary victims. While it is true that the secondary victims response cannot be sustained if a certain view of responsibility is assumed, however, it is not true that this poses a problem for the theory. If it is assumed that an offender is not really responsible for any of the harms he causes to the rest of society, then it is true that the theory of pure restitution will imply that the offender owes no debt to society, but given the assumption, this result should seem perfectly reasonable. 6 Abel and Marsh, for example, make the same mistaken assumption throughout their 1984 book on punishment and restitution, and Golash’s brief discussion of

Punishment, Restitution, Incarceration  141 restitution in her more recent book seems to do the same (2005: 162–5). Ellin (2000: 308) notes that victim-offender reconciliation could be pursued as an alternative to monetary compensation, but that is the only other suggestion he offers beyond some limited form of punishment. 7 The idea of using such devices to monitor the location of such people as mental patients and those on probation or parole has been around since the early 1960s (see Schmidt and Curtis 1987: 137), and has become much more promising with the more recent development of global position system technology (for some assessments of the practice prior to that point, see, e.g., Vaughn 1987, Blomberg, Waldo, and Burcroff 1987, as well as Schmidt and Curtis 1987). 8 Intensive forms of probationary supervision are often assigned to intermediate offenders such as nonviolent felons. Although the details of such programs vary considerably, conditions that are commonly employed include such practices as frequent personal contact with a probation officer, drug and alcohol testing and counseling, payment of a supervision fee, and mandatory employment, as well as curfews, all of which could be justified on restitutive grounds if needed to help restore the community to its previous level of well-being. For a useful description of some of these programs, see Petersilia (1987: 15–19) and Clear, Flynn and Shapiro (1987: 31–41). For various views on the success of these programs see, in addition to these sources, Pearson (1987), Latessa (1987) and Bennett (1987). 9 A bill signed into law in New Mexico, for example, could be justified along the same restitution-based lines. The bill requires some repeat drunken drivers to install ignition locks on their cars: “The device, at a cost of $60 to $70 per month for the offender, requires the driver to blow into a tube connected to the ignition. If the driver’s blood-alcohol level is over 0.02 percent, the car will not start” (Sink 2002: A15). 10 See also Holmgren (1989: 143–4). The same position is defended as a debt-based theory of punishment in Hershenov (1999). I respond to this position more directly in section 5.13. 11 Note that Holmgren herself recognizes this. She accepts Benn’s definition of punishment on which “the unpleasantness is essential to it, not an accompaniment to some other treatment” (the definition is cited in 1983: 37). So, on Holmgren’s own terms, locking an offender up with the aim of restoring the community to its previous level of security cannot count as punishing an offender. See also Wood (1997) for a useful discussion of the way in which preventative detention need not be associated with punishment. 12 Emphasis added. 13 Golash, it should be noted, does acknowledge that some forms of intensive supervision and preventative detention can be consistent with a theory of pure restitution, but she does not recognize that they can be justified by the theory as well (e.g., 2005: 160, 166). 14 See, e.g., Schmidt and Curtis (1987: 141–2).

References Abel, Charles F. and Marsh, Frank H. (1984). Punishment and Restitution: A Restitutionary Approach to Crime and the Criminal. Westport, CT: Greenwood Press. Barnett, Randy (1977 [1985]). “Restitution: A New Paradigm of Criminal Justice.” In Punishment and Rehabilitation, 2nd ed., ed. Jeffrie G. Murphy. Belmont, CA: Wadsworth Publishing Co. Baylis, Charles A. (1968). “Immorality, Crime and Treatment.” In Philosophical Perspectives on Punishment, ed. Edward H. Madden, Rollo Handy and Marvin Farber. Springfield, IL: Charles C. Thomas.

142  David Boonin Bennett, Lawrence A. (1987). “A Reassessment of Intensive Service Probation.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda R. McCarthy. Monsey, NY: Criminal Justice Press. Blomberg, Thomas G., Waldo, Gordon P. and Burcroff, Lisa C. (1987). “Home Confinement and Electronic Surveillance.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Boonin, David (2008). The Problem of Punishment. Cambridge: Cambridge University Press. Clear, Todd R., Flynn, Suzanne and Shapiro, Carol (1987). “Intensive Supervision in Probation: A Comparison of Three Projects.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Dagger, Richard (1991). “Restitution: Pure or Punitive?” Criminal Justice Ethics 10: 29–39. Ellin, Joseph (2000). “Restitutionism Defended.” The Journal of Value Inquiry 34: 299–317. Golash, Deirdre (2005). The Case Against Punishment: Retribution, Crime Prevention, and the Law. New York and London: New York University Press. Hoekema, David (1991). “Trust and Obey: Toward a New Theory of Punishment.” Israel Law Review 25: 332–50. Holmgren, Margaret R. (1983). “Punishment as Restitution.” Criminal Justice Ethics 2: 36–49. Kleinberg, Stanley S. (1980). “Criminal Justice and Private Enterprise.” Ethics 90: 270–82. Latessa, Edward J. (1987). “The Effectiveness of Intensive Supervision with High Risk Probationers.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Loewy, Arnold H. (2000). Criminal Law in a Nutshell, 3rd edition. St. Paul, MN: West Group. Martin, Rex (1970). “On the Logic of Justifying Legal Punishment.” American Philosophical Quarterly 7: 253–9. McCarthy, Belinda R. (1987). “Introduction.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Miller, Franklin G. (1978). “Restitution and Punishment: A Reply to Barnett.” Ethics 88, no. 4 (July): 358–60. Pearson, Frank S. (1987). “Taking Quality into Account: Assessing the Benefits and Costs of New Jersey’s Intensive Supervision Program.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Petersilia, Joan. (1987). “Georgia’s Intensive Probation: Will the Model Work Elsewhere?” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Schmidt, Annesley K. and Curtis, Christine E. (1987). “Electronic Monitors.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press.

Punishment, Restitution, Incarceration  143 Sink, Mindy (2002). “New Mexico: Ignition Locks for D.W.I. Offenders.” New York Times, National ed., Friday, March 8, A15. Tunick, Mark. (1992). Punishment: Theory and Practice. Berkeley, CA: University of California Press. Vaughn, Joseph B. (1987). “Planning for Change: The Use of Electronic Monitoring as a Correctional Alternative.” In Intermediate Punishments: Intensive Supervision, Home Confinement and Electronic Surveillance, ed. Belinda McCarthy. Monsey, NY: Criminal Justice Press. Wood, David. (1997). “Reductivism, Retributivism, and the Civil Detention of Dangerous Offenders.” Utilitas 9, no. 1 (March): 131–46.

8 Communicative Theories of Punishment and the Impact of Apology Eddy Nahmias and Eyal Aharoni1

1. Introduction The incarceration system in the United States is broken. It does not effectively—much less efficiently—serve the forward-looking goals of punishment, which rely heavily on fear of incarceration. It neither rehabilitates criminals nor prevents recidivism effectively. Evidence suggests the current system fails to deter any more than shorter sentences or better conditions would (e.g., Cullen, Jonson, and Nagin 2011). And the incarceration system does not serve the backwards-looking retributive goals of punishment effectively either, in part because, as we discuss below, most people are not satisfied with punishment involving merely impersonal incarceration or the attendant suffering of wrongdoers, except in response to the worst crimes. Rather, we want punishment not only to send a message to wrongdoers but also for the message to be received, internalized, and acted upon. The US penal system is not effective at leading wrongdoers to recognize the harm they have done, to want to repair the harm, and to change so that they will not repeat such harms. Criminal behavior may indicate that the offender does not sufficiently value the community’s norms. Punishment might be most effective in achieving multiple aims if it expresses both this (backward-looking) message about the offender’s undervaluation of the community’s norms and the (forward-looking) message that the offender must increase his valuation of those norms. Here, we will argue that the communicative theory of punishment may best combine features of, and intuitions supporting, backward-looking and forward-looking theories of punishment, while also going beyond the expressivist theory’s limited aim of expressing the community’s condemnation of crime. We will then suggest some empirical predictions of such a theory and present a pilot study aimed to test some of these predictions. A background assumption of our discussion is that improving our broken incarceration system requires a better understanding of people’s punitive psychology, such that proposed reforms to the system can serve the functions of public safety and rehabilitation without requiring deviations from people’s punitive psychology that would risk a rejection of those reforms.

Punishment and Apology  145

2. The Communicative Theory as a Unifying Theory of Punishment The communicative theory of punishment states that the aim of punishment, including but not limited to incarceration and ‘hard treatment,’ is to communicate to offenders (and to the rest of the community) both the message that they have violated the norms of their community and the demand that they must respond to this message appropriately. Punitive acts ideally express the condemnatory message in ways that will lead offenders to repent their crimes, to repair harms they have caused, and to reform so that they can be trusted not to repeat their wrongdoing (see Duff 2001). Though most often advanced as a justificatory theory of punishment, the communicative theory provides a framework for a descriptive account of our punitive psychology that helps explain the diversity of people’s intuitions about just punishment, in part by explaining people’s close attention to the mental states of offenders both leading up to crimes and in response to being caught. We will focus here on the descriptive advantages of the communicative theory, highlighting some justificatory features as they arise. Below, we describe traditional theories of punishment in terms meant to be as neutral as possible between their being descriptive or justificatory. The traditional theories of punishment (except for ‘mixed theories’) are ‘one-directional.’ Pure retributive theories are solely backward-looking, stating that we punish in order to give offenders what they deserve for their past wrongdoing, typically meted out in terms of suffering proportionate to the harms they have willingly brought about. Such theories attend to past mental states of offenders to assess what punishment they deserve, since responsibility for outcomes depends in part on what the agent knew, intended, and controlled. But retributive theories do not attend to the offender’s mental states in response to punishment (except perhaps the amount of his suffering)—for instance, whether he better understands his community’s values. While some suggest that retributive punishment is most satisfying for victims and members of the community, the theory does not suggest a role for the responses of victims or the community to the punishment of offenders. And it may be, as we explain below, that people are most satisfied with punishment when any suffering it engenders serves to make the offender understand the nature of his crime and motivate him to change accordingly. Most rival theories to retributivism are purely forward-looking, stating that we punish in order to bring about particular consequences, such as specific or general deterrence, incapacitation, or rehabilitation (if one defines ‘punishment’ as requiring offenders to suffer pain or setbacks through hard treatment solely because they deserve it, then these rival theories will be alternatives to—not theories of—punishment). Consequentialist theories attend less closely to the past mental states of offenders, since they are not assessing desert. Instead, they consider mental states only to the extent that they are relevant in determining what it will take to deter or incapacitate

146  Eddy Nahmias and Eyal Aharoni actual or potential offenders or whether an offender has been rehabilitated sufficiently to release him. And they are concerned with victims and community responses only to the extent that these responses are relevant to general deterrence or acceptance of the legal system’s practices. None of these one-directional theories adequately captures people’s various interests in or attitudes towards punishment, including our obsession with the mental states of offenders—their intentions, beliefs, attitudes, and traits both leading up to their crimes and in response to their conviction and punishment. A more expansive theory is needed to unify the plurality of our punitive aims and intuitions. As Bedau and Kelly put it, “A strait-laced purely retributive theory of punishment is as unsatisfactory as a purely consequentialist theory with its counter-intuitive conclusions (especially as regards punishing the innocent). The practice of punishment, to put the point another way, rests on a plurality of values, not on some [single] value to the exclusion of all others” (2015). (See also the pluralistic purposes for sentencing outlined in the US Model Penal Code [American Law Institute §1.02.] and recent suggested revisions that reference both retributivist aims and restorative justice.) The communicative theory of punishment is ‘multi-directional’ in a way that can address some of these limitations of other theories. It is backward-looking, like retributive theories, in that it requires that punishment be communicated only towards those who deserve it because of what they actually did. And it suggests close attention to the mental states of the offender before and during the crime to assess what punishment he deserves to have communicated to him. To know what punishment is appropriate, we must assess the offender’s intentions, plans, and knowledge. The communicative theory also includes elements of the expressive theory of punishment, which assesses the guilt of the offender and nature of the crime in order to express a corresponding level of moral condemnation to him and to the community for the offense (e.g., Feinberg 1965). But unlike the expressive theory, the communicative theory takes punishment to involve more than just one-way expression of the community’s negative response to norm violations. The communicative theory goes further to demand a communicative response from the offender. And this responsive feature also explains its forward-looking aims. The punishment system commits the community (through its designated authorities) to carry out punitive acts that serve as directives or commands to the target offenders to carry out those directives. Hence, the communicative theory suggests that punitive acts aim to ach­ ieve outcomes, many of which overlap with the standard forward-looking theories of punishment. Punishment aims to alter the offender’s relevant mental states so that he increases his valuation of the community’s norms (his respect for the law). Hence, it is essentially rehabilitative, though not primarily by inducing fear of punishment, but rather by helping the offender become reintegrated into the community once he has signaled that

Punishment and Apology  147 he better understands and values its norms. Such signals suggest that two crucial features of communicative punishment will be evidence of remorseful apology and costly compensation. Given the facts of human psychology, these demands of remorse, rehabilitation, and reparation will require that the offender experience deprivation and suffering, and in some cases may be achieved best through the ‘hard treatment’ sometimes presented as the essential feature of punishment. However, the primary aim of punitive acts, on this view, is not merely or mainly the suffering suggested by retributive theories. Furthermore, current incarceration practices are poorly designed to encourage the outcomes suggested by the communicative theory. To demand of offenders that they reform, repent, or restore requires substantial reforms to the penal system so that it provides the means to allow that these communicative demands can in fact be met by convicted criminals. Indeed, communicative punishment treats offenders as autonomous members of the community—albeit, ones who have not properly valued its norms—rather than aiming to exclude them or treating them as ‘objects’ to be fixed, or as ‘outlaws’ literally outside the community’s legal norms. In this way, it avoids the potential dehumanization of criminals suggested by purely forward-looking theories of quarantine or rehabilitation (e.g., C.S. Lewis 1949). Rather than taking what P.F. Strawson (1962) calls the ‘objective stance’ towards offenders—aiming to manage, control, or fix them—communicative punishment treats offenders from within the ‘participant stance’ with the expectation that they have the cognitive capacities to respond to the demands communicated to them. (The connections between the communicative theory of punishment and Strawson’s influential theory of moral responsibility have not been sufficiently explored, but see McGeer 2011.) In treating offenders as persons who are members of the community, communicative punishment might be more effective at changing their behavior over the long term than punishments targeted directly and exclusively at behavior modification. Offenders motivated to internalize social norms regulate themselves, whereas motivating them with extrinsic reinforcers will only work as long as those reinforcers are present. Nonetheless, people will be inclined to take the objective stance towards some criminals, especially those who commit the most heinous crimes, who recidivate, who are perceived as lacking relevant psychological capacities, or who are perceived as members of out-groups or having little potential value to the community. Some of these offenders may in fact lack the capacities to receive or respond to the message communicated by punishment. It may be that the communicative theory must ‘bifurcate’ and allow for different types of treatment for some of these criminals, just as Strawson allows that it will be appropriate to take the objective stance towards some who wrong us (see Duff 2001: 166). However, the ‘worst of the worst’ pose challenges for every theory of punishment (e.g., forward-looking theories may require permanent quarantine; retributivism may suggest execution). But the communicative theory can, in practice, suggest ways to increase the

148  Eddy Nahmias and Eyal Aharoni scope of those offenders whom we are inclined to see as appropriate targets of inclusive communication rather than exclusory quarantine or dehumanizing treatment. This process will likely require encouraging people to see fewer criminals as irredeemable or as members of an out-group and more as members of the community with whom communication is possible, with the hope that it can lead to reintegration. Doing so will presumably require reforms to the way the legal system describes and treats crimes, criminals, and punishment.

3. The Psychological Plausibility of the Communicative Theory The communicative theory of punishment seems to accord well with humans’ punitive psychology. First, it is an extension of the attitudes and expressions used in our interpersonal relations when others treat us in ways that show a lack of proper valuation (or respect) for our relationship with them. When our friends, family members, or colleagues violate our normative expectations, we feel and express indignation or resentment, which communicates to them demands for expressions of remorse, potentially for reparative actions, and especially for commitment to better future behavior. Our ultimate aim is typically not to make them suffer, except insofar as they experience negative emotions associated with remorse or the setbacks required to make up for their wrongs. The aim of these emotions from the offender and the offended—from moral anger to remorse (what Strawson calls the ‘reactive attitudes’)—is ultimately to motivate us to repair relationships. To the extent that we are able to see criminal offenders as members of the community and aim to reintegrate them, the communicative functions of punishment are extensions of, not replacements for, our interpersonal punitive attitudes and responses. Second, the communicative theory resonates well with evolutionary models of punishment. According to these models, punitive motivations evolved to protect members of the social group against those who threaten the norms that enable the group to thrive. Those who were unwilling to support the punishment of cheaters would have suffered a fitness cost at the hands of those cheaters, breaking down the benefits of participating in a social group. For example, if some members consistently reap the spoils of group hunting without contributing to the hunt, and without facing punitive sanctions, these cheating behaviors will be selected over cooperative behaviors. Hence, punitive responses would be selected for to allow the possibility of large-scale cooperation. In the long term, potential cooperators are valuable enough to the community that efforts to try to reintegrate cheaters back into the community could be more beneficial than the costs of trying to ostracize or kill them. If so, then these pressures would likely select for punitive sentiments that motivate us to punish in a way that appeals not just to cheaters’ fears of

Punishment and Apology  149 retaliation, but to their internal values. Specifically, these responses should signal to cheaters that they have undervalued the norms of the community and must update that valuation in order to reap the future benefits of community membership (Petersen et al. 2012; Sell 2005; Sznycer et al. 2016; Tooby et al. 2008; Tooby and Cosmides 2008). In this context, it would be useful to be able to accurately assess, and possibly increase, the capacity and willingness of offenders to internalize the social norms. Humans’ strong interest in gauging offenders’ intentions, previous behavior, and sincerity of apology seem, therefore, to accord with the punitive psychology suggested by communicative punishment. Ultimately, the forward-looking selective processes that allowed large-scale cooperation in our ancestors may have given us the backward-looking reactive attitudes that motive punishment. But the goal of these punitive emotions and the actions they motivate is to appraise and communicate the need and possibility for reconciliation. Furthermore, the communicative theory of punishment suggests that our punitive responses should be sensitive to whether the offender understands the message. Indeed, research supports this hypothesis. For example, Gollwitzer and colleagues (2009, 2014) found that participants were more satisfied with their punitive acts when they received evidence that the offender understood why he was being punished and even more so when he indicated even a minimal commitment to avoid future wrongdoing. These studies challenge some earlier studies (e.g., Carlsmith, Darley, and Robinson 2002; Darley, Carlsmith, and Robinson 2000) suggesting that people are motivated to punish primarily by the retributive aim of harming the offender and to vent their anger (see Nadelhoffer et al. 2013). The Gollwitzer et al. studies, however, involved economic games, so the stakes were relatively low and the signals were minimal (e.g., “OK—I was greedy . . . and now I see what’s wrong with that. . . . I shouldn’t be such a jerk in a situation like this”). Whether our desire to motivate such understanding of wrongdoing in the offender in these low-stakes contexts also applies to more serious wrongdoing has not been sufficiently studied. Thus, one motivation for our study is to test the effects of such responses to punishment, in the form of apologies, in the context of criminal wrongdoing.

4.  Testing the Impact of Apology on Punishment Judgments The communicative theory suggests a range of empirical predictions that would be useful to test. A few of these have been explored (e.g., research on restorative justice programs, reparations, and apologies; e.g., Schinkel 2014; Gold and Weiner 2000; Jehle, Miller, and Kemmelmeier 2009; Peterson et al. 2012; Petrucci 2002; Rachlinski, Guthrie, and Wistrich 2012; Robinson, Jackowitz, and Bartels 2012; Scher and Darley 1997; O’Hara and Yarn 2002). Here, we focus on how information about relevant mental states of a criminal offender impacts punitive judgments, specifically information from which people can make inferences about the intentions and

150  Eddy Nahmias and Eyal Aharoni planning of the offender in carrying out the crime, about his character based on criminal history, and about the sincerity of his apology in response to his crime. To determine whether the offender adequately values community norms, it is necessary to appraise his initial motivation and to rule out relevant excuses or justifications for the violation. For example, it would be useful to know if the offense was intentional in a way that indicates recognition of and disregard for the risks of harm to others. Those who commit the same act opportunistically, for instance, likely do not undervalue the community’s norms as much, and therefore may be less likely to reoffend than someone who has planned their crime. Thus, punishing people whose violations were less intentional would be less efficient. Similarly, to be sure that a punishment will be understood and will effectively reform the offender, we would need to know whether the offender is capable of internalizing the community’s norms. One way to do this is by evaluating the offender’s offense history. A consistent pattern of norm violations indicates greater future risk, in part because it reveals something about that offender’s characteristic (low) valuation of others. If punishment is aimed at communicating a message to the offender, then people should respond to evidence that the offender will be responsive to that message. Another way to increase confidence that punitive messages will be internalized by offenders is to evaluate their commitment to changing their behavior. According to the communicative theory, one of the most credible ways for an offender to signal relevant changes of mental state and future behavior is with a sincere apology. Apologies can include various elements which provide evidence for different mental states and levels of sincerity. These elements include: (1) acknowledgment that one’s action violated a norm (i.e., was wrong) and of the harm it caused; (2) an expression of remorse for one’s action and its harmful outcomes; (3) a promise to avoid repeating such actions; and (4) an offer to make up for the harm caused by the action (e.g., compensation). Apologies might also include some explanation for why one did the action (though sincere apologies do not include attempts to excuse or justify one’s behavior), emotional expressions such as sorrow or shame, a request for forgiveness, or an expression of hope for an improved relationship (Scher and Darley 1997; Petrucci 2002; O’Hara and Yarn 2002). These theoretical considerations lend themselves to specific predictions, namely that punishers will be responsive to information about criminal intent, criminal history, and the perceived sincerity of apology. We developed a pilot study to examine these three features of people’s punitive psychology and their influence on judgments about punishment in response to crime. While previous research has examined each of these features, we consider how they act in combination to influence people’s punishment judgments. To examine the influence of intent (1), we varied whether the criminal planned his crime (robbery) or whether he carried it out opportunistically.

Punishment and Apology  151 To examine the influence of past wrongdoing (recidivism) (2), we varied whether or not the criminal had been convicted and served a sentence for a similar crime. And to examine the influence of apology (3), we varied whether the criminal offers few or most elements of apology for his crime, suggesting an insincere versus a sincere apology. We manipulated these factors between subjects (high vs. low) and also within subjects (pre vs. post) and measured the effects on judgments of sentence severity and length. All our manipulations were designed to be relatively subtle in order to limit the influence of potential confounds and isolate the unique effects of the relevant features of each manipulation. Notice that if people are punishing for expressivist or retributive reasons, it is not clear why they should or would lower punitive responses in response to apology or even offers of compensation. If the point of punishment is to express our condemnation for the crime or to cause the criminal to suffer in proportion to the harm he caused, then the criminal’s response should not be relevant to punishment, unless perhaps it is taken as evidence of diminished intent or as signaling diminished need to express condemnation. Pure retributive theories would also predict that people punish in response to what is deserved for the specific crime in question, and hence should not increase punishment in light of earlier crimes so long as those crimes had already been justly punished. Furthermore, on a general deterrence theory, apology might be seen as problematic, since lowering punishment in response to apologies or reparations might signal to would-be perpetrators that they have an ‘escape valve’ if they are caught. This might be part of the reason that judges appear unlikely to lower sentences in response to apology (see Rachlinski et al. 2012). On a specific deterrence or a rehabilitative theory, our reactions to apology should be based solely on our taking them to provide reliable information about recidivism. Since the communicative theory is also concerned with offenders’ responding to punishment by reforming and committing not to recidivate, it will be difficult to discern which of these theories best explains people’s responses to apologies unless we can discern whether offenders’ improved behavior is due primarily to fear of punishment or internalization of the community’s norms. These predictions motivated the pilot study described below. Participants read a realistic case summary of a crime that has resulted in a conviction and were asked to make punishment recommendations, before and after receiving information about the offender’s prior crimes and the type of apology he offers.

5.  Study Design Participants were 601 undergraduates (67.2% female; 30% male) from Georgia State University participating for course credit in philosophy (n = 493) or psychology (n = 108). Although the punitive attitudes of GSU

152  Eddy Nahmias and Eyal Aharoni undergraduates are not expected to represent that of juries or judges, this population is valuable for the purpose of assessing folk psychological punitive attitudes and piloting new experimental methods. Furthermore, compared to many US universities, the sample was demographically diverse with respect to age (M = 22.7, SD = 3.9; 57.7% between age 18–22), race (Black: 35.9%; White: 29.1%; Asian: 17.3%, Mixed/Other/Unknown/Prefer not to answer: 16.6%) and ethnicity (Hispanic/Latino: 13.7%; Mixed/Other/ Unknown/Prefer not to answer: 7.9%). The study used a 2 (intent) × 2 (criminal history) × 2 (apology) mixed factorial design, permitting measurement of differences in recommended punishment both between- and within-subjects. Intent was defined as the offender’s level of criminal intent being “high” (planned crime) versus “low” (opportunistic crime). Criminal History was defined as the presence or absence of a prior criminal record (i.e., “second-time” versus “first-time” offender). Apology was defined as the presence of a “sincere” versus “insincere” verbal apology by the offender. All participants were presented with the information in the order Intent, Criminal History, Apology. Our primary dependent measure was a sentence length measure describing number of months in prison on an internally developed, logarithmic scale from 1 year to 20 years and 2 months. The within-subjects comparison was achieved by delivering the sentencing measure three times, once after presentation of each of the three independent variables. This design feature, though it departs from actual legal procedure, enabled us to measure each participant’s change in punitive sentiment as he/she learns more about the details of the offender’s history and apology, thereby revealing something about the underlying cognitive updating process likely involved in punishment evaluations. We tested the following hypotheses:

• H1: High intent will evoke more severe punishments than low intent •

•

(i.e., main effect of intent). H2: Within-subjects punishment will increase following presentation of prior criminal history (i.e., for second-time offenders). Likewise, second-time offenders will evoke more severe punishments than first-time offenders (i.e., main effect of criminal history). H3: Within-subjects punishment will decrease in the presence of a sincere apology. Likewise, the sincere apology will evoke lesser punishment than the insincere apology (i.e., main effect of apology).

The study design yielded eight vignette-based surveys, corresponding to each of the eight between-subjects conditions. (See Appendix for vignette stimuli.) The survey was administered online (Qualtrics) via personal computers. For each survey, after providing consent, participants were instructed to imagine that they are serving as a trial court judge for the sentencing phase of a criminal case. The participant’s task was to consider the details of the case and

Punishment and Apology  153 produce a sentence recommendation. All vignettes were closely matched for length and reading level. The key dependent measure, sentence length, was introduced using the following instructions: “According to Georgia law, a person convicted of robbery is to be punished by imprisonment for not less than 1 nor more than 20 years. Given the following options for sentencing, how much time in prison should Mr. Jones receive for this offense?” For the second and third punishment measurements, participants were told that before their initial sentence recommendation had gone into effect, new facts about the case became available for consideration—i.e., information first about criminal history and then about apology. In addition to our key dependent measure, several supplemental questions were examined to identify the factors most likely to influence and explain our primary effects. These included questions designed to capture punitive attitudes pertaining to the vignette, including deontological (e.g., the moral wrongfulness of the crime), consequentialist (e.g., the offender’s dangerousness), and communicative (e.g., how much the offender understands the purpose of the punishment) attitudes measured on a 9-point scale. We also queried participants’ philosophical attitudes about the purpose of punishment more generally, using a scale (adapted from Nadelhoffer et al. 2013), asking participants to rank the importance of six justifications for punishment such as giving offenders “what they deserve” and sending offenders “the message that they should feel genuinely remorseful and should make up for the harm they caused.” We also included questions to assess the effectiveness of our manipulations such as the degree of agreement with the statement “Mr. Jones’ apology was sincere.” Finally, participants were asked to provide demographic information (e.g., gender, age, race and ethnicity, political orientation, religiosity).

6.  Does Apology Reduce Punishment? This study was designed to determine whether apology mitigates punishment independently of criminal intent and criminal history. To answer this question, we examined two types of comparisons. First, we sought to evaluate the within-subjects changes in punishment before vs. after presentation of prior criminal history and before vs. after presentation of the (in)sincere apology. Then, we sought to understand whether the changes in sentences from pre- to post-criminal history and pre- to post-apology differed between experimental groups. We tested our hypotheses using a mixed Analysis of Variance with Intent (planned vs. opportunistic), Criminal History (1st time vs. 2nd time), and Apology (sincere vs. insincere) as between-subjects factors and Sentence Recommendation as a within-subjects factor. The sentence recommendation was collected at three points: (1) following the Intent, (2) Criminal History, and (3) Apology manipulations, respectively. We used planned comparisons to evaluate within-subjects changes in punishment independently for each level of our independent variables (corrected for multiple comparisons).2

154  Eddy Nahmias and Eyal Aharoni (H1) Did high intent evoke more severe initial punishments than low intent? Yes. Prior to the introduction of criminal history or apology information, participants in the high intent condition responded with longer sentences (M = 4.23 years) than participants in the low intent condition (M = 3.33 years), p < .01. (H2) Did participants increase punishment following presentation of prior criminal history (i.e., for second-time offenders)? Yes. Punishment was greater following presentation of prior criminal history (M = 5.39 years) compared to that preceding criminal history information (M = 3.36 years), p < .001, and this pre/post increase was only evident when that information described a second-time offender (MD = 2.03 years) rather than the first-time offender (MD = -0.54 years) p < .001. (H3) Did participants decrease punishment following presentation of a sincere apology? Yes. Punishment was lower following either apology (M = 4.02 years) compared to that preceding apology information (M = 4.54 years), at least under conditions of high criminal intent, p < .05. Furthermore, this decrease in punishment was greater for the sincere apology (MD = -0.52 years) than for the insincere apology (MD = -0.02 years), but only in the high intent condition, p < .05. (See Figure 8.1.) In addition to our primary hypotheses, several supplemental questions were examined to identify factors likely to clarify our primary effects. Most notably, participants assigned to the sincere apology condition were significantly more likely (M = 1.60) than those in the insincere condition (M = 0.32)

Figure 8.1 When the offense was committed with high criminal intent, a sincere apology significantly mitigated punishment relative to an insincere apology.

Punishment and Apology  155 to agree with the statement that: “Mr. Jones seems to understand why he needs to be punished” (p < .001). The same participants were less likely to agree that Mr. Jones will commit another crime after release (p < .05). Furthermore, we asked participants to rank the importance of six justifications for punishment. On average, participants ranked specific deterrence and communication of a message as the two most important justifications. These two justifications, while statistically indistinguishable from each other (p = .44), were ranked proportionally higher than their nearest neighbor (p’s < .001) according to a series of Wilcoxon Signed Rank tests with correction for multiple comparisons. (See Figure 8.2.) These results are consistent with the predictions of communicative theory that sincere apology should mitigate punishment by sending a credible message that the offender values the societal norms and will make efforts to compensate for the harms he has caused. However, the mitigating effect of apology is not large in comparison to the aggravating effect of criminal history. To the extent that apology communicates a commitment to play by the rules, the small effect should not be surprising alongside evidence of recidivism. That is, the presence of a criminal history could undermine the credibility of an apology in the way that “actions speak louder than words.” Yet, even with evidence of recidivism, we still observe a mitigating effect of sincere apology. The intent by apology interaction was not predicted. We found that sincere apology was more effective in lowering sentencing judgments when

Figure 8.2 The two highest ranking punishment justifications were to deter offenders from reoffending and to send a message to offenders that they should feel remorseful and make up for the harm caused.

156  Eddy Nahmias and Eyal Aharoni intent was high. One possible explanation for this effect is that for the milder, low-intent offense, participants believed that there was less to apologize for. The absolute values of participants’ sentences for the low-intent crime also tended to be low, so these participants had less room on the scale to reduce their sentence. However, future research would be required to evaluate these explanations. We limited the apology manipulation to a subset of established elements of apology. Specifically, the sincere apology uniquely included (1) an acknowledgement that the action was wrong and harmful, (2) an expression of remorse for causing the harm, (3) a promise to desist from such actions, and (4) a voluntary offer of costly compensation. However, the two apologies did not differ in the compensation outcome (the money was returned in both scenarios) or explicit request for forgiveness (no such request was made). These two elements may influence the mitigating effect of apology. We also presented minimal information about the emotions expressed in the sincere apology (or the victim’s emotions in response to apology), elements which have been shown to play a significant role in impacting punitive responses (Gold and Weiner 2000).

7.  Future Research and Concluding Suggestions We plan to consider several directions in future research. Extensions of this line of research should test the responses of those who actually make punitive judgments, including lawyers and ideally criminal trial court judges, permitting a direct comparison between experts’ and laypersons’ punitive intuitions. If the communicative theory of punishment is descriptively on track, it should be detectable even within the constraints that professional judges must oblige, such as federal sentencing guidelines. It would also be useful to test the effects of offender’s apologies and offers of reparation made directly to the victim of the crime, and of the victim’s accepting the apologies and offers and/or offering forgiveness versus not doing so (e.g., as occurs in some restorative justice procedures). The effects of more detailed descriptions of (or even video presentations of) emotional expressions by offender and victim would also be useful to explore. The communicative theory of punishment assumes that punishment is designed to reform eligible individuals within the social community. As such, it is less clear about how we are likely to treat outsiders who violate the social norms of our ingroup. On one hand, we might not expect outgroup members to be familiar with our ingroup norms; on the other hand, we have less reason to invest in outgroup members in the first place. Thus, laypersons should be adept at distinguishing between ingroup and outgroup members, and should employ different types or amounts of punishment to achieve their different goals for outgroup members. It would hence be useful to test the effects on punitive judgments of participants’ perception of offenders as members of the ingroup or outgroup.

Punishment and Apology  157 The communicative theory of punishment has been neglected in discussions of penal reform, perhaps because it is not obvious what practical reforms it suggests, though theorists have discussed reforming and expanding the parole system and various restorative justice programs (e.g., Duff 2001; Schinkel 2014). To the extent that the theory suggests increased attention to the mental states of offenders before and during their crimes and after conviction of their crimes, it may face practical concerns about how best to assess these mental states with objective measures. Furthermore, there are procedural hurdles to the use of apologies in the legal system. Since a crucial element of apology is the offender’s recognition and communication of his guilt, apology before conviction will be strongly discouraged by defense attorneys, except as part of a plea bargain, in which admission of guilt typically leads to sentence reduction—e.g., an automatic 2–3 level reduction in sentence (2012 United States Sentencing Commission §3E1.1). Apologies after conviction can seem—and can be—insincere if they are perceived as—or aimed at—an attempt to minimize punishment (see Rachlinski, Guthrie, and Wistrich 2012). Relatively recent attempts to incorporate apology into restorative or reparative justice frameworks offer some practical guidance for reforms. However, these frameworks are typically presented as, and understood as, alternatives to punishment. One feature of the communicative theory of punishment is that punitive acts aim to encourage, or even force, offenders to apologize, to compensate, and to take on the work of rehabilitation. If we reform our broken incarceration system to more effectively induce offenders to understand the norms they have violated, to increase their valuation of them, and to make efforts to compensate for violating them, such reforms will not be replacing punishment. They may instead be moving us towards what we already believe punishment aims to do.

Notes 1 Authorship is equal. We owe much gratitude to Julia Watzek, who provided valuable assistance in survey design, analysis, and reporting. 2 We also tested absolute differences in punishment between groups. However, unlike our within-subjects models, the corresponding between-subjects test yielded no significant differences between sincere vs. insincere apology. Given the lower sensitivity of between-subjects comparisons for small effects, we speculate that these tests may have suffered from Type II error. Additional measures were collected which are beyond the scope of the present research question and so are not reported here.

References Bedau, Hugo and Kelly, Erin. (2015). “Punishment.” In The Stanford Encyclopedia of Philosophy, ed. E. Zalta. Available at http://plato.stanford.edu/archives/ fall2015/entries/punishment. Carlsmith, Kevin M., Darley, John M. and Robinson, Paul. (2002). “Why Do We Punish? Deterrence and Just Deserts as Motives for Punishment.” Journal of Personality and Social Psychology 83: 284–99.

158  Eddy Nahmias and Eyal Aharoni Cullen, Francis T., Jonson, Cheryl L. and Nagin, Daniel S. (2011). “Prisons Do Not Reduce Recidivism: The High Cost of Ignoring Science.” The Prison Journal 91: 48S–65S. Darley, John M., Carlsmith, Kevin M. and Robinson, Paul. (2000). “Incapacitation and Just Deserts as Motives for Punishment.” Law and Human Behavior 24: 659–83. Duff, Anthony. (2001). Punishment, Communication, and Community. New York: Oxford University Press Feinberg, Joel. (1965). “The Expressive Function of Punishment.” The Monist 49: 397–423. Funk, Friederike, McGeer, Victoria and Gollwitzer, Mario. (2012). “Get the Message: Punishment Is Satisfying If the Transgressor Responds to Its Communicative Intent.” Personality and Social Psychology Bulletin 40: 986–97. Gold, Gregg and Weiner, Bernard. (2000). “Remorse, Confession, Group Identity, and Expectancies About Repeating a Transgression.” Basic and Applied Social Psychology 22: 291–300. Gollwitzer, Mario and Denzler, Markus. (2008). “What Makes Revenge Sweet: Seeing the Offender Suffer or Delivering a Message?” Journal of Experimental Social Psychology 45: 840–4. Jehle, Alayna, Miller, Monica and Kemmelmeier, Markus. (2009). “The Influence of Accounts and Remorse on Mock Jurors’ Judgments of Offenders.” Law and Human Behavior 33: 393–404. Lewis, C.S. (1949). “The Humanitarian Theory of Punishment.” Twentieth Century: An Australian Quarterly Review 3: 5–12. McGeer, Victoria. (2011). “Co-Reactive Attitudes and the Making of Moral Community.” In Emotions, Imagination, and Moral Reasoning, ed. R. Langdon and C. Mackenzie. New York: Taylor & Francis. Nadelhoffer, Thomas, Heshmati, Saeideh, Kaplan, Deanna and Nichols, Shaun. (2013). “Folk Retributivism and the Communication Confound.” Economics and Philosophy 29: 235–61. O’Hara, Erin and Yarn, Douglas. (2002). “On Apology and Consilience.” Washington Law Review 77: 1121–91. Petersen, Michael B., Sell, Aaron, Tooby, John and Cosmides, Leda. (2012). “To Punish or Repair? Evolutionary Psychology and Lay Intuitions about Modern Criminal Justice.” Evolution and Human Behavior 33: 682–95. Petrucci, Carrie J. (2002). “Apology in the Criminal Justice Setting: Evidence for Including Apology as an Additional Component in the Legal System.” Behavioral Science and the Law 20: 337–62. Rachlinski, Jeffrey, Guthrie, Chris and Wistrich, Andrew J. (2012). “Contrition in the Courtroom: Do Apologies Affect Adjudication?” Cornell Law Review 98: 1189–243. Robinson, Paul, Jackowitz, Sean and Bartels, Daniel. (2012). “Extralegal Punishment Factors: A Study of Forgiveness, Hardship, Good Deeds, Apology, Remorse, and Other Such Discretionary Factors in Assessing Criminal Punishment.” Vanderbilt Law Review 65: 737–826. Scher, Steven and Darley, John. (1997). “How Effective Are the Things People Say to Apologize? Effects of the Realization of the Apology Speech Act.” Faculty Research and Creative Activity. Paper 26.

Punishment and Apology  159 Schinkel, Marguerite. (2014). “Punishment as Moral Communication: The Experiences of Long-Term Prisoners.” Punishment & Society 16: 578–97. Sell, Aaron. (2005). “Regulating Welfare Tradeoff Ratios: Three Tests of an Evolutionary-Computational Model of Human Anger.” Doctoral dissertation, University of California Santa Barbara. Strawson, Peter. (1962). “Freedom and Resentment.” Proceedings of the British Academy 48: 1–25. Sznycer, Daniel, Tooby, John, Cosmides, Leda, Porat, Roni, Shalvi, Shaul and Halperin, Eran (2016). “Shame Closely Tracks the Threat of Devaluation by Others, Even Across Cultures.” Proceedings of the National Academy of Sciences 113: 2625–30. Tooby, John and Cosmides, Leda. (2008). “The Evolutionary Psychology of the Emotions and Their Relationship to Internal Regulatory Variables.” In Handbook of Emotions, 3rd ed., ed. M. Lewis, J. Haviland-Jones and L. Barrett Feldman. New York: Guilford Press. Tooby, John, Cosmides, Leda, Sell, Aaron, Lieberman, Debra and Sznycer, Daniel (2008). “Internal Regulatory Variables and the Design of Human Motivation: A Computational and Evolutionary Approach.” In Handbook of Approach and Avoidance Motivation, ed. A. Elliot. New York: Taylor and Francis.

Appendix Case Summary (Differences between Conditions Are in Bold)

Frank Jones has been convicted of one count of robbery. At trial, the following evidence was presented. On the evening of March 13, 2016, thirty-year-old Mr. Jones entered Wilson’s Convenience Store in Alpharetta, Georgia. Mr. Jones is seen on video surveillance walking into the store at 10:37 p.m. Mr. Jones took an item from the refrigerator and brought it to the counter. Mr. Jones handed a bill to the cashier, who was the only other person in the store. The cashier opened the register and then walked into an adjacent workroom to get change. Mr. Jones waited approximately 15 seconds, looked around the empty store, then walked around the counter and put all the bills from the cash register into his pockets. The cashier then returned to the area behind the counter, standing in between Mr. Jones and the exit. Mr. Jones pushed the cashier out of his way, and she crashed into a glass display case, breaking the glass and losing consciousness. Mr. Jones then fled through the exit door. The cashier was found unconscious at 11:04 p.m. and taken to the nearest hospital, where she received four stitches. Based on the store’s electronic transaction history, the amount stolen was $674. Low intent (opportunistic crime)

High intent (premeditated crime)

Mr. Jones’ phone was investigated for any indication that the robbery was planned, but no evidence of planning was found. Investigators found only one relevant text message from a friend of Mr. Jones asking him to pick up some milk from the store.

Mr. Jones’ phone was investigated for any indication that the robbery was planned, revealing a text message from earlier in the day in which he bragged to a friend about his plans to “strong arm” the store when he knew there was only one cashier on duty.

Additional Facts Before you announce Mr. Jones’s sentence, you are given additional facts about the case. First, you are informed that after he was found guilty of robbery, prosecutors discovered that. . .

Punishment and Apology  161 Low Criminal History

High Criminal History

Mr. Jones has never before been convicted or arrested for any other crimes in any US state.

Mr. Jones has been previously convicted of another robbery in a different US state and served his full prison sentence for that crime.

At the end of the current trial, Mr. Jones was asked if he would like to make a statement to the victim in the courtroom. He provided the following statement: Sincere Apology

Insincere Apology

I am truly sorry for causing you all this pain. I understand that what I did was wrong. There’s no excuse for my behavior. I can’t undo what I did, but I promise I will do what it takes so that I will never do that kind of thing again. Also, I will work to repay you fully, not just for the money I stole, but also for the pain I caused.

I am sorry about what happened to you. I didn’t think it would turn out this way. I can’t undo what I did, but I definitely don’t want to go to prison. Also, you’ll get your money back because they got it out of my car when I was arrested.

9 A Reparative Approach to Parole-Release Decisions Kristen Bell

When Richard Martin1 was a teenager, he ran away from an abusive home. A 46-year-old man took him in, and repeatedly drugged and raped him. When Richard was 16, his abuser told him to kill a man and steal a truck. He did so. Three weeks later, Richard brought himself to a police station and confessed. He was convicted of murder and sentenced to life in prison with the possibility of parole. A sentence of life with the possibility of parole (“term-to-life” sentence) means that Richard had to serve a set number of years in prison, after which he became eligible for a hearing in which a parole board could grant or deny release on parole. Unlike most states, California recognizes a legal presumption in favor of granting release on parole which can be defeated by some evidence that a person poses a current danger. The parole board decided to approve Richard for release on three separate occasions, and each time the decision was reversed. In 2013, California passed Senate Bill 260, a statute requiring that parole hearings be a “meaningful opportunity for release” for people who committed crimes at a young age. The law essentially put a thumb on the scale in favor of release at youthful offender parole hearings. Over 250 people have been released from prison under this law to date, and it was expected that Richard would be among them (Kuznia 2016). Despite the fact that Richard has a clean record in prison and has engaged in almost all the rehabilitation programs that the prison has to offer, the parole board again denied his release. As of this writing, Richard is 55 years old and remains incarcerated in a California State Prison. He struggles with depression; as he has told the parole board, he has lost faith in the process. Richard’s story is not unique, a number of similar stories are chronicled in “False Hope,” a report by the American Civil Liberties Union that details the broken nature of the parole-release process across the nation (Mehta 2016). In California, one-third of the entire prison population is serving a term-to-life sentence (“lifers”), meaning they will die in prison unless the parole board approves their release (Nellis 2013). Across the nation, over 110,000 people are serving term-to-life sentences (Nellis 2013: 6). “The

Parole-Release Decisions  163 number of people serving life sentences has more than quadrupled since 1984—a faster rate of growth than the overall prison population.” (Ghandnoosh 2017: 3). Procedural protections at parole hearings are minimal and variable (Ruhland et al. 2016: 23), as is the rate at which parole boards approve prisoners for release (Mehta 2016; Schwartzapfel 2015). In California, for example, the parole board granted parole in approximately 27% of all the lifer parole hearings that it conducted in 2015, but almost no one was released under the reign of Governor Gray Davis in 1999–2003 (Board of Parole Hearings 2015; Kobrin 2005; Weisberg 2011: 15). In Florida, the parole grant rate is less than one percent, and in Ohio the grant rate ranges between four percent and nine percent (Mehta 2016: 46). In stark contrast, the Arkansas parole board granted parole in 71 percent of the cases it reviewed in 2015 (Mehta 2016). Many state parole boards also have release authority over hundreds of thousands of prisoners who are serving sentences that are shorter than term-tolife (Ruhland et al. 2016: 14). In non-term-to-life cases that go before the parole board, people are generally serving sentences within a range fixed by a judge (for example, 7–10 years), and the parole board decides on a release date at some point within that range. The rate at which parole boards release people before the maximum date of their term is generally higher than the rate of release in term-to-life cases. For example, the Georgia parole board approved release for 11 percent of people serving term-to-life sentences, but its general grant rate (including non-life sentences) was 56 percent (Mehta 2016: 47). Of note, however, research from one state shows that the parole grant rate among non-life sentences is disparate across racial lines: a study of the general parole grant rate in New York State in 2013–2016 showed a release rate of 25% for white prisoners and 15% for African American prisoners (Winerip 2016). An older study of the New York parole system found that the variability and unpredictability of parole-release decisions increased as the severity of the crime increased (Vera Institute of Justice 1978). The arbitrary and politicized nature of parole-release systems is an easy target to criticize, but proposals for effective reform are few and far between. Some consider reform hopeless and have advocated for abolishing parole-release systems in favor of sentences with pre-determined end dates and specific good-time reduction credits (American Bar Association 1994). The federal government and several states abolished or otherwise contracted parole-release systems in the 1970s through the 1990s (Rhine 2012: 631–632), but many of these jurisdictions have seen increases in the amount of time that people serve in prison and little (if any) improvement in the fairness of sentencing (Nellis 2013). In Louisiana, for example, people serving term-to-life sentences in prison used to be regularly released on parole after having served a decade in prison, but after the abolition of parole-release, 4,657 people are now serving life without the possibility of parole sentences (11% of Louisiana’s prison population). Of those 4,657 people, 73.4% are

164  Kristen Bell African American (Nellis 2013). In recent years, states that had previously contracted parole-release systems are once again expanding them (Ruhland et al. 2016: 15). In 2016 and 2017, advocates and scholars have proposed several recommendations for meaningful reform of the parole-release process (Mehta 2016; Rhine 2017; Ghandnoosh 2017). A primary objective of the recommendations is to curtail discretion in parole-release decisions; as one expert has said, parole-release decisions have long been the most visible display of discretion in the criminal justice system (Rhine 2012). To reduce arbitrary decision making, the recommendations propose depoliticizing parole boards, reducing the amount of time served before the first parole hearing, increasing procedural protections at hearings, and establishing a strong presumption in favor of release. Left largely unexplored, however, is the normative project of analyzing what substantive criteria ought to guide parole-release decisions. I undertake this normative project here, reviewing how four different moral theories of punishment would structure parole-release criteria. I consider two general objections to each of these approaches, and then turn to draw lessons from how one state’s parole-release system functions in practice. Learning from both the theoretical and practical perspectives on the question of parole-release criteria, I suggest that a fundamentally different approach to parole-release criteria is required. I propose a “reparative approach” that builds on aspects of restorative justice and takes seriously respect for the moral agency of prisoners, victims, and the broader political community.

Four Theoretical Approaches to Parole-Release Decision Criteria In this section, I describe the approach to parole-release decision criteria from the perspective of four received theories of punishment: a retributive theory, deterrence theory, rehabilitation theory, and communicative theory. I do not sketch multifaceted policy proposals, but seek only to identify the basic decision-release criteria that are grounded in the core commitments of these theories of punishment. Although the four theories of punishment I consider purport to provide a moral justification of punishment, it is an open question whether any of them actually justify the contemporary practice of term-to-life sentences or other prison sentences with the option of parole-release. I do not address the moral justifiability of such sentences here, but assume for the purpose of the paper that they are a relatively fixed point in the American landscape for the foreseeable future. The project here is to identify what guidance (if any) these theories of punishment can provide in the design of parole-release decision criteria. I aim to be ecumenical regarding theories of punishment in this paper; that is, I do not herein object to any of the theories themselves.

Parole-Release Decisions  165 I argue in the next section, however, that each of the parole-release decision criteria are objectionable on two basic grounds. In focusing on how different theories of punishment would inform parole-release decision criteria, I am suggesting that the central question in making a parole-release decision is when and whether punishment should end. The central question, however, may be considerably more complex, depending on how one conceives of punishment and the function of parole.2 Arguably, punishment does not end when a person is released on parole. When a person is granted release on parole, she is placed under state supervision and instructed to abide by a specific set of parole terms. If she is found to be in violation of those terms, she is returned to prison on her original sentence. The central inquiry of the parole-release decision thus may not be whether punishment should cease, but rather whether punishment should occur in the community rather than in the prison. Research on this question is critical, but I save it for future work. For the purpose of this paper, I assume that a full normative account of parole will need to address the question of when and whether punishment should end, and I consider that question from the perspective of four received views of punishment. Retribution. The core commitment of a retributive theory is that punishment is morally justified on the basis of the principle that people who commit crimes deserve to suffer in proportion to the gravity of the crime. Given that the amount of punishment is keyed to the gravity of the crime—which is known at the time of conviction—retributivists have reason to favor determination of sentence lengths at the time of conviction. For various reasons, retributivists tend to favor sentences that cover a general range of time (for example, 7–10 years) rather than a specific time period. Within the range of time, a retributivist may favor basing parole-release decisions on a variety of factors that are morally relevant, but not grounded in retributivism; for example, by an assessment of dangerousness, by employment skills, or by the needs of dependent family members who need the prisoner to take care of them. If pressed to identify release criteria that are grounded in the core commitments of retributive theory, a theorist would have (at least) two options. First, a retributivist could base release decisions on whether a person has actually experienced an amount of suffering that is proportional to the gravity of the crime. The amount of actual suffering that individuals experience from the same length of incarceration varies considerably due to differences in conditions of confinement as well as people’s individual characteristics. A person who has actually experienced more suffering compared to others convicted of a similarly grave crime would have grounds for release earlier within the sentencing range. A second parole-release criterion based in retributive theory concerns transformational change in the person who committed the crime. Some retributivists maintain that the fittingness of a punishment to a given crime

166  Kristen Bell loses its grip if the person who committed the crime undergoes a deep change in character. The person who committed the crime deserves punishment because she stands behind the crime, or is committed to the moral view that the crime expressed. If a person changes over time and continually separates herself from the crime as a reprehensible action which she deeply regrets, she no longer stands behind the crime and it may no longer makes sense for her to be punished for it (Hampton 1988: 154). A person who demonstrates this change would have grounds for release earlier within the sentencing range. Communicative. The core insights of a communicative theory of punishment are that (i) political communities ought to communicate condemnation of crime in order to express the wrongness of criminal actions, (ii) punishment communicates such condemnation, and (iii) insofar as criminal punishment is morally justified, its justification relies to some degree on the aim of punishment to express condemnation.3 A communicative theory, like a retributive theory, may rely on criteria for parole-release decisions that are exogenous to the communicative aim. If pressed to develop criteria within the theory, a natural criterion for a communicative theorist would be whether a perpetrator has understood and recognized the community’s message about the wrongness of her past criminal conduct. This would include understanding both that the violation is wrong and why it is wrong, and embracing a commitment to abide by the law in the future not (only) as a prudential matter, but as a matter of what is owed to the community (Duff 2011: 373). Parole boards might seek to answer the central inquiry by prompting a person to explain how the crime happened, why she did it, and inviting her to express her current perspective on the action. The parole board might also look to whether the person has done her best to understand the attitudes or values that led them to her crime, and to explain what she has done to change those attitudes. In grave crimes, the parole board may consider whether a person has taken action to make amends (for example, paying restitution, fundraising for charities, repairing relationships, etc.) as a way of showing that she fully understands the depth of her wrongdoing. Deterrence. The core claim of a general deterrence-based theory of punishment is that punishment is justified when (and if) punishing people for crime is an effective method of reducing or preventing crime in the community. Deterrence theorists take punishment to be justified only up to the point where resources used on punishment return greater crime-reduction outcomes compared to alternatives. A deterrence theorist would base release decisions on a utility function that is sensitive to factors including crime rates, economic incentives for committing crimes, public perception of the harshness of punishment, cost of incarceration, and predictions of how a person will behave upon release. A person would be released from prison when the utility function indicates that the cost of incarcerating a person would have smaller public safety benefits than releasing the person.

Parole-Release Decisions  167 Some deterrence theorists are less interested in “general deterrence” (which purports to justify punishing a person in order to deter others from crime), and instead focus on “special deterrence” or “incapacitation” (which purports to justify punishing a person in order to prevent that person from committing crime in the community). On a theory of special deterrence, the parole board would not consider the full utility function, but limit itself to consideration of the probability that a person would commit a future crime if released. A person would be released if the probability falls below a given benchmark. The theory itself does not settle what the benchmark should be. Of note, some have proposed grounding parole-release decisions in a mixed view of retribution and special deterrence (Rhine et al. 2016: 14–17). Sentences would be structured such that there is a range of time that “fits” the crime, and release decisions within that range would be made on the basis of judgments about whether the individual is likely to commit another crime if released. Rehabilitation. A rehabilitation theory of punishment holds that punishment is justified only when (and if) punishing people for crime is an effective method of improving people’s ability to function as pro-social members of the community. Punishment in prison is thought to help people improve themselves, but after a certain point, improvement is better achieved outside of prison walls. On this view, a parole board ought to release an individual at “just the right time” to help a person succeed in re-entering society. A person should not be released too early (i.e., when she would still benefit more from time in prison), nor too late (i.e., when she would be stagnated by spending more time in prison). To determine “just the right time” for release, parole boards look to whether people have taken full advantage of rehabilitation programs available in the prison, to their re-entry options, and to the opinions of prison staff and psychologists. Of note, the rehabilitation framework was taken to be the theoretical underpinning of operative parole-release systems in the 1960s and 1970s, although it was unclear whether actual practice reflected this ideal (Rhine 2012).

Two General Objections Each of the criteria described above is subject to (at least) two general types of objection. The first type of objection concerns the limits on our ability to assess fairly or measure the various proposed criteria. The second concerns failure to respect the moral agency of people in prison. Limits on fair assessment. Given the nature of the criteria proposed by each of the four theories, we have limited ability to know whether or not the criteria are fulfilled in any given case. The limitations on our knowledge are somewhat different for each theory. With respect to both the retributive and communicative theories, we are limited because we are not equipped to peer into a person’s psyche and discern whether she has undergone personal transformation and/or understood the depth of her wrongdoing. Outward

168  Kristen Bell indicators like expressions of remorse, apology, and efforts toward amends can provide some evidence of transformation and/or understanding, but we have no mirror into a person’s authentic conscience (Hampton 1988). With respect to the deterrence and rehabilitation theories, we are limited because we lack a “crystal ball” that can predict a person’s future behavior upon release from prison. Actuarial risk assessment tools and expert opinions of psychologists may provide estimated probabilities of whether a person will commit a crime or improve herself if released from prison, but such estimates are not proof of future conduct. Early empirical studies of parole hearing decisions and recidivism underscore the point (Scott 1974; Garber and Maslach 1977; Carroll 1978; Holland 1978; Carroll et al. 1982). Consider one study that compiled case summaries of 200 former parolees, half of whom violated parole and half of whom successfully completed parole and were discharged. Two groups of people—(i) experienced parole officers and (ii) accountants—were asked to read the case summaries and decide who would fail on parole and who would succeed. The parole officers did slightly worse than the accountants, and neither group made better predictions than would have resulted from chance alone (Garber and Maslach 1977: 264). Research in actuarial risk assessment tools have progressed since the time of that study, and parole boards are relying increasingly on these tools. (Ruhland et al. 2016). There is reason to believe that the accuracy of parole boards in making predictions about future conduct has improved and will continue to improve. Nevertheless it is unlikely that accuracy will rise to the level of proof beyond a reasonable doubt. The problem with the inherent uncertainty in the criteria proposed by the four theories is not simply a lack of accuracy, but the fact that the uncertainty invites decision makers to exercise idiosyncratic discretion that in turn engenders arbitrary decision making. Worse, given the nature of some of the criteria, the exercise of discretion will tend to disadvantage some prisoners relative to others on the basis of morally irrelevant factors. With respect to the retributive and communicative criteria, people will fare better at convincing decision makers of transformation and understanding if they are intelligent, understand the culture of the decision makers, and are skilled at comporting themselves and articulating themselves within that culture. People from linguistic and cultural minorities, people who have difficulties in expressing themselves, and people with deficits in understanding or in communicating complex ideas will tend to fare worse. With respect to deterrence and rehabilitation criteria that look to predictions of future conduct, people of color and people from poor, disadvantaged communities will tend to fare worse. Actuarial risk assessment tools tend to overestimate the level of risk they pose because they have fewer past indicators of stability (Larson et al. 2016). Robust procedural protections might mitigate some of the unfairness here. But we should think twice about designing a system which relies on criteria that invite arbitrary decisions and unfairness from the outset.

Parole-Release Decisions  169 Failure to Respect Moral Agency. Even if we could reliably and fairly measure each of the proposed criteria, there are deontic moral reasons to object to each of them. In different ways, each of the criteria fail to respect people in prison as moral agents—that is, as agents with the capacity to accept moral norms and values through their own judgment, and with the capacity to conform their actions to those norms and values. Consider first the deterrence view which involves conditioning release from prison on a prediction about whether a person will commit a crime in the future. Keeping a person in prison because a state official has deemed it necessary to prevent her from responding to some imagined future stimuli treats her as an animal, not an autonomous agent equipped with the capacity to understand moral reasons and make choices. To be clear, the problem is not that acting on predictions about a person’s future conduct necessarily fails to respect that person’s agency. The problem is that the state is determining the harshness of criminal punishment—the amount of time a person spends locked up in a place overtly designed to set back the interests of culpable individuals—on the basis of a prediction about a person’s response to some imagined future stimuli. This treats a person as a dangerous, caged animal; not a person who is a co-member of the political community. In a different way, the other three proposed criteria also fail to respect the moral agency of people in prison. Personal transformation, understanding, and rehabilitation all include a person accepting norms and values for pro-social conduct in society. When the state conditions release from prison on acceptance of prevailing norms and values, it is not respecting a person’s status as an agent with the capacity to accept those norms on the basis of her own judgment. Instead, the state is using the threat of prolonged incarceration to coerce a person into accepting those norms and values by force. Such use of force raises moral concern in two ways: first, it is generally wrong to use force on a moral agent unless or until she changes her mental state. Doing so fails to respect her moral agency; in the words of Martin Luther King, this type of action is “immoral because it seeks to humiliate the opponent rather than win his understanding.” Second, it is generally wrong for the state to so coerce a person because of limits on legitimate state authority. We recognize that such state coercion over the content of mental states is wrong as applied to free people, and the same principle ought to apply to people in prison (Tadros 2011: 355). Given this concern, and the importance of respect for agency that both the retributive and communicative embrace, pure retributive and communicative theorists may ultimately oppose the proposed criteria I sketched above. Most notably, a leading defender of communicative theory, Antony Duff, has made it clear that punishment “must seek to persuade (but not to coerce or manipulate) [a person convicted of a crime] to repent his crime and to accept his punishment as a penance for that crime, while leaving him free to remain unpersuaded and unrepentant” (Duff 2001: 177, emphasis added.) Conditioning release from prison on adopting an attitude of understanding

170  Kristen Bell would violate this principle. It may be that communicative and retributive theorists would ultimately favor the reparative approach that I describe at the end of this chapter, but the issue merits further thought into the weeds of these theories that is beyond the scope of this paper.

The Practice of Making Parole Decisions Having considered parole-release criteria from the perspective of four different theories of punishment, I now consider how one jurisdiction has implemented some of these criteria in the context of term-to-life sentences. I consider California because it has the third largest prison population in the country after the federal government and the state of Texas (Carson 2015: 3), and the largest population of people serving term-to-life sentences (Ghandnoosh 2017). As of 2012, approximately one in three people in California prisons was serving a term-to-life sentence, meaning that they will die in prison unless the parole board finds them suitable for release on parole (Nellis 2013: 6). Parole is far from a guarantee for these prisoners; from 2000 to 2011, more prisoners serving life sentences for murder died in prison than were released on parole (Mullane 2012: 147). Although California has some of the most robust procedural protections at parole hearings compared to other states (Mehta 2016), its parole-release grant rate nevertheless varies remarkably from one political regime to the next (Ghandnoosh 2017: 16). Social scientists and legal commentators have written a great deal about California’s parole process (for example, Weisberg 2011; Young 2016; Caldwell 2016); my discussion here is limited to a brief summary of the legal framework and a critical evaluation of the underlying normative framework. California law requires the parole board to grant parole unless it finds that a person poses an unreasonable risk to public safety. The law’s sole focus on risk to public safety makes it appear that the California parole system is designed to operate under the “special deterrence” framework described above. After a person has served the minimum amount of time proportionate to her offense, she is to be released unless the parole board predicts that she will commit another crime or violate a term of parole if released (In re Reed [2009] 171 Cal.App.4th 1071, 1081). If the board denies parole, it must exhaustively list its reasons for doing so and explain why there is evidence of current dangerousness. That evidence cannot rely solely on facts about the crime or a person’s history prior to the crime (In re Lawrence [2008] 44 Cal.4th 1181). The facts of the crime and other past history can be included as evidence of current dangerousness, however, if there is a “rational nexus” between the crime and the person’s recent conduct or mental state (In re Shaputis [2011] 53 Cal.4th 192, 218). With this cursory description of the framework in place, I turn to describe three morally problematic features of the California parole process. First, as discussed above, grounding decisions on predictions about future conduct

Parole-Release Decisions  171 fails to respect people as moral agents and invites a great deal of arbitrariness into decisions. Of two people with similar case factors, one may be deemed dangerous and denied, whereas the other is deemed not dangerous and granted. Although the parole board states its reasons for every decision, the factors are so murky that the real reason behind a given denial are often left a mystery. What one prisoner said about the mysteriousness of the parole process in another state is equally true in some cases in California: “I feel like I am chasing a ghost” (Mehta 2016). Second, California’s focus on special deterrence at parole hearings is inconsistent with its concern for proportionate sentencing. California law recognizes a proportionality principle in that prisoners are required to serve a minimum number of years that is proportional to the crime before becoming eligible for a parole hearing. But because current law does not recognize a ceiling on the maximum punishment (other than natural death), the system produces outcomes that are perverse from the perspective of proportionality. Consider two 18-year-old co-defendants who had equal culpability in committing a crime, and who were both given a sentence of 7-years-to-life. Suppose one co-defendant was a smart, articulate prisoner who followed rules and was released on parole at age 25 after having served 7 years in prison. Suppose the second co-defendant is continually denied parole because she is frequently written up for non-serious violations of prison rules which the parole board believes to be correlated with a risk of recidivism. Suppose the second co-defendant dies in prison at age 68 after having served 50 years. She has served a sentence that is not only more than 7 times longer—but an entire lifetime longer—for the very same crime. Given that the state has adopted a retributive principle as the reason to set a floor on the minimum punishment, it is inconsistent with that principle to permit such an extraordinarily wide range of punishment without a ceiling. Third, in practice, there is a gulf between the criterion that the law requires the parole board to use—whether or not a person will re-offend if released (special deterrence)—and the criteria that the parole board actually considers in exercising its discretion. As the following discussion shows, the board is strongly influenced by concerns apart from deterrence that are rooted in the retributive, rehabilitation, and communicative aspects of punishment. My discussion here is informed by several studies of California parole hearings as well as reading transcripts from California parole hearings for people serving life sentences for juvenile convictions (Weisberg 2011; Young 2016; Caldwell 2016). California lifer parole hearings are laden with expressions of condemnation for the crime, and with questions about whether a person recognizes the depth of her wrongdoing. Questions about the crime and past criminal history often consume much of the hearing; in some cases, questions about a decades-old crime have lasted over three hours. Hearings close with a statement by the district attorney and the victim or victim’s next of kin (if they choose to make a statement) which are generally prolonged expressions

172  Kristen Bell of condemnation. In addition, the parole board seeks to determine whether the prisoner has “gone deep” in taking personal responsibility understanding her role in crime. If a person attributes the commission of her crime to “external factors” such as drug use and peer pressure within a street gang, the board is likely to think she is “minimizing” her own criminal mindset at the time. The board told one man that they wanted to hear him say he was “a monster” for having sold crack on a street corner at age 16 (transcript on file with author). The board regularly cites the heinousness of the crime, inadequacy of remorse, and insufficient change in attitude since the time of the crime as reasons to deny parole. The discourse is far closer to what you would expect to hear from a retributivist or a communicative theorist (or a hard-nosed priest and a penitent in a confessional) than it is to a collection of data points on predicting future conduct. The parole board also spends roughly 10–20% of a lifer parole hearing asking questions about rehabilitative programs that a prisoner has participated in. When the parole board decides to deny parole, it often cites lack of participation in rehabilitation programs as among the reasons for denial—even if the person did not have access to those programs in prison (Caldwell 2016: 289). In one case, the parole board denied a man who served decades in prison for a juvenile conviction who scored “low risk” on a psychological risk assessment and had a near perfect disciplinary record in prison (transcript on file with author). The board remarked that the man had not participated in any programs in prison, and did not appreciate his attitude that he wanted to simply serve out the time that he owed for his crime. More generally, the parole board looks fondly upon a prisoner if she has written book reports (Caldwell 2016: 289). Logistic regression analysis of over 700 California lifer parole hearings has shown that failure to answer a question about the 12 Steps of Alcoholics/Narcotics Anonymous (a substance abuse program offered in California prisons) significantly impacts the likelihood of being granted parole (Young 2016: 275). It is unclear whether writing book reports or being able to recite the 12 Steps reduces the chance of committing crime upon release. But it is clear that these pieces of evidence show engagement with rehabilitation and with what the parole board thinks is “good for inmates” to do in prison. Finally, the purported focus on special deterrence does not square with the fact that parole is granted in murder cases at a far lower rate than other crimes. In California, people convicted of murder have the lowest recidivism rates compared to any other crime; according to the California Department of Corrections and Rehabilitation’s most recent outcome report, only 2.6% of people convicted of murder who were released returned to prison, whereas 44.6% of all people released returned to prison (California Department of Corrections 2016). The same trend holds true across the nation (Mehta 2016: 67). If the parole board’s concern were solely on preventing future crimes, one would expect people convicted of murder to have relatively high rates of release on parole. But the opposite is true. Nationally,

Parole-Release Decisions  173 parole rates for people convicted of less serious offenses (especially nonviolent offenses) are far higher than for those convicted of murder (Mehta 2016: 46–7). For these reasons, the California parole board does not in practice appear to satisfy its purported goal of making release decisions solely on the basis of predictions about future crime. One lesson advocates could draw from this practice is that more strict procedures are needed in order for the parole board to live up to its purported goal. For example, strict rules of evidence could be implemented to specify what counts as relevant, reliable evidence for measuring the risk of dangerousness. The rules could preclude consideration of factors (like a decades-old crime) whose relevance tends to be outweighed by prejudicial value. The state could train decision makers to focus only on the specified types of reliable evidence for recidivism, and could engage the courts in robust judicial review of the decisions to enforce those rules. Although such procedures are necessary and would improve the parole system, it is unclear whether they would actually succeed in getting parole board members to make decisions solely on the basis of current dangerousness. Practice suggests that retributive, communicative, and rehabilitation concerns sincerely matter to people as they make these decisions. It may be psychologically unrealistic to expect people to weed out values that matter to them in the context of such high-stakes decisions. A better way forward may be to design parole-release decision making in a way that recognizes a variety of considerations in a structured fashion, rather than asking decision makers to contort their vision into seeing through a single lens. Moreover, even if improved procedures could make parole decisions turn solely on the criterion of dangerousness, the system would remain unjust in light of the two objections articulated above: the looseness of “dangerousness” invites arbitrary decision making, and continued incarceration based on predictions of future conduct fails to respect the moral agency of people in prison. Furthermore, an additional injustice also arises given that the system is operating upon non-ideal social conditions. Many, if not most, people serving life sentences in California have endured deep injustice and trauma in their lives that affected their commission of the crime. In a sample of 107 people serving life sentences for juvenile convictions, 62% experienced sexual, physical, or emotional abuse prior to the crime (Caldwell 2016: 276). In addition to having experienced direct abuse, people discuss having been homeless, having witnessed their mothers or siblings being raped or beaten, having been shot at or having seen friends shot at, having developed an addiction to drugs and alcohol at a very young age (for example, under 10 years old), having been brought into a street gang as a teen by older family members, and not having received any help in school for learning disabilities or trauma. As Duff and others have argued, such circumstances of disadvantage and exclusion from the community ought to have a mitigating effect on

174  Kristen Bell punishment. Among other things, such circumstances undercut the state’s moral standing to make the perpetrator of crime answer for that crime (Duff 2001: 186–200). In the context of a parole system ostensibly grounded in special deterrence, however, these circumstances are used as reasons in favor of denying parole. Like many actuarial risk assessment tools, the California parole board explicitly deems an “unstable social history” to be predictive of dangerousness (California Code of Regulations., Title 15, § 2402). The very features that deprived a person of an equal chance in life before the crime are used to disadvantage them further in parole decisions. Further, the state stands in a position of hypocrisy when it demands that prisoners accept full personal responsibility for the crime and express understanding of their role in the crime, while the state itself never acknowledges or addresses the role that it played in engendering conditions of injustice that underlay the crime.

The Reparative Approach Having identified some of the challenges in designing parole-release criteria from a theoretical and practical perspective, I turn now to suggest a fundamentally different approach to parole-release decisions. The approach described below aims to decrease arbitrary decision making by setting clear, individualized criteria for each person at the outset of the prison sentence. In setting specific, individualized criteria at the outset, the approach also aims to respect the agency of people in prison by providing a roadmap on which they can exercise their own judgment and realistically earn their way out of prison. Further, it takes the task of crafting substantive release criteria out of the hands of the state (which lacks the moral standing to do this task in many cases), and places this task largely in the hands of a circle of people directly impacted by the crime. I sketch the approach below, leaving further development of it for future research. I call the approach a reparative approach because it is based on a restorative justice paradigm, but only loosely so. A restorative justice paradigm generally rejects state-centered, punitive responses to crime in favor of community-centered responses that aim to repair the harm of the crime and rebuild broken relationships. The reparative approach incorporates elements of restorative process, but it is not strictly under the banner of restorative justice because it assumes the existence of a system in which the state is still in the business of punishing people in prison for a period of time before they become eligible for release on parole.4 I have adopted this assumption not on the basis of ideal theory, but because I do not think it is reasonable to expect that the American public will take imprisonment off the table in the foreseeable future, at least not for grave crimes like murder and rape. The reparative approach would apply in cases where a person is serving a prison sentence that includes the possibility of release. The approach would begin by convening a restorative justice circle at the outset of that sentence.

Parole-Release Decisions  175 As Johanna Luttrell describes in her article in this volume, such a circle brings together various members of the community who are impacted by a problem (or crime), who talk and listen to each other about what happened, who express the harm and discuss what can be done to repair the harm, and create an individualized and specific reparation plan that can include harsh punishment among other things. In this context, the circle would include the person who committed the crime, her friends or family, the victim and/or friends or family of the victim (the victim and people close to the victim may choose not to participate, and if so, a surrogate could participate instead), the friends or family of the victim, and representatives from the neighborhood where the perpetrator lived, where the victim lived, and where the crime happened. The circle would also include people who worked for the state and who had an influence on either the perpetrator or the victim—for example, a social worker, a teacher, a parole or probation officer, and staff at the prison where the person will be incarcerated. The circle would discuss the full depth of the harm of the crime, as well as any unjust disadvantages that the perpetrator or the victim faced prior to the crime. After discussing the full extent of the harm, the circle would articulate two sets of goals for repairing the harm: (1) criteria that the perpetrator must meet in order to repair the harm, and (2) criteria that the community ought to work toward in order to repair the harm of the crime and any underlying conditions of injustice. For example, the community could be asked to provide the victim with individual therapy, and with an opportunity to participate in a carefully mediated victim-offender dialogue. A perpetrator could be asked to work a job in the prison to pay for those services, and to participate in programs such as substance abuse treatment, trauma-based counseling, and educational and/or vocational training. If the circle discovers that the perpetrator suffered abuse or other forms of disadvantage that affected the crime, the community could also be asked to conduct an investigation of the perpetrator’s background and work toward ameliorating remaining conditions of injustice. For example, if the circle discovers that the perpetrator was sexually abused at a foster home, they could order an investigation into that foster home (or others in the area) to determine whether other children are being abused. Critically, the criteria would be established at the outset of a person’s sentence. Setting forth individualized criteria at the outset expresses that the perpetrator is a responsible agent capable of making choices for herself in the prison and acting on those choices. Once the criteria are clearly established and she knows what they are, it is up to her to fulfill them and so in some sense “earn” her way out of prison. Further it is not “the state” or its representative that would establish the criteria—as Duff has argued, the state lacks standing to hold the perpetrator accountable in cases where marked disadvantage and exclusion underlie the crime (Duff 2001). Instead, the people who are impacted by the crime create the terms of what it is to appropriately hold the perpetrator accountable. They set the standards, and

176  Kristen Bell the state acts as a mediator (of sorts) to monitor compliance with those standards. Although the state may lack moral standing to substantively craft and impose the standards, no such standing is needed to engage in the task of monitoring compliance with those standards. (The solution is imperfect because the state is still engaging in the punishment of incarceration for the years prior to the parole hearing, and in some cases the state may lack the moral standing to impose this punishment.) Ideally, the prison sentence would come to be understood as a structured opportunity for the offender to work toward repairing the harm of the crime and for the community to work toward repairing underlying injustice that surrounded the crime. The period would be complete—and thus release on parole would be warranted—when the offender has served a minimum number of years required by the judge at sentencing and has completed the clear, individualized criteria that are set forth by those impacted by the crime. In order to reduce subjective judgments on the part of the parole board, the criteria would be specific and include a clear metric for measuring fulfillment of the criteria. For example, a criterion such as “the perpetrator should address trauma underlying the crime” would be too vague, but “the perpetrator must complete a minimum of 10 sessions of trauma-based counseling that focus on his history of being sexually abused.” So long as criteria are specific and measurable, the parole-release decision would become a fairly ministerial task of measuring completion of those criteria based on established metrics. The criteria that the restorative circle would establish would need to be reasonably achievable by the perpetrator during the minimum period of punishment. What counts as reasonably achievable will vary from person to person based on individual capacities. For example, earning a college degree and multiple vocational certifications is not reasonably within the reach of a person who has severe cognitive impairments. In addition, the criteria must be achievable with respect to the conditions of imprisonment. A representative of prison administration ought to be part of the circle to inform this discussion, and an independent body would likely be needed to ensure that the prison follows through on its commitments. (This is no small task given the current level of resources devoted to programming in prisons: according to a study of state parole boards, 44 require rehabilitation programs for release, but only two report having enough programs (Ghandnoosh 2017: 32).) After the circle decides on proposed criteria, a judge would need to determine whether the criteria are sufficiently specific, and that they are reasonably achievable from the perspective of both individual and institutional capacities. If the criteria fail on any of these fronts, the circle would be reconvened to propose an alternative plan. In addition to setting initial criteria, procedures may be developed to assess progress at various intervals during the period of incarceration and to allow for a small number of modifications to the criteria based on unanticipated developments.

Parole-Release Decisions  177 A thorough description and defense of the reparative approach is beyond the scope of this paper. It may be that sufficiently specific criteria are too difficult to agree upon in the context of grave crimes. It may also be implausible to enforce commitments that the prison provide programs that are necessary to meet the criteria. These are just some of the issues that require further research. But given the problems with existing parole-release systems and the pressing need to reform these systems, there is no better time to conduct this further research into a fundamentally different approach.

Notes 1 While I have changed the name of this person to protect his identity, all other cited facts about this case are true. 2 I thank Gideon Yaffe for calling my attention to the complexity of this question. 3 Theorists disagree on whether communication plays a central role in the moral justification of punishment. Duff, for example, maintains that the communicative aim of punishment grounds the very legitimacy of a system of crimes and punishments. On the other end of the spectrum, Victor Tadros argues that while the communicative aim of punishment is morally significant in considering how we ought to tailor punishments, it plays a cursory role (at best) in justifying the existence of a system of criminal punishment (Tadros 2011: 99–110). 4 I am drawing here on a distinction between restorative process (victim-offender mediation, sentencing circles, family-group conferences) and restorative justice (paradigm of not seeking to complement the criminal justice system but to replace it) (Robinson 2011: 353–5).

References American Bar Association (1994). Standards for Criminal Justice: Sentencing, 3rd ed. Chicago: American Bar Association Press. Board of Parole Hearings (2015). “Board of Parole Hearings Workload Summary.” Available at www.cdcr.ca.gov/BOPH/docs/LSTS_Workload_CY2015.pdf, accessed December 27, 2016 Braithwaite, John (2002). Restorative Justice and Responsive Regulation. Oxford: Oxford University Press. Caldwell, Beth (2016). “Creating Meaningful Opportunities for Release: Graham, Miller, and California’s Youth Offender Parole Hearings.” N.Y.U. Review of Law and Social Change 40: 245–304. California Department of Corrections and Rehabilitation Office of Research (2016). “2015 Outcome Evaluation Report: An Examination of Offenders Released in Fiscal Year 2010–2011.” August. Available at www.cdcr.ca.gov/Adult_Research_ Branch/Research_Documents/2015_Outcome_Evaluation_Report_8-25-2016. pdf, accessed December 28, 2016 Carroll, John S. (1978). “Causal Attributions in Expert Parole Decisions.” Journal of Personality and Social Psychology 36: 1501–11. Carroll, John S., Weiner, Richard, Coates, Dan, Albrio, James (1982). “Evaluation, Diagnosis, and Prediction in Parole Decision Making.” Law and Society Review 17: 199–228.

178  Kristen Bell Carson, E. Ann (2015). “Prisoners in 2014.” September. Department of Justice, Bureau of Justice Statistics. Available at www.bjs.gov/content/pub/pdf/p14.pdf, accessed December 27, 2016 Conley, John A and Zimmerman, Sherwood (1982). “Decison-Making by a Part-Time Parole Board: An Observational and Empirical Study.” Criminal Justice and Behavior 9 (December): 396–431. Duff, R.A. (2001). Punishment, Communication, and Community. Oxford: Oxford University Press. Duff, R.A. (2011). “Restorative Punishment and Punitive Restoration.” In Why Punish? How Much? ed. M. Tonry. Oxford: Oxford University Press, 367–82. Garber, Robert M and Maslach, Christina (1977). “The Parole Hearing: Decision or Justification.” Law and Human Behavior 1: 261–81. Ghandnoosh, Nazgol (2017). “Delaying a Second Chance: The Declining Prospects for Parole on Life Sentences.” The Sentencing Project. Available at www.sen tencingproject.org/wp-content/uploads/2017/01/Delaying-a-Second-Chance.pdf, accessed March 12, 2017 Hampton, Jean (1988). “The Retributive Idea.” In Forgiveness and Mercy, ed. J.M. Hampton. New York: Cambridge University Press, 111–61. Holland, Terrill, Holt, Norman, and Brewer, David (1978). “Social Roles and Information Utilization in Parole Decision-Making.” Social Psychology 106: 111–20. Justice Policy Institute (2016). “Defining Violence: Reducing Incarceration by Rethinking America’s Approach to Violence.” August. Available at www. justicepolicy.org/uploads/justicepolicy/documents/jpi_definingviolence_final_ report_9.7.2016.pdf, accessed December 28, 2016 Kobrin, Sandra (2005). “Dying on Our Dime.” Los Angeles Times, June 26. Kuznia, Rob (2016). “An Unprecedented Experiment in Mass Forgiveness.” The Washington Post, February 8. Larson, Jeff, Mattu, Surya, Kirchner, Lauren, and Angwin, Julia (2016). “How We Analyzed the COMPAS Recidivism Algorithm.” May 23. Available at www.pro publica.org/article/how-we-analyzed-the-compas-recidivism-algorithm. Mehta, S. (2016). “False Hope: How Parole Systems Fail Youth Serving Extreme Sentences.” American Civil Liberties Union. November. Available at www.aclu. org/feature/false-hope-how-parole-systems-fail-youth-serving-extreme-sentences. Mullane, N. (2012). Life After Murder: Five Men in Search of Redemption. New York: Public Affairs. Nellis, A. (2013). Life Goes On: The Historic Rise in Life Sentences in America. Available at http://sentencingproject.org/wp-content/uploads/2015/12/Life-GoesOn.pdf, accessed December 27, 2016. Rhine, Edward E. (2012). “The Present Status and Future Prospects of Parole Boards and Parole Supervision.” In The Oxford Handbook of Sentencing and Corrections, ed. K.R. Joan Petersilia. Oxford: Oxford University Press, 627–56. Rhine, Edward E., Petersilia, Joan, and Reitz, Kevin (2017). “The Future of Parole Release: A Ten Point Plan.” In Crime and Justice: Reinventing American Criminal Justice, ed. M. Tonry. Chicago: University of Chicago Press. Robert Weisberg, Mukamal, Debbie, and Segall, Jordan (2011). “Life in Limbo: An Examination of Parole Release for Prisoners Serving Life Sentences with the Possibility of Parole in California.” September. Available at http://law.stanford.edu/ wp-content/uploads/sites/default/files/child-page/164096/doc/slspublic/SCJC_ report_Parole_Release_for_Lifers.pdf

Parole-Release Decisions  179 Robinson, Paul (2011). “The Virtues of Restorative Process, The Vices of ‘Restorative Justice.’ ” In Why Punish? How Much, ed. M. Tonry. Oxford: Oxford University Press, 353–67. Ruhland, Ebony, Rhine, Edward, Robey, Jason, and Mitchell, Kelly Lyn (2016). “The Continuing Leverage of Releasing Authorities: Findings from a National Survey.” Available at https://robinainstitute.umn.edu/publications/continuing-leveragereleasing-authorities-findings-national-survey, accessed December 27, 2016 Schwartzapfel, Beth (2015). “How Parole Boards Keep Prisoners in the Dark and Behind Bars.” The Washington Post, July 11. Scott, Joseph (1974). “The Use of Discretion in Determining the Severity of Punishment for Incarcerated Offenders.” Journal of Criminal Law and Criminology 656: 214–24. Tadros, Victor (2011). The Ends of Harm: The Moral Foundations of Criminal Law. Oxford: Oxford University Press. Vera Institute of Justice (1978). Final Report on Parole Decision-Making Project Sumbitted to NYS Department of Correctional Services. Vera Institute of Justice: New York. Winerip, Michael, Schwirtz, Michael, and Gebeloff, Robert (2016). “For Blacks Facing Parole in New York State, Signs of a Broken System.” The New York Times, December 4. Young, Kathryne M., Mukamal, Debbie, and Favre-Bulle, Thomas (2016). “Predicting Parole Grants: An Analysis of Suitability Hearings for California’s Lifer Inmates.” Federal Sentencing Reporter 28: 268–77.

10 Restorative Justice in High Schools A Roadmap to Transforming Prisons Johanna C. Luttrell

The School-to-Prison Pipeline In recent decades, criminal justice trends and policies like “broken windows” policing, mandatory minimum sentencing, and the mass incarceration of people of color, have found their way into schools, in the form of “zero-tolerance” policies which target young African American and Latinx students (Heitzig 2016). Experts name the phenomena, the points of connection between the criminal justice system and education, as the “school-toprison pipeline.” The “school-to-prison pipeline” manifests itself in multiple ways: firstly, the school buildings and administration increasingly resemble prisons, with metal detectors at the doors and hired police in the hallway ready to do search-and-seizure operations (Mallett 2015). Secondly, schools are mimicking the punitive character of the criminal justice system, in popular “zero-tolerance” approaches that criminalize minor infractions, suspending and expelling students for first and second offenses like fighting in the hallway, swearing, or being disrespectful to a teacher. Suspension, temporary removal from school, and expulsions, semi-permanent to permanent removal from school, are extreme mechanisms of marginalization and isolation, yet they are often used as a first line of defense against difficult behavior (Heitzig 2016). Loïc Wacquant contextualizes this marginalization that black and brown children experience in school within the history of the evolution of racialized practices of punishment in the United States (Wacquant 2008). In fact, zero-tolerance school policies were modeled after the “get tough on crime” drug enforcement laws of the 1980s (Skiba and Peterson 1999). Thirdly, schools are feeding their children into the criminal justice system. Increasingly, middle and high schools are referring their students to the police and the courts on issues that could be resolved within the school, and students end up in juvenile detention or prison instead of continuing to learn in the classroom. Also, as with the increase in private prisons, so there has been an increase in private “alternative” schools specializing in discipline (Gonzalez 2012). Like the criminal justice system, schools have severe racial punishment gaps: black and brown children are consistently disciplined more harshly than white

Restorative Justice in High Schools  181 children. The Department of Education’s most recent report on disparities within school punishment is damning. Black preschoolers, while only 18% of the preschool enrollment, comprise 48% of preschool students who receive more than one out-of-school suspension. Surely, there is a better way to punish 4-year-olds than exiling them from the classroom. Across the age spectrum, black students are suspended and expelled at rates three times greater than their white counterparts. On average, 4.6% of white students are suspended, compared to 16.4% of black students. 20% of black boys and 12% of black girls receive an out-of-school suspension (Dept. of Education 2014). Suspensions are also the number one predictor of dropout rates, and there is a strong correlation between dropout rates and imprisonment (Flannery 2015). In 2003, 56% of federal inmates, 67% of inmates in state prisons, and 69% of inmates in local jails did not finish high school (Harlow 2003). The trauma from being labeled a “bad kid,” and marginalized by separation from the community, so early in a child’s educational career, can become a self-fulfilling prophecy as they become adults (Brown 2007). The sense of “not belonging” that suspensions and expulsions induce is replicated within the prison system (Wearmouth and Berryman 2011). There is another connection between education and the criminal justice system, in the similarity of extreme expectations put upon both teachers and the police. In the mess of trying to teach children who come to school with poor nutrition, facing poverty, violence, and sometimes homelessness in their family life, teachers often become the “first line of defense” against a deluge of social problems that they cannot, by their singular effort, solve. In some ways, the increasingly strict disciplinary measures like zero-tolerance come about as coping mechanisms to school over-population and lack of funding (DeLauri and Eaton 2010: 16). Police, too, labor under impossible expectations. Police speak of becoming “the thin blue line” against any number of neighborhood social ills that they themselves cannot solve (Katz and Walker 2012).

Restorative Justice in Schools Understanding these alarming connections and similarities between school and criminal justice systems, faculty and administrators working in restorative justice aim to “disrupt the school-to-prison pipeline.” Restorative Justice (RJ) is an alternative approach to traditional forms of punishment, an approach which emphasizes repairing harm and restoring relationships and wholeness for all parties involved: victims, offenders, and their communities. RJ seeks solutions that will strengthen communities, keep people accountable for their actions, and reintegrate those who committed harm. In general, restorative justice practices can take a variety of forms. In the United States, the term “restorative justice” first became popular as a way for victims to be involved in the punishment of their offender by expressing how the crime impacted their lives, and what would need to happen in order

182  Johanna C. Luttrell for them to feel like justice had been served (Zehr 1990). Now, the most popular forms of RJ are: victim-offender mediations (VOM), where the victim and offender, under the auspices of a mediator, meet to dialogue, and the victim expresses how they were affected by the crime, and if there are any avenues of remediation; community reparative boards, where citizen boards work with the victim and offender to create a punishment that addresses the concerns of the victim and the community in which the harm was enacted; and circles, the most common sort of restorative justice seen in high schools. In additions, schools run other RJ oriented programming like student conferences, peer mediation, restitution, and community service. What all of these forms have in common is that they do not treat a harm that has been done as a harm against an impersonal entity or institution, like the state or the school. These practices view harms as violations against specific people and communities, and thus must be resolved on a “local” level, between people and the communities themselves. Anita Wadhwa, founder of the Houston Restorative Justice Collaborative, and RJ teacher at Spring Branch Public Schools in Houston, Texas, primarily uses circles. As Wadhwa explains, circles have their roots in indigenous courts (Wadhwa 2015). The “circle” is just what it sounds like: everyone sits in a circle, and passes around a talking piece. In many urban public high schools RJ practitioners do three kinds of circles: talking circles, healing circles, and accountability circles. “Talking circles” allow students to build bonds with each other, develop empathy, share stories, and influence behavior (Wadhwa 49). There does not necessarily have to have been a “harm done” in order for the talking circles to occur. “[Teachers] used circles to check in about how students were feeling about school or to talk about topics such as love and loss . . .” (Wadhwa 42). The talking circles occur at regular intervals, in a preventative format of building bonds between community members, and not as a reaction to a specific harm. Oftentimes, teachers in other classes will flag students for the talking circles—these typically include students who are having a hard time, who are losing ground in their studies. The student begins to engage in the circle just to say what is on their heart, and what they are going through. They are met by empathetic peers, whose only job it is, at first, to listen and affirm each other’s feelings. The students report that the circles allow them simply to know more of their peers in school, and to get out of their individual bubbles, cliques, and gangs (Wadhwa 65). Wadhwa also encourages other teachers to participate in the talking circles, and to be just as vulnerable as the students. In participating, students begin to see teachers as human beings rather than mere disciplinarians, and also begin to see that poor behavior in school would hurt individual, real people. That is, they begin to understand the consequences of their actions. After the circles, the teachers report feeling more “grounded” and “connected” to both their colleagues and their students in the process. One teacher described the circles as tools that combat burnout:

Restorative Justice in High Schools  183 “They put me back together at the end [of the week] when we’re doing them on Friday. I wouldn’t have been able to make it this far at all, if we didn’t have a bunch of those . . . It helped me remember what it is we’re trying to do and give chances to connect with everybody in a meaningful way.” (Wadhwa 66) The second kind of circle is a healing circle, which occurs after a harm has been done. Healing circles are places where students reflect on their problems and, in conjunction with their peers and, at times, their victim[s], make a plan to repair the harm. The kind of plan that comes from a healing circle is an interpersonal type, not in an institutionally mandated discipline. Healing circles most closely resemble victim-offender mediations, and what is crucial about these circles is that everyone involved feels heard, understood, and that they belong. Any possible plan for reparation comes from the agents involved, and everyone must feel that their agency is respected. Thirdly, accountability circles often include representatives from the community or institution, including, at times, the children’s parents, who work together to make an institutionally “official” plan for the student to make reparations. These circles most closely resemble courts, but they still follow the principles of restorative justice. If all parties feel it is the most just outcome, it is possible, in certain instances, that the reparation plans which come out of accountability circles are harsh punishments, like suspensions. However, since the process is participatory from the beginning, RJ practitioners feel that the process encourages accountability and inclusion in ways that quick, “top-down” disciplinary measures fail to accomplish. All three of kinds of circles work in conjunction with each other, and RJ practitioners stress their interdependence; it is imperative that the students form the bonds with each other and feel accountable for their actions, not in terms of punishments that are put upon them by their superiors, but in terms of their own emotional growth. Importantly, though, as Wadhwa writes, “the purpose of the [any of the] circle[s] is not to decide whether someone is a good or bad person”; in other words, whether or not a person belongs in the community in the first place (Wadhwa 51). The purpose of RJ is that the students never feel like they deserve to be expelled from the community as a whole. Each person in the process must feel the critical and paramount sense of belonging, and feeling thus, they cultivate feelings of accountability to the community. Importantly, Wadhwa and other RJ practitioners in urban schools also use their class as a sort of consciousness-raising space, for both the students and the teaching staff. Students are encouraged to take a critical perspective on their own situation, many as people of color subject to the exigencies of the criminal justice system. Different RJ classes read, for example, Michelle Alexander’s The New Jim Crow together, and reflect on the ways in which the topics bear on their own lives (Alexander 2010). As students are asked

184  Johanna C. Luttrell to, for example, “talk about a time when they were treated differently for their race,” teachers themselves can be faced with uncomfortable realities. As David Simson writes, “[In RJ circles] students have a greater ability to share their belief that a teacher’s behavior toward them was inappropriate and their belief that their race played a role in this process” (Simson 2014: 558). Further, “[RJ’s] focus on rehabilitation and reintegration might lead to a more open dialogue about stigma’s extent, lived reality, and hurtful consequences with which minorities have to contend even today and which is likely to influence both behavior and disciplinary decision making at school” (Simson 559). Thus, the critical-consciousness development aspect of an RJ curriculum has the potential, at least, to transform institutions, as well as individuals. The impact of the circles is broader than the circles themselves. Students are to consider themselves as agents of this change. Wadhwa trains her own students as RJ practitioners, and, in her words, “flips the hierarchy” between teacher-as-expert and student-as-observer (Wadhwa 116). The goal is that they “shift the paradigm” of the teacher/ student relationship from one of powerful/powerless to “adults and youth in equal partnership” (Wadhwa 115). The high school students themselves become experts in mediating circles, and replicate the circles in middle and elementary school classes, and in their own personal contexts, without the teacher present as the sole authority. Training and treating the students as authorities in RJ work allows for the first steps in the students seeing themselves as experts in their own experience, and is an important part of the goal of education for a critical consciousness in general. Wadhwa’s students do not only describe the effect of their education in personal terms. Instead, they started a (large!) collaborative across the city of Houston, one whose motto is: “Let’s start a movement!” (RJ Collaborative of Houston 2016). Restorative Justice in urban high schools works. It deters, prevents, and provides just deserts. The International Institute for Restorative Practices found that RJ correlated to a decrease in disciplinary referrals, detentions, suspensions, and overall disruptive behavior (Lewis 2009). One study in Oakland, California, found that “schools implementing restorative justice have had a 24% reduction in chronic absenteeism, a 128% increase in ninth-grade reading levels, a 56% decline in the high-school dropout rate, and a 60% increase in the four-year high school graduation rate. So more students are in school, learning and graduating on time” (Wall Street Journal 2015). Beyond prevention and deterrence, though, RJ “helps educators and students uncover the root of conflicts, create space for individuals to be accountable for their action, teach social and emotional literacy, foster a sense of community and improve academic outcomes, relationships, and school climate overall” (Kline 2016: 99). That is, restorative justice is transformative, not only for achieving the institutional goals of disciplinary outcomes, but for the students’ own lives. In fact, many RJ practitioners

Restorative Justice in High Schools  185 find the term “Restorative Justice” to be a misnomer, and find Transformative Justice to be a better description, because RJ does not only “restore” non-existent communities and relationships, it creates new, working ones. However, a recent “This American Life” radio program on Lyons Community School in Brooklyn, NY, illuminates the fact that, no matter the success rate of these RJ programs in high schools, the students still face a criminal justice system which is not restorative or transformative; it is punitive, unjust, impersonal, and racially biased (“This American Life” 2014). Lyons is a completely RJ-oriented school; the entire curriculum relies on RJ practices for its guiding pedagogy, in both RJ-specific classes and in regular academic courses like math and science. The students in the program express that, until they came to Lyons, they did not feel that any of their superiors cared about them. At Lyons, they felt a sense of belonging, many were deterred from juvenile detention, and began to become interested in their studies once again. They formed community bonds with their peers and the teachers. The teachers generally felt secure that their work was accomplishing its goals until one day when, while on a field trip to an art museum in Manhattan, a class group was traveling on the subway together. One of the students got in a minor altercation with a very aggressive man, whom the student took to be mentally ill. The man turned out to be a plainclothes police officer, who arrested four of the black, male students, arrests which later resulted in felony charges. The teachers were horrified. They thought they could talk to the officer, explain that they were on a field trip, they were responsible for the students, and that no arrests were needed. They found out that was not possible. The officer quickly escalated the situation, and was unresponsive to the teacher’s entreaties. The radio narrator describes this moment as when “restorative justice meets plain old criminal justice.” The felony charges inspired major soul searching on the part of the teachers and administration of Lyons School—they wondered if they, as believers in the RJ curriculum, were preparing their minority students for a world that did not exist. Or (and this is a larger worry about RJ approaches in general) that they were creating compliant subjects re-integrated into a community that is, still, unchanged and unjust. Currently, many schools closely resemble prisons. However, it’s worth considering whether prisons should, instead, resemble schools. These urban schools are on the “front lines,” so to speak, of movements of mass incarceration in the US, and they have developed serious and effective ways to combat the momentum of such trends, to advocate for children they care about. These methods are worth importing into adult contexts, if not for any other reasons than they are effective in prevention, deterrence, and justice. These RJ methods also originate in environments of care, concern, and inclusion. In order to unravel the school-to-prison pipeline, these educators realize that the work cannot begin and end with education. The criminal justice system must be transformed as well.

186  Johanna C. Luttrell

Restorative Justice in Prisons The US criminal justice system is not, nor was ever meant to be, a completely rational cost-benefit analysis of just deserts, deterrence, or crime prevention. Even an abbreviated history of prisons in the US shows the degree to which they were a legacy of antebellum slavery, and, afterwards, forced labor during the Reconstruction.1 In the United States, prisons were constructed, in great measure, for the purposes of the oppression and marginalization of people of color. Around the world, restorative justice practices are gaining popularity around the world, in many different forms: not only victim-offender mediation (VOM), but alternative youth courts, community dialogues, and truth and reconciliation committees. RJ is gaining in popularity in the United States, as well. However, in the US, RJ practices may be used and valued within prisons, but they have little to no influence on sentencing, parole, or probation. For instance, the seminal guide to victim-offender mediation explicitly stresses that VOM is “a process that has absolutely no direct influence on the classification, custody, parole, probation, or release date status of the offender” (NAVSPIC VOD National Standards 2016). It specifies that the VOM process is “personal” and not “legal,” and that the mediators’ case notes are absolutely confidential, and should not be shared with any institutional entity, including parole board or probation files (NAVSPIC VOD National Standards 2016). In an important way, this restriction makes sense: in order for the VOM process to be genuine and authentic, the victim needs to be free to express the entirety of their experience and feelings about the crime, and there cannot be serious incentives that would induce the offender to respond in predictable, desirable ways. For example, if the VOM influenced sentencing, an offender might be tempted to act more repentant than they actually feel. If there is a serious outcome at stake for the offender, they might be inclined to be disingenuous about their reaction to the victim. Also, one must consider the impossibility of “measuring” feelings of regret or accountability—this would be an impossible task for any mediator, and would distract from the important process of victim-offender dialogue in the first place. Most prison-oriented RJ programs have the same confidentiality and outcome-oriented restrictions. For example, Insight Development Group (IDG), an influential RJ program at the Oregon State Penitentiary, offers typical programming for offenders: philosophy, skill-building, and psychology courses, grief group counseling, and life coaching (IDG 2016). The program, like many others, is all volunteer-run, and operates independently from prison staffing. Many inmates involved in these courses find them to be transformational on a personal level. However, there is no accompanying material transformation while in prison. However, RJ programs have a serious impact on recidivism rates once inmates are released, even in the cases of very minimal programming (Bergseth

Restorative Justice in High Schools  187 and Bouffard 2013). Researchers at Sam Houston State found that “restorative justice programs, such as victim-offender mediation and community impact panels, are more effective in reducing recidivism rates among juvenile offenders than traditional court processing” (Bergseth and Bouffard 2013: 1071). Full programs in direct mediation in prison have a recidivism rate of 33%, in comparison with a recidivism rate of 50% in regular court. RJ programs in US prisons should be more influential in decisions regarding sentencing lengths. One reason RJ in high schools is so effective is precisely because it has real consequences, effects on reducing suspensions and expulsions. Educational practitioners do not separate their so-called “political” goal, ending the school-to-prison pipeline, from their educational goal, disciplining and transforming students. Similarly, any goal of “transforming prisons” must include transforming prisoners. Both the victims and offenders in educational RJ contexts have a say in the disciplinary outcome. Also, educational RJ aims to address systemic racism in a way that incorporates the experiences and voices of those who are most victimized by it. To be quite honest, I am not convinced that justice in prisons could look any other way. To illustrate the need for more consequential RJ programs in prison, I recall an account from Bryan Stevenson’s book, Just Mercy. One of the cases that most affected him in his work for the Equal Justice Initiative, an influential NGO that advocates for equal treatment in criminal justice and challenges poverty and racial injustice (Stevenson 2014), is Stevenson’s work with Ian Manuel. Manuel is a black inmate in a Florida penitentiary who spent eighteen years in uninterrupted solitary confinement. In 1990, when Manuel was 13, he and some friends tried to rob a couple in Tampa, and he ended up shooting his victim, Debbie Baigre (a white woman), in the face. Her injury was severe, damaging, and traumatic. At this time in Florida, juveniles could be sentenced to life in prison without parole, and this was Manuel’s sentence. Because of prior offenses, he was sent to an adult prison. Because he was young and small, he was vulnerable to sexual assault, and the prison staff put him in solitary confinement. His mental health became severely compromised—he started self-harming, which only caused his isolation to be extended. Two years after the crime, Manuel unexpectedly called Baigre, on Christmas Eve, expressing deep and uninhibited emotional regret for the incident. Baigre was moved, and over time, she and Manuel formed a relationship. After communicating for several years, Baigre began to advocate to prison officials and the press on Manuel’s behalf. Baigre wrote, “No one knows more than I do how destructive and reckless Ian’s crime was. But what we’re currently doing to him is mean and irresponsible. When this crime was committed, he was a child, a thirteen-year-old boy with a lot of problems, no supervision, and no help available. We are not children” (Stevenson 153). The relationship that Baigre and Manuel formed was an important one, for both of them. As Stevenson recounts it, Manuel did not have much family support. Baigre became a crucial encouragement for Manuel, and over time, she became friends with his mother and brother. Now, Manuel’s

188  Johanna C. Luttrell mother, father, and brother have died, and Baigre is the only “family” Manuel considers he has left (Kristof 2014). Because of Manuel’s case, Stevenson and the Equal Justice Initiative were able to change the law in Florida to prevent juveniles from getting life sentences. Still, Manuel remains in prison, and Baigre’s testimony about Manuel’s character has little impact on his continued sentence. Intuitively, one feels that Baigre and other victims should be allowed to have some impact on their offenders’ sentences, and this intuitive insight is really the starting point of RJ approaches. I cannot give a full defense here of why the victim should have a say in the outcome of a punishment,2 but I do want to stress, again, that the effectiveness of the RJ programs in high schools rests on its potential impact on actual disciplinary outcomes, through the interdependency of the three types of circles. That is, the talking circle is not independent of the healing or accountability circle, and, by and large, it is the same crew of characters who are involved in the circles from the beginning. As they stand now, RJ programs in prisons have been cut off at the knees. They achieve what traditional sentencing measures are, at least in public discourse, aiming to achieve: accountability. Yet, any benchmarks of accountability offenders come to in RJ programs remain independent of sentencing, parole, and probation measures. During parole board hearings, especially, questions of what constitutes authentic remorse and intention, what constitutes a sincere apology and what does not, are in play. If insights from RJ practitioners were included, this could expand the scope of community knowledge, and parole boards could reach their decisions with more relevant information. Furthermore, the critical consciousness RJ students develop in high schools is instructive for prisoners as well. This can be heard as a more radical proposition: that, like the training of high school students to become experts in RJ mediation, those most affected by the institution are those most poised to change it. Thus, prisoners themselves are the agents of change in a criminal justice system. This insight is in line with research in feminist standpoint theory—that, in situations of oppression, those experiencing the oppression have, like Marx’s proletariat, a critical insight on the workings of the system as a whole.3 Prisoners, like high school students, must feel themselves as agents, and as authorities in their own experience. Of course, it is controversial to say that prisoners themselves need to be empowered in a critical consciousness. However, as we learn with the experience of RJ in high schools: one does not become the kind of person who can feel themselves fully responsible and accountable for offenses they commit, if one does not feel oneself as an agent in the first place, as the kind of person who has the power to harm at all. The students did not realize they were actually harming the teachers until they saw themselves as equals. RJ in high school shows us that accountability and agency go hand-in-hand. This is why the accountability circles do not work if there are not also regular talking circles. The harm the students commit is a symptom, the tip of

Restorative Justice in High Schools  189 the iceberg, of their deeper sense of self and well-being. The talking circles form bonds between people that allow students to express and address all parts of their lives. If talking circles are not present to form the foundation of any participatory disciplinary action, the student may not feel as much a part of the community, and thus less accountable to it. Thus, if the goal of punishment is accountability, prisoners too must see themselves as agents not only in the crime they commit, but also in all other aspects of their lives, living as part of a community. Since their current community is prison, they will most likely address aspects of their life in that context, and be agents of change in criminal justice. Of course, there are serious objections to transposing RJ educational methods and outcomes onto processes involving adult offenders. First, the harm adult criminals commit is of a different scale and magnitude than juveniles. Any sort of so-called “talking cure” appears highly suspect in the face of horrific and violent crimes, offenses that are much more serious than petty fights in the hallway, marijuana in locker rooms, or swearing at a teacher. Unfortunately, this objection underestimates the danger present in many students’ lives. RJ practitioners in high schools are often worried about homelessness, domestic and gang violence, and drug addiction, not only the minor ways in which their students act out from these experiences at school. In addition, it must be noted that black and brown students are often read as “adults,” much older than they are, and less innocent than white students (APA 2014). If these children are treated as adults in the first place, and are subject to “adult” forms of punishment so early on in their education, racial disparities have already erased a clear and distinct line between “adult” and “juvenile” offenses. As the school-to-prison pipeline literature suggests, the trauma of early marginalization in schools often stays with adult prison inmates. Until racial disparities in punishment are mediated, then, in both schools and prisons, one cannot draw such a distinct line between adult and juvenile offenses. The second serious objection to applying educational RJ approaches to the criminal justice system goes like this: it is precisely because children are malleable and can change that RJ approaches in schools are so effective. Because adult offenders cannot change, mainstreaming RJ in prisons will not be as effective. In part, this is an empirical objection, and there are studies showing that even in offenders of the most horrific crimes, emotional learning is still possible (Walters 2014). However, beyond the empirical evidence, I must note again that the proliferation of zero-tolerance policies in preschools demonstrates that currently, the prevailing belief is that even children as young as four cannot “change.” Rather, American educational policies reflect a belief that some people are born “bad,” and that the best way to deal with these bad seeds is to isolate, marginalize, and banish them from our collective communities. Given that we might find that presumption to be absurd in the case of four-year-olds, we might extend that intuition of absurdity to older children and young adults as well. Also, because that

190  Johanna C. Luttrell intuition of absurdity rests on the reality of racial disparities in education, the scaffolding of a belief in unchangeability of character should become less stable. A third and real objection is that RJ approaches are not appropriate in all kinds of crimes. For example, feminist critics of RJ worry about the appropriateness of RJ in cases of sexual assault (Quince Hopkins and Koss 2005). Firstly, any meeting between a sexual assault survivor and their offender could seriously re-traumatize the victim. Re-victimization is a possibility in any victim/offender interaction, but especially so in the case of sexual assault. Further, social expectations already exist that women should be “forgiving” and compliant, and, while the goal of RJ is not necessarily forgiveness, feminists worry that forgiveness will be more expected in these cases than others. Similarly, because of the weight of those social expectations of the “forgiving woman,” women themselves could hold adaptive preferences that make the process too easy and streamlined for the offender, garnering less accountability. These are serious and valid concerns. One reply is that, while RJ has much to inform contemporary criminal justice, it is not a universal panacea to every offense. There are some situations, like sexual assault, where more traditional sentencing measures may be more appropriate. Other offenses, like victimless drug crimes, would require other methods as well. Yet, even in the most traumatizing cases, RJ principles could frame the process in a useful way. For example, the victim could be consulted in what she/he wanted: whether to be independent of sentencing processes or involved, whether they want an eventual meeting or not, whether they want an acknowledgement from the offender. Thus, even in cases where RJ circles or VOM is not appropriate, RJ principles of victim empowerment and offender accountability could still be useful (Quince Hopkins and Koss 2005; Van Wormer 2009). There is an opposite kind of concern, though, from the worry about the victim not being able to demand enough accountability from their offenders: that is, it might be the case that victims are too punitive, and specifically that the RJ methods demand confessions from offenders that are intrusive and traumatizing. Again, I must stress that this is a community endeavor, not an individualized one. As in high schools’ circles, RJ approaches are not a dialogue between two people, they are a community conversation. While no process is perfect, in the community, and in trained RJ mediators, there are ideally safeguards built in, professionals who have these same worries that all parties involved (the victims, their offenders, and their communities) are treated justly. At any rate, it would seem a worthy experiment to incorporate RJ approaches into sentencing, parole, and probation decisions. The worry about disingenuous victim/offender dialogues is a serious and remaining concern, and RJ approaches are not a universal panacea. However, again, if the aim of criminal justice measures is safety, accountability, deterrence, prevention, and/or justice, RJ measures in educational contexts have proven

Restorative Justice in High Schools  191 themselves far more effective in achieving these aims. It is worth taking a serious look at their success.

Notes 1 For a history of race and the criminal justice system, see, for instance, Douglas A. Blackmon, Slavery by Another Name (New York: Doubleday, 2008). See also Ida B. Wells Barnette, Southern Horrors: Lynch Laws in All Its Phases (Boston, MA and New York: Bedford St. Martin’s, 1993). See also Phillip Dray, At the Hands of Persons Unknown: The Lynching of Black America (New York: Modern Library, 2007). See also Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Color Blindness (New York: The New Press, 2010). See also NAACP. “Criminal Justice Fact Sheet.” naacp.org/pages/criminal-justice-fact-sheet 2 See Howard Zehr’s Changing Lenses (1990) for a full defense of the voice of victims in sentencing. 3 For an explanation of feminist standpoint theory, see, for instance, Lorraine Code, “Taking Subjectivity into Account,” in Code’s Rhetorical Spaces: Essays on Gendered Locations (New York: Routledge, 1995). See also Sandra Harding, Whose Science? Whose Knowledge? Thinking from Women’s Lives (Ithaca, NY: Cornell University Press, 1991).

References Alexander, Michelle (2010). The New Jim Crow: Mass Incarceration in the Age of Color Blindness. New York: The New Press. American Civil Liberties Union Factsheet. “What Is the School-to-Prison Pipeline?” Available at www.aclu.org/fact-sheet/what-school-prison-pipeline American Psychological Association (2014). “Black Boys Viewed as Older, Less Innocent than Whites, Research Finds.” This American Life (2014). “Is This Working?” Episode 508. Available at www. thisamericanlife.org/radio-archives/episode/538/is-this-working Bergseth, Kathleen and Bouffard, Jeffrey (2013). “Examining the Effectiveness of a Restorative Justice Program for Various Types of Juvenile Offenders.” International Journal of Offender Therapy and Comparative Criminology 57: 1054–75. Brown, Tara M. (2007). “Lost and Turned Out: Academic, Social, and Emotional Experiences of Students Excluded from Schools.” Urban Education 42: 432–55. DeLauri, Linda, and Eaton, Susan. “Building Equalizing Schools Within Inclusive Communities: Strategies that Re-Direct the School-to-Prison Pipeline in the Classroom and Beyond.” The Charles Hamilton Institute for Race and Justice at Harvard Law School. April 1, 2010. http://www.issuelab.org/resources/14506/14506.pdf. Dhami, Mandeep K., Mantle, Greg and Fox, Darrell (2009). “Restorative Justice in Prisons.” Contemporary Justice Review 12: 433–48. Flannery, Mary-Ellen (2015). “The School-to-Prison Pipeline: Time to Shut It Down.” NEA Today Magazine 33: 42–5. Gonzalez, Thalia (2012). “Keeping Kids in Schools: Restorative Justice, Punitive Discipline, and the School to Prison Pipeline.” Journal of Law and Education 41: 281–336. Harlow, Caroline Wolf (2003). “Educational and Correctional Populations: Bureau of Justice

192  Johanna C. Luttrell Statistics Special Report.” US Dept. of Justice. https://www.bjs.gov/content/pub/pdf/ ecp.pdf. Heitzig, Nancy A. (2016). The School-to-Prison Pipeline: Education, Discipline, and Racialized Double Standards. Santa Barbara: Praeger Insight Development Group, Salem, Oregon. Available at www.insightdevelopmentgroup.org Hopkins, C. Quince and Koss, Mary P. (2005). “Incorporating Feminist Theory and Insights into a Restorative Justice Response to Sex Offenses.” Violence Against Women 11: 693–723. International Institute for Restorative Practices (2009). “Improving School Climate: Findings from Schools Implementing Restorative Justice.” Accessible at www.iirp.edu/iirpWebsites/web/uploads/article_pdfs/92115_IIRP-ImprovingSchool-Climate.pdf Katz, Charles and Walker, Samuel (2012). The Police in America, 8th ed. New York: McGraw Hill. Kline, Dana M. Stewart (2016). “Can Restorative Practices Help to Reduce Disparities in School Discipline Data? A Review of the Literature.” Multicultural Perspectives 18: 97–102. Kristof, Nicholas (2014). “A Shooter, His Victim, and Race.” NY Times. Available at www.nytimes.com/2014/12/14/opinion/sunday/nicholas-kristof-a-shooterhis-victim-and-race.html?_r=0 Lewis, Sharon (2009). Improving School Climate: Finding from Schools Implementing Restorative Practices. Bethlehem, PA: IIRP Graduate School: International Institute for Restorative Justice. Loury, Glenn C., Karlan, Pamela S., Waquant, Loic and Shelby, Tommie (2008). Race, Incarceration, and American Values. Cambridge, MA: MIT Press. Mallett, Christopher (2015). The School-to-Prison Pipeline: A Comprehensive Assessment. New York: Springer. NAVSPIC VOD National Standards Subcommittee (2016). “Guide to VOD: 20 Essential Principles.” Available at http://restorativejustice.org/am-site/media/ victim-centered-victim-offender-dialogue-in-crimes-of-severe-violence.pdf Restorative Justice Collaborative of Houston (2016). Available at http://restorativehouston.wixsite.com/rjch Simson, David (2014). “Exclusion, Punishment, and Racism in our Schools: A Critical Race Theory Perspective on School Discipline.” UCLA Law Review 506: 508–563. Skiba, R. and Peterson, R. (1999). “The Dark Side of Zero-Tolerance: Can Punishment Lead to Safe Schools?” Phi Delta Kappan 80: 372–82. Stevenson, Bryan (2014). Just Mercy: A Story of Justice and Redemption. New York: Spiegel & Grau. US Department of Education (2014). “Civil Rights Data Collection Data Snapshot: School Discipline.” Available at http://www2.ed.gov/about/offices/list/ocr/docs/ crdc-discipline-snapshot.pdf Van Wormer, Katherine (2009). “Restorative Justice as Social Justice for Victims of Gendered Violence: A Standpoint Feminist Perspective.” Social Work 42: 555–67. Wadhwa, Anita (2015). Restorative Justice in Urban Schools: Disrupting the School to Prison Pipeline. New York: Routledge. Wall Street Journal (2015). “Suspension, Restorative Justice and Productive Schools.” Available at www.wsj.com/articles/suspension-restorative-justice-andproductive-schools-letters-to-the-editor-1428524841

Restorative Justice in High Schools  193 Walters, Helen (2014). “Can Prison Be a Place of Redemption?” Available at http:// ideas.ted.com/prison_rehab_from_the_inside/ Waquant, Loïc. (2008) Urban Outcasts: A Comparative Sociology of Advanced Marginality. Cambridge: Polity. Wearmouth, Janice and Berryman, Mere (2012). “Viewing Restorative Approaches to Addressing Challenging Behaviour of Minority Ethnic Students Through a Community of Practice Lens.” Cambridge Journal of Education 42: 253–68. Zehr, Howard (1990). Changing Lenses: A New Focus for Crime and Justice. Penn: Herald Press.

11 Reforming Youth Incarceration in the United States Cara H. Drinan1

Introduction The United States was once, not too long ago, a pioneer on juvenile justice matters. Legislators in Illinois invented the concept of a juvenile court at the end of the 19th century (Zimring and Tanenhaus 2014). Quickly thereafter, every state in the union instituted its own juvenile court system, and developed nations around the world emulated the American juvenile court model (Zimring and Tanenhaus 2014). The early juvenile court was premised on the notion that childhood is a period of dependency and risk; the state’s obligation was to assist a child in jeopardy, typically by providing social services (Kupchik 2006: 10–11). Dramatically and rapidly, though, the United States became an international outlier in the severity of its juvenile sentencing and correctional practices. Today, each year in America, police arrest more than one million juveniles (US Dep’t of Justice: Office of Juvenile Justice and Delinquency Prevention, Statistical Briefing Book: Juvenile Arrests), and about 250,000 of those kids are charged with a crime and processed in adult court (Campaign for Youth Justice, Key Facts: Youth in the Justice System). In some states, children as young as six can be transferred out of juvenile court into adult court without any judicial oversight.2 Once there, they face sentences—often mandatory ones—that were drafted with adults in mind. If convicted, these children are sentenced to a term of years in a correctional facility fraught with problems, not the least of which is that it was designed for adults. Until 2005, the United States was the only developed country that subjected children to the death penalty,3 and today we are the only nation that employs juvenile life without parole. Because of their physical and mental vulnerability, youth inmates experience the highest rates of sexual and physical assault,4 as well as suicide (Justice Policy Inst. 1997, The Risks Juveniles Face When They Are Incarcerated with Adults, 2). The Pope, U.N. officials, and international human rights organizations have condemned American juvenile sentencing practices. In other works I have explored the question of how American juvenile justice lost its way (Drinan 2017). At the same time, legal academics have

Reforming Youth Incarceration in the US  195 extensively documented the harms that we inflict upon minors in our criminal justice system. In this chapter, my goal is to map out three juvenile justice reform measures for lawmakers to consider today in light of the Supreme Court’s recent juvenile sentencing decisions. I begin with a brief overview of those crucial Eighth Amendment cases. I then move to argue for three specific reform measures. First, I contend that juvenile transfer to adult court should be onerous for the state and rare—that we should return to a default rule whereby juveniles are dealt with in juvenile court. Second, I make the case for abolishing mandatory minimum sentencing schemes as they apply to juveniles. Finally, I argue that, while the Supreme Court’s recent juvenile sentencing decisions have not squarely addressed conditions of confinement, the spirit of those cases calls for an overhaul of youth incarceration in America. If lawmakers pursue these and other measures, this country can begin to return to the rehabilitative ideal that prompted the creation of the juvenile court a little more than a century ago.

Recent Supreme Court Decisions on Juvenile Sentencing In the last decade, the United States Supreme Court has scaled back the extent to which juveniles may be exposed to the most severe punishments. In 2005, in Roper v. Simmons,5 the Supreme Court held that the practice of executing those who had committed their crimes prior to the age of 18 was unconstitutional.6 The Roper Court employed longstanding Eighth Amendment analysis for the capital setting: it examined juveniles as a group and asked whether the use of execution was proportionate given the diminished culpability of youth offenders.7 Further, in assessing proportionality, the Court looked at the “objective indicia of consensus, as expressed in particular by the enactments of legislatures that have addressed the question” and then exercised its own “independent judgment” as to “whether the death penalty is a disproportionate punishment for juveniles.”8 In that process, the Roper Court found that a majority of states forbade the practice of juvenile capital punishment; that it was rarely employed in the states that permitted it; and that the national trend was moving away from subjecting juveniles to the death penalty.9 At the same time, the Court focused on three reasons why juveniles are categorically different from adults and thus should not be exposed to capital punishment: they lack maturity; they are far more susceptible to external pressures; and their moral character is still fluid.10 On this basis, the Court held that the Eighth Amendment forbids juvenile execution. In 2010, in Graham v. Florida, the Supreme Court revisited the question of juveniles’ diminished culpability and held that states may not impose a life without parole (LWOP) sentence on a juvenile who commits a non-homicide crime.11 Writing for the Court, Justice Kennedy relied upon what neuroscientists had been saying for some time: given their stage of development, children are both less culpable for their crimes and more amenable to reform than adults. As a result, the Court held, a juvenile who commits a

196  Cara H. Drinan non-homicide crime may not be sentenced to die in prison: “A State is not required to guarantee eventual freedom to a juvenile offender convicted of a nonhomicide crime. What the State must do, however, is give defendants like Graham some meaningful opportunity to obtain release based on demonstrated maturity and rehabilitation.”12 Only two years later, the Supreme Court revisited the issue of extreme juvenile sentences in America and held once again in favor of the juvenile petitioner. Evan Miller had been sentenced to mandatory LWOP for a homicide he committed at the age of 14. In Miller v. Alabama, the Court held that his sentence, like Mr. Graham’s, was cruel and unusual.13 As the Miller Court explained: “[life without parole] is an ‘especially harsh punishment for a juvenile,’ because he will almost inevitably serve ‘more years and a greater percentage of his life in prison than an adult offender . . . The penalty when imposed on a teenager, as compared with an older person, is therefore ‘the same . . . in name only.’ ”14 Under the Miller decision, states may still impose an LWOP sentence upon a juvenile defendant, but only after taking into account youth and all of its mitigating attributes.15 Together, these decisions—referred to as the Miller trilogy—stand for the proposition that children are different in the eyes of the law, and state and federal sentencing practices must reflect that fact. The Supreme Court’s decisions in the Miller trilogy were about much more than LWOP for kids. The Miller Court, in particular, made clear that its opinion was an indictment of broader juvenile justice practices, such as transfer laws that permit children to be tried in adult court and sentencing guidelines that render youth irrelevant. Perhaps most important, the Court sent a clear message that kids are different—their brains have not fully developed, and thus they are less culpable and more amenable to rehabilitation. Because of this biological difference, it violates basic notions of fairness for the state to proceed with criminal charges against juveniles “as though they [are] not children.”16 State and federal actors are in the process of implementing these recent Supreme Court decisions, and it will take years for that process to fully unfold. In the meantime, though, juvenile justice advocates can pursue reform measures that flow logically from these decisions. In the following sections I address three such measures.

Reform #1: Keep Kids in Juvenile Court When the juvenile court system was created, it was difficult and rare for a child to be transferred out of that system and into the adult court system (Kupchik 2006). Today, though, transfer provisions that put kids in adult court are the norm. All states have some mechanism, and many have several, that permit adult criminal prosecution of juvenile defendants (National Criminal Justice Reference Service, Trying Juveniles As Adults: An Analysis Of State Transfer Laws And Reporting). Twenty-nine states have statutory

Reforming Youth Incarceration in the US  197 exclusion transfer laws—laws that mandate adult criminal court for children charged with certain offenses (National Criminal Justice Reference Service, Trying Juveniles As Adults: An Analysis Of State Transfer Laws And Reporting). For years, scholars have argued against transfer laws (Brink 2004), and lawyers have challenged various transfer laws in court—without success.17 Today, in light of the Miller trilogy, there is newfound traction to the claim that mandatory transfer laws are unconstitutional and nonsensical. After Miller, it is possible to challenge automatic transfer laws as impermissible “one size fits all” treatment of juveniles. In fact, the Miller Court not only took issue with conflating adult and juvenile sentencing generally, but it also criticized mandatory transfer provisions explicitly. The Miller Court explained that mandatory transfer laws, depending upon their operation, can vest prosecutors with too much unbridled discretion; can force judges into making extreme sentencing choices; and can jeopardize a child’s well-being.18 The language, logic, and science of the Miller trilogy, then, have further eroded the legitimacy of transfer laws—laws that have been under attack for decades. Many scholars have mapped out this newfound basis for challenging juvenile transfer laws (Hess 2014: 312, 331–2; Hamack 2014: 775, 805–27; Slobogin 2013: 103, 121–9; Jacobs 2013: 989, 992). For example, Tulane Law Professor Nancy Hoeffel has noted that transfer and death penalty proceedings have much in common in their stakes and in their finality. She argues that the transfer decision should be done on an individual basis, just as capital sentencing proceedings have to be, incorporating all relevant mitigation evidence (Hoeffel 2013: 51–5). Lawyers today can draw on this new law and scholarship to challenge juvenile transfer laws and to argue for a return to the default of keeping kids in juvenile court. The goal of keeping more kids in the juvenile system is not unrealistic, and, in fact, efforts to accomplish this goal are already underway in several states. To begin, while most states and the District of Columbia deal with adolescents in juvenile court until they turn 17 (at least as a default), in the last seven years, five states have amended their laws to keep children in juvenile court until they turn 18: Illinois, Connecticut, New Hampshire, Massachusetts, and Mississippi (Childress 2016). Similar efforts are ongoing in Michigan and New York (Thomas 2016). At the same time, legislators are trying to curb the power of prosecutors to “direct file” juvenile cases in adult court in California, Florida, and Vermont (Evans 2016). These efforts reflect a national recognition that both kids and society are better off when we keep juveniles in juvenile court. Given what we know today about brain development and given what the Supreme Court has done in the juvenile sentencing arena, there is no reason for children to be transferred to adult court as a matter of course. Instead, we should return to the default that a child who breaks the law ought to be dealt with in juvenile court unless a judge determines that extraordinary circumstances warrant a transfer to adult court.

198  Cara H. Drinan

Reform #2: Abolish Mandatory Minimum Sentences for Kids In its Miller trilogy, the Supreme Court acknowledged what “any parent knows”:19 that children are less mature and less responsible than adults; that children do not have the same capacity to appreciate the long-term consequences of decisions; and that children may be overwhelmed by potentially coercive environments, even when a reasonable adult would not be. Yet, while the Court has restricted the states’ ability to sentence youth to death and LWOP, it has not addressed the legitimacy of generally applicable mandatory minimums. As long as children are charged in the adult system, we must work toward the goal that youth always counts as a mitigating variable on its own at the sentencing phase. For context, it is important to understand the kinds of sentencing regimes that states employ. Sentencing guidelines range from mandatory to advisory. If a sentence is mandatory, it means that once the jury has convicted the defendant of a certain charge, the judge has no choice but to impose the sentence prescribed by the legislature for that crime. A presumptive sentencing guideline, however, suggests a predetermined sentence for a crime, but permits the judge to impose a more lenient alternative sentence if the judge determines that there are mitigating circumstances. Typically, the legislature determines in advance what mitigating factors might justify a downward departure from the presumptive sentence (Hunt and Connelly 2005; Connecticut General Assembly, Legislative Program Review and Investigations Comm., Connecticut Mandatory Minimum Sentences Briefing). Finally, advisory guidelines are voluntary in that they provide a benchmark for the sentencing judge, but the judge may depart from the suggested sentence with or without explanation (Maryland State Commission on Criminal Sentencing Policy, Sentencing Guidelines Overview). After the Miller trilogy, at a minimum, juvenile justice advocates should insist that youth itself be a relevant mitigating factor when presumptive sentencing guidelines apply. As the Miller Court explained, there are many “ ‘mitigating qualities of youth.’ ”20 Youth is a “time of immaturity, irresponsibility, ‘impetuousness[,] and recklessness,’ ” and it is a period during which “a person may be most susceptible to influence and to psychological damage.”21 Thus, youth alone should at least be permissible grounds for a judge to impose a more lenient sentence than what the presumptive guideline suggests. But not all presumptive sentencing guidelines include youth as a mitigating factor in its own right. For example, Alaska provides presumptive sentencing guidelines for felonies, and the statute separately lists aggravating factors and mitigating factors.22 The Alaska statute lists twenty separate mitigating factors that may “allow imposition of a sentence below the presumptive range.”23 Only one of the twenty mitigating factors relates to youth, and it does not recognize youth in its own right as a mitigating

Reforming Youth Incarceration in the US  199 variable. The statute permits a lesser sentence than the presumptive one if “the conduct of a youthful defendant was substantially influenced by another person more mature than the defendant.”24 Moreover, as with any of the mitigating variables, the burden is on the defendant to prove to the judge by clear and convincing evidence each mitigating factor.25 Alaska is not alone in its disregard for youth as a mitigating factor in and of itself.26 Because the Supreme Court has elevated youth in its own right to a mitigating factor of constitutional significance, states must consider youth at sentencing even in a presumptive or advisory sentencing context. Moreover, going forward, mandatory minimums should not apply to children. Not only are mandatory minimums especially harmful and unfair to children, but after Miller, they are simply unjustifiable.27 The Miller opinion is replete with discussion of process and the importance of discretion for juvenile sentencing. The Court explained: “[s]uch mandatory penalties, by their nature, preclude a sentencer from taking account of an offender’s age and the wealth of characteristics and circumstances attendant to it.”28 And later: “Mandatory life without parole for a juvenile precludes consideration of his chronological age and its hallmark features—among them, immaturity, impetuosity, and failure to appreciate risks and consequences.”29 The Miller Court was examining and speaking of LWOP, but in an earlier part of the decision, the majority recognized that “none of what [Graham] said about children—about their distinctive (and transitory) mental traits and environmental vulnerabilities—is crime specific.”30 It is also true that none of what Roper/Graham/Miller said about children is sentence-specific. The sentencing process and discretion called for by the Miller Court are simply incompatible with a mandatory sentencing scheme—whether it is a mandatory sentence of life without parole or a mandatory sentence of thirty-five years. To be sure, the claim that Miller precludes mandatory minimums for juveniles sounds radical, at least at first blush. Only one state supreme court to date has read Miller in this way.31 In an expansive opinion, documenting the evolution of juvenile justice in this country and the United States Supreme Court’s recent juvenile cases, the Iowa Supreme Court rejected the concept of mandatory minimums for children in Iowa v. Lyle.32 “Mandatory minimum sentences for juveniles are simply too punitive for what we know about juveniles,” the Court held.33 Moreover, the Court anchored its decision in its reading of Miller: “Miller is properly read to support a new sentencing framework that reconsiders mandatory sentencing for all children. Mandatory minimum sentencing results in cruel and unusual punishment due to the differences between children and adults. This rationale applies to all crimes, and no principled basis exists to cabin the protection only for the most serious crimes.”34 While several other state courts have rejected Iowa’s position,35 the Iowa Supreme Court’s reading of Miller is entirely defensible. As discussed above, the language of the Miller Court repeatedly emphasized the importance of

200  Cara H. Drinan process and discretion when sentencing juveniles. In addition to its language, the logic of the Miller decision also precludes mandatory sentencing for juveniles. The Miller Court drew on two separate strands of precedent: its cases dealing with categorical bans on certain sentencing practices and its line of cases prohibiting the mandatory imposition of capital punishment.36 The first line of cases says “that children are constitutionally different from adults for purposes of sentencing.”37 The Miller Court reiterated what Roper and Graham had recognized: that brain and social science confirm children are less culpable and more amenable to reform and that these differences must be taken into account at sentencing.38 Because the Miller Court cemented this “kids are different” approach, one cannot claim post-Miller that such differences are irrelevant outside the context of LWOP. Second, the Miller Court drew on its line of cases requiring “that capital defendants have an opportunity to advance, and the judge or jury a chance to assess, any mitigating factors,” especially those dealing with “the ‘mitigating qualities of youth.’ ”39 This line of cases requires the states to provide defendants with an opportunity to present mitigating factors that may impact the sentence—including youth, substance abuse, a history of violence within the family, developmental challenges, or traits that suggest amenability to rehabilitation. The Miller Court borrowed from this line of cases to say that kids are different and that these differences should be illuminated in an individualized, discretionary sentencing scheme. Thus, the logic of Miller, in addition to its language, suggests that mandatory minimums—schemes that preclude individual consideration of mitigating factors, including youth—are incompatible with the Miller trilogy. Beyond the question whether mandatory minimums are permissible in the wake of Miller, critics will argue that eliminating mandatory minimums for juveniles will generate unworkable administrative burdens for state courts. The dissenting Justices in the Lyle case emphasized this point. Justice Zager, for example, lamented that there were more than 100 Iowan inmates serving a mandatory sentence that was imposed upon them as a juvenile, and that revisiting those sentences would “take hundreds, if not thousands, of hours.”40 Further, the dissenting justices posited that, without mandatory minimums for juveniles, trial courts would be required to hear expert witnesses regarding juveniles’ diminished culpability.41 “In sum, ‘the trial court must consider all relevant evidence’ of the distinctive youthful attributes of the juvenile offender . . . The possibilities are nearly endless.”42 The logical conclusion of this type of administrative burden argument is that, if indeed juveniles cannot be subject to mandatory sentences, the entire process of sentencing juveniles in adult court is undermined, as determinate sentencing schemes are the national norm. That outcome does not necessarily follow. Prohibiting mandatory minimums for juveniles does not preclude their appearance in adult criminal court. It may make juvenile sentencing in adult court more time-consuming and resource intensive, but the Supreme Court has consistently held that

Reforming Youth Incarceration in the US  201 efficiency and fiscal constraints must yield to the observance of constitutional rights.43 States may not, for example, refuse to provide lawyers to poor criminal defendants or doctors to inmates because those obligations are too expensive.44 In the same way, states cannot defend a juvenile sentencing scheme simply on the grounds of administrative ease and cost containment. Further, if it is simply too onerous for states to sentence juveniles in adult court without relying upon mandatory sentencing schemes, that reality may compel prosecutors and legislators to reconsider when, and how frequently, children should be transferred to adult court. Suppose, for example, that prosecutors in Iowa do not want to pursue an adult criminal sentence except in rare cases because of the burden of justifying such sentences under the Miller factors. It may turn out that precluding juvenile mandatory minimums forces state actors to internalize the full costs of prosecuting children as adults. And it may follow that, as a result of internalizing those costs, over time, state actors charge juveniles as adults only very sparingly. Given what science has revealed about juveniles and their capacity for change, and given the Supreme Court’s incorporation of that science, such an outcome seems logical. Moreover, such an outcome—the reluctant charging of children in adult court—would merely be a return to the juvenile justice model that was founded in this country more than a century ago.

Reform #3: Rehabilitation, Not Incarceration, for Youth In its recent juvenile sentencing decisions, the Supreme Court has focused on what sentences the states may impose rather than the conditions under which juveniles are required to serve those sentences. Yet, the Court has repeatedly expressed concern with the vulnerability of youth in its recent juvenile Eighth Amendment cases, and in other constitutional settings, too. Juvenile justice advocates should leverage the Court’s emphasis upon the vulnerability and susceptibility of youth to seek improved conditions of confinement for youth in the years to come. We currently rely upon youth detention where it is unnecessary and where it routinely does more harm than good. Only about 25% of children in detention facilities have been charged with or convicted of one of the four most serious felonies, and nearly a third are serving time based on a public order violation, a technical parole or probation violation or a status offense (Department of Justice, Juveniles in Residential Placement, 2013). At the same time, the majority of incarcerated youth enter the system with a mental health or special education need that often goes unmet and with a history of trauma that is only exacerbated by the hostility of detention. Once a child enters the correctional system he or she is more likely to have diminished educational and employment opportunities and to commit crime in the future. While we are spending almost 6 billion dollars a year on youth corrections, the prevailing method of juvenile detention is an utter failure: between 70 and 80 percent of juveniles released from detention centers are

202  Cara H. Drinan rearrested within two or three years for a new offense (The Annie E. Casey Foundation, The Missouri Model: Reinventing the Practice of Rehabilitating Youthful Offenders, 2). Youth incarceration is a tremendous waste of human potential and taxpayer dollars. Going forward, incarceration for juveniles should be a rare measure of last resort, and it should be rehabilitative when it needs to happen. In recent years, states like Alabama, California, Louisiana, New York, and North Carolina have reduced their detained youth populations, and these jurisdictions can serve as a model for keeping youth out of the correctional system (The Annie E. Casey Foundation, The Missouri Model, 5). For example, Alabama lawmakers passed a juvenile justice reform act in 2008 that recognized the state was confining too many youth—often for minor offenses (Cason 2015). Since its passage, the state has cut its detained youth population roughly in half and has been able to redirect the money spent on incarceration to diversion programs set up in 52 counties. Through these diversion programs, low-risk youth have access to substance abuse and mental health treatment, as well as counseling, and they avoid the harmful effects of incarceration. Similarly, in the 1990s the Annie E. Casey Foundation began exploring alternatives to juvenile incarceration and launched the Juvenile Detention Alternatives Initiative (JDAI) (The Annie E. Casey Foundation, Juvenile Detention Alternatives Initiative: Progress Report 2014). The goal of the JDAI was to demonstrate in pilot locations that youth detention populations could be substantially reduced without impacting public safety (The Annie E. Casey Foundation, Progress Report 2014: 8). JDAI jurisdictions use objective indicators to screen out youth who do not pose a risk to public safety and to offer appropriate alternatives to detention, such as mental health treatment and Multisystemic Therapy, a family and community based approach to addressing the root cause of delinquency. As of 2013, more than 250 localities in 39 states and the District of Columbia were participating in the JDAI (The Annie E. Casey Foundation, Progress Report 2014: 10). Overall, these participating jurisdictions have achieved a cumulative 43% reduction in detention populations (The Annie E. Casey Foundation, Progress Report 2014: 13), and 38 JDAI sites have reduced their average daily populations in detention by 50 percent or more (The Annie E. Casey Foundation, Progress Report 2014: 15). At the same time, JDAI sites have experienced reductions in youth crime indicators, such as juvenile arrests and felony petitions (The Annie E. Casey Foundation, Progress Report 2014: 17–18). These jurisdictions demonstrate that diverting youth away from incarceration can be a win-win proposition for taxpayers and youth. In the small percentage of cases where youth represent a risk to society and must be incarcerated, we need to overhaul the mode of detention altogether. Large prison-like detention centers are the most common mode of incarceration for kids in America (Nat’l Center for Juvenile Justice, Juvenile

Reforming Youth Incarceration in the US  203 Offenders and Victims: 2006 National Report), and they are the least effective, with recidivism rates hovering around 75% (The Annie E. Casey Foundation, The Missouri Model, 2).45 However, there is a well-documented, affordable and more effective alternative that has been successfully employed in Missouri for nearly four decades with bipartisan support.46 Instead of warehousing youth for long periods of time in large centers that offer few services, Missouri shifted to small centers that emphasize routine, therapy, and relationship-building for detained youth. Politicians across the political spectrum, non-profit organizations, and mainstream media outlets have applauded this “Missouri Model” for its efficacy, humanity, and affordability. The Missouri youth corrections model emphasizes treatment over punishment, and it has six defining features (The Annie E. Casey Foundation, The Missouri Model, 13–14). First, Missouri youth who require confinement are placed in small facilities near their homes when possible. Second, the correction model focuses heavily on group therapy and group accountability rather than leaving youth to find their own way among equally troubled youth in a large prison environment. Third, the model employs staff who are trained in therapy and supportive treatment, eschewing guards and coercive techniques that are the norm in juvenile prisons. Fourth, confined youth in Missouri are engaged in educational and vocational training that will enable them to transition back into society. At the same time they spend a great deal of time in group therapy seeking to discern the root cause of their delinquency so that they can avoid repeating that behavior in the future. Fifth, when possible, the staff in Missouri’s system seeks to include family members in youth treatment and after-care plans. Finally, when youth are deemed rehabilitated and ready to return to society, Missouri provides intensive transition resources and after-care support as youth re-enter educational and job opportunities. The radical differences between confinement in Missouri and in other states’ juvenile correctional facilities cannot be overstated. Instead of being in an environment with one hundred other delinquent youth, in Missouri, the largest youth confinement center has only 50 beds, and within facilities youth are assigned to 12-member teams of kids with whom they work (The Annie E. Casey Foundation, The Missouri Model, 15, 20). Instead of sleeping on metal cots in cells, Missouri kids are in carpeted rooms with bunkmates and walls decorated with personal effects (The Annie E. Casey Foundation, The Missouri Model, 19). While many detained youth across the country are simply sitting in a cell waiting for their release date, kids in the Missouri system are scheduled all day, primarily working on academics, work details, and group therapy sessions. Moreover, Missouri youth are serving indeterminate sentences; release hinges not on the passage of time but on personal growth. Before they are released, Missouri youth must demonstrate self-awareness, personal growth, an ability to control their emotions, and the capacity to live and work in a community.

204  Cara H. Drinan Perhaps the biggest difference between the juvenile incarceration status quo and the Missouri model is its underlying value system. The Missouri model is predicated on the “belief that delinquent youth can succeed and the expectation that most will” (The Annie E. Casey Foundation, The Missouri Model, 47). With this value in mind, the entire program is geared toward enabling that success, giving youth the tools to understand their past behaviors, learn from them and develop skills to make alternative choices in the future. Not only do youth develop self-awareness through group therapy with their peers, but also they develop leadership skills and responsibility for others. For example, while the group of 10–12 kids in each team “checks in” at least five times a day to discuss the group dynamics, share feelings, and air tensions, at any point in time staff members and the youth themselves can “call a circle” (The Annie E. Casey Foundation, The Missouri Model, 29). When a circle is called, all team members sit or stand facing each other so that a grievance can be aired or a conflict can be addressed. In this way, the youth learn self-care and regulation, but also interaction with others and positive conflict resolution. It is easy to be cynical about delinquent youth and to dismiss the Missouri model as something only a Pollyanna would seriously consider. Yet the Missouri model has been in use for nearly four decades, and it has a proven track record of success that far outstrips national average indicators of success. To begin, the Missouri model is far more effective at treating the underlying cause of delinquency than the traditional juvenile detention model. While recidivism measurements vary widely by state, Missouri’s rates are typically half that of other states (The Annie E. Casey Foundation, The Missouri Model, 6–7). When looking at Missouri youth three years after release, more than 65% were law abiding and had not been recommitted to the youth system or the adult system (The Annie E. Casey Foundation, The Missouri Model, 7 fig. 4). Despite the fact that the Missouri model rejects the use of coercive correctional measures like handcuffs and solitary confinement, its residential centers are safer than other facilities. In 2006, for example, Ohio’s Youth Corrections performed a comparison of Ohio and Missouri detained youth safety. Even though Ohio had a little more than twice the amount of youth in confinement, Ohio had four times as many youth-on-youth assaults, and nearly seven times as many youth-on-staff assaults as Missouri (The Annie E. Casey Foundation, The Missouri Model, 9). Perhaps most revealing on the safety front, since closing its traditional detention centers, Missouri has not had a single youth in detention suicide (The Annie E. Casey Foundation, The Missouri Model, 10). Because of the physical and emotional safety in Missouri’s facilities, its detained youth are experiencing educational gains that detained youth elsewhere are not. One such indicator is the percentage of confined youth who are making at least one year of academic progress for every year in confinement. The national average is 25%, while in Missouri nearly 75% of detained youth are staying on track academically if not accomplishing more than they could in a public school setting (The Annie E. Casey Foundation, The Missouri Model, 10). Finally, Missouri is achieving all of this success

Reforming Youth Incarceration in the US  205 for its detained youth at a fraction of the cost for juvenile detention in other states. Criminologists estimate that “steering just one high-risk delinquent teen away from a life of crime saves society $3 million to $6 million in reduced victim costs and criminal justice expenses, plus increased wages and tax payments over the young person’s lifetime” (The Annie E. Casey Foundation, The Missouri Model, 12). In short, the Missouri model is not only more humane for the youth it treats, but also it is more productive and cost-effective in the long run. To date, a number of states and localities have attempted to pursue the Missouri model of juvenile justice. Louisiana, Washington, DC, and counties within New Mexico, Virginia, and California are trying to implement pieces of the Missouri model (Taylor-Thompson 2014: 143, 198, n. 389; see also The Missouri Model at 51–52). Yet, “emulation of the system has been slower than one might expect” (Taylor-Thompson 2014: 143, 197). Going forward, the Missouri model is clearly superior, and it ought to be aggressively pursued nationwide.

Conclusion In recent years the Supreme Court has provided legal precedent and moral leadership that juvenile justice advocates can leverage today. The three proposals outlined herein—returning to a default rule of keeping kids in juvenile court, abolishing mandatory minimum sentences for kids, and overhauling youth incarceration—are all good starting points for correcting the course of American juvenile justice.

Notes 1 Extracts (c.7000w) from “Introduction,” Chapter 5, “Progress and Hope from the Nation’s High Court”, and Chapter 7, “A War for Kids” from The War on Kids: How American Juvenile Justice Lost its Way, edited by Drinan, Cara (2017). By permission of Oxford University Press, USA. 2 Miller v. Alabama, 132 S. Ct. 2455, 2473, n.14 (2012). 3 Roper v. Simmons, 543 U.S. 551 (2005) (abolishing the death penalty for minors). 4 Cf. Prison Rape Elimination Act, 42 U.S.C.A. § 15601(4) (2003)(noting that young inmates were five times more likely to be raped in adult facilities than juvenile ones, often within 48 hours of incarceration). 5 Roper v. Simmons, 543 U.S. 551, 578 (2005). 6 543 U.S. 551, 578. 7 Id. at 564. 8 Id. at 564. 9 Id. at 567–8. 10 Id. at 569–70. 11 Graham v. Florida, 560 U.S. 48, 82 (2010). 12 560 U.S. 48, 75. 13 Miller v. Alabama, 132 S. Ct. 2455 (2012). 14 132 S. Ct. 2455, 2466. 15 Id. at 2468. 16 Id. at 2466.

206  Cara H. Drinan 17 See e.g., In re Interest of D.M.L., 254 N.W.2d 457, 459 (S.D. 1977)(rejecting claim that juvenile transfer statute was unconstitutionally vague); Commonwealth v. Cotto, 753 A.2d 217, 224 (Pa. 2000)(rejecting claim that juvenile transfer statute violated Due Process Clause). 18 Miller, 132. S. Ct. 2455, 2474. 19 Miller, 132 S. Ct. 2455, 2464 (citation omitted). 20 132 S. Ct. 2455, 2467 (citation omitted). 21 Id. (citations omitted). 22 Alaska Stat. § 12.55.155 (c)(2014)(listing aggravating factors); id. § 12.55.155 (d)(listing mitigating factors). 23 Id. § 12.55.155 (d). 24 Id. § 12.55.155 (d)(4). 25 Id. § 12.55.155 (f)(1). 26 See e.g., Considerations in Imposing Sentence, Ind. Code § 35-38-1-7.1 (2015) (listing eleven mitigating circumstances the court may consider, none of which relate to youth); Imposition of Presumptive Sentence; Jury Requirements; Departure Sentencing; Substantial and Compelling Reasons for Departure; Mitigating and Aggravating Powers, Kan. Stat. Ann. 21-6815 (c)(1)(2014)(listing non exhaustive mitigating factors, none of which include youth); Factors to Consider in Felony Sentencing, Oh. Stat. Ann. § 2929.12 (2014)(same). 27 I first made this argument in two related law review articles. See Cara H. Drinan, “Misconstruing Graham & Miller,” Washington University Law Review 91 (2014): 785, 789 n. 26;Cara H. Drinan, “The Miller Revolution,” Iowa Law Review 101 (2016): 1787, 1816–24; Martin Guggenheim, “Graham v. Florida and a Juvenile’s Right to Age-Appropriate Sentencing,” Harvard Civil Rights-Civil Liberties Law Review 47 (2012): 457. 28 Miller v. Alabama, 132 S. Ct. 2455, 2467. 29 Id. at 2468. 30 Id. at 2465. 31 It is worth noting that other state legislatures have implicitly recognized the harshness of mandatory minimum sentences as they apply to children in adult court. For example, Oregon and Washington statutory law prohibit mandatory minimums for juveniles tried as adults, except for very serious crimes, such as aggravated murder. See Or. Rev. Stat. § 161.620 (2003)(providing juvenile tried as an adult shall not receive a mandatory minimum sentence except for aggravated murder or felonies committed with a firearm); see also Wash. Rev. Code. Ann. § 9.94A.540(3)(a)(2010)(prohibiting mandatory minimum sentences for juvenile offenders except for aggravated first-degree murder). 32 State v. Lyle, 854 N.W. 2d 378 (Iowa 2014). 33 854 N.W. 2d 378, 400. 34 Id. at 402. 35 See e.g., People v. Banks, 36 N.E. 3d 432, 506 (Ill. 2015)(mandatory minimum sentence for first-degree murder did not violate juvenile’s Eighth Amendment rights); State v. Vang, 847 N.W. 2d 248, 262–3 (Minn. 2014)(mandatory minimum of life with possibility of parole after thirty years for first-degree felony murder for juvenile did not violate Eighth Amendment); Comm. v. Lawrence, 99 A.3d 116, 121–2 (Pa. Super. Ct. 2014)(statute giving mandatory minimum of thirty-five years to juvenile defendant convicted of first-degree murder did not violate Eighth Amendment). 36 Miller v. Alabama, 132 S. Ct. 2455, 2463–4. 37 Miller, 132 S. Ct. 2455, 2464. 38 Id. at 2464–5. 39 Id. at 2467. 40 854 N.W.2d 378, 419 (Zager, J., dissenting).

Reforming Youth Incarceration in the US  207 41 Id. 42 Id. at 420. 43 See e.g., Gideon v. Wainwright, 372 U.S. 335 (1963) (holding Sixth Amendment right to counsel applies to states and thus imposing burden on states to pay for that representation); Brown v. Plata, 131 S. Ct. 1910 (2011) (affirming finding of Eighth Amendment violation due to prison overcrowding and requiring state to either improve conditions at state’s expense or release inmates). 44 Id. 45 “In state after state, 70 to 80 percent of juveniles released from youth corrections facilities are rearrested within two or three years for a new offense.” 46 See generally The Missouri Model; Marian Wright Edelman, “Juvenile Justice Reform: Making the ‘Missouri Model’ the American Model,” The Huffington Post, May 25, 2011. [online]. www.huffingtonpost.com/marian-wright-edelman/juvenile-justice-reform-m_b_498976.html.

References The Annie E. Casey Foundation (2010). “The Missouri Model: Reinventing the Practice of Rehabilitating Youthful Offenders,” 2. Available at http://static1.1. sqspcdn.com/static/f/658313/9749173/1291845016987/aecf_mo_fullreport_ webfinal.pdf?token=qFAbECf2WN5xmvuKsMQloWE58d0%3D, accessed December 29, 2016 The Annie E. Casey Foundation. “Juvenile Detention Alternatives Initiative: Progress Report 2014.” Available at http://cms.aecf.org/m/resourcedoc/aecf-2014JDAI ProgressReport-2014.pdf#page=5, accessed December 29, 2016. Brink, David (2004). “Immaturity, Normative Competence, and Juvenile Transfer: How (Not) to Punish Minors for Major Crimes.” Texas Law Review 82, 1555–85. Campaign for Youth Justice (2012). “Key Facts: Youth in the Justice System.” Available at www.campaignforyouthjustice.org/documents/KeyYouthCrimeFacts.pdf, accessed December 28, 2016. Cason, Mike (2015). “New DYS Girls Facility Reflects Alabama’s Reformed Approach to Juvenile Crime.” Available at www.al.com/news/index.ssf/2015/10/ new_dys_girls_facility_reflect.html, accessed December 29, 2016 Childress, Sarah (2016). “More States Consider Raising the Age for Juvenile Crime.” Frontline, June 2. Available at www.pbs.org/wgbh/frontline/article/more-statesconsider-raising-the-age-for-juvenile-crime/, accessed December 28, 2016. Connecticut General Assembly (2005). “Legislative Program Review and Investigations Comm., Connecticut Mandatory Minimum Sentences Briefing.” Available at www.cga.ct.gov/2005/pridata/Studies/Mandatory_Minimum_Senteces_Briefing.htm, accessed December 29, 2016. Department of Justice (2016). “Juveniles in Residential Placement, 2013.” Available at www.ojjdp.gov/pubs/249507.pdf?ed2f26df2d9c416fbddddd2330a778c 6=ixggxbbxfa-icgedfbc, accessed December 29, 2016. Drinan, Cara H. (2014). “Misconstruing Graham & Miller.” Washington University Law Review 91, 785–795. Drinan, Cara H. (2016). “The Miller Revolution.” Iowa Law Review 101, 1787–1832. Drinan, Cara H. (2017). The War on Kids: How American Juvenile Justice Lost Its Way. Oxford: Oxford University Press.

208  Cara H. Drinan Edelman, Marian Wright (2011). “Juvenile Justice Reform: Making the “Missouri Model” the American Model.” The Huffington Post, May 25. Available at www. huffingtonpost.com/marian-wright-edelman/juvenile-justice-reform-m_b_498 976.html Evans, Brian (2016). “ ‘Raise the Age,’ ‘Direct File,’ and More: States Pursuing Youth Justice Reforms in 2016.” Campaign for Youth Justice, May 11, 2016. Available at http://cfyj.org/news/blog/tag/Campaign%20for%20Youth%20Jus tice, accessed December 29, 2016. Guggenheim, Martin (2012). “Graham v. Florida and a Juvenile’s Right to Age-Appropriate Sentencing.” Harvard Civil Rights-Civil Liberties Law Review 47, 457–500. Hamack, Brice (2014). “Go Directly to Jail, Do Not Pass Juvenile Court, Do Not Collect Due Process: Why Waiving Juveniles into Adult Court Without a Fitness Hearing Is a Denial of their Basic Due Process Rights.” Wyoming Law Review 14, 775–828. Hess, Wendy N. (2014). “Kids Can Change, Reforming South Dakota’s Juvenile Transfer Law to Rehabilitate Children and Protect Safety.” South Dakota Law Review 59, 312–342. Hoeffel, Janet C. (2013). “The Jurisprudence of Death and Youth: Now the Twain Should Meet.” Texas Tech Law Review 46, 29–69. Hunt, Kim S. and Connelly, Michael (2005). “Advisory Guidelines in the Post-Blakely Era: Federal Sentencing Report.” Federal Sentencing Reporter 17, 233–242. Jacobs, Rachel (2013). “Waiving Goodbye to Due Process: The Juvenile Waiver System.” Cardozo Journal of Law & Gender 19, 989–1017. Justice Policy Inst. (1997). “The Risks Juveniles Face When They Are Incarcerated with Adults,” 2. Available at www.justicepolicy.org/images/upload/97-02_rep_ riskjuvenilesface_jj.pdf, accessed December 28, 2016 Kupchik, Aaron (2006). Judging Juveniles: Prosecuting Adolescents in Adult and Juvenile Courts. New York: New York University Press. Maryland State Commission on Criminal Sentencing Policy. “Sentencing Guidelines Overview.” Available at www.msccsp.org/Guidelines/Overview.aspx, accessed December 28, 2016 Nat’l Center for Juvenile Justice (2006). “Juvenile Offenders and Victims: 2006 National Report.” Available at www.ojjdp.gov/ojstatbb/nr2006/downloads/ chapter7.pdf, accessed December 29, 2016. National Criminal Justice Reference Service (2011). “Trying Juveniles As Adults: An Analysis of State Transfer Laws and Reporting,” 2. accessed December 28, 2016. Slobogin, Christopher (2013). “Treating Juveniles Like Juveniles: Getting Rid of Transfer and Expanded Adult Court Jurisdiction.” Texas Tech Law Review 46, 103–132. Taylor-Thompson, Kim (2014). “Minority Rule: Redefining the Age of Criminality.” New York University Review of Law & Social Change 38, 143–200. Thomas, Jeree (2016). “Support of Michigan’s Bill Package to Raise the Age.” Available at http://cfyj.org/news/blog/tag/Campaign%20for%20Youth%20Justice, accessed December 28, 2016. U.S. Department of Justice: Office of Juvenile Justice and Delinquency Prevention. “Statistical Briefing Book: Juvenile Arrests.” Available at www.ojjdp.gov/ ojstatbb/crime/qa05101.asp?qa, accessed December 28, 2016 Zimring, Franklin E. and Tanenhaus, David S. (2014). Introduction to Choosing the Future for American Juvenile Justice. New York: New York University Press.

12 Policing for “Profit” The Political Economy of Private Prisons and Asset Forfeiture Abigail R. Hall and Veronica J. Mercier

1 Introduction In 2016, the US Department of Justice (DOJ) announced that it would end its use of “private prisons” in the United States, facilities which are owned or operated by private firms on behalf of state and federal authorities. The memo, released by Deputy Attorney General Sally Q. Yates, states the DOJ decision comes in light of data showing that private prisons do not provide the same level of correctional services, fail to save money, and do not maintain the same level of safety and security as government-run facilities (Yates 2016). As a result, the DOJ directive outlined intentions to phase out the use of private prisons by declining to renew contracts. The goal is to significantly reduce and ultimately eliminate the use of these facilities (Ibid). The announcement comes after years of expansion in the use of private prison facilities throughout the United States. As the number of federal prisoners increased some 800 percent between 1980 and 2013, the Bureau of Prisons, unable to accommodate so many inmates, turned to private corporations to fill the gap. The United States now incarcerates between 1.8 and 2 million citizens on any given day (Dyer 2000: 1). Some 130,000 are housed in private facilities (Carson 2015: 1). Studies on private prisons offer mixed reviews. Some researchers find that contracted prisons offer a minimal cost advantage to their public counterparts. Reports consistently find that privately contracted prisons provide lower quality of service (Mason 2012: 10). The problems with private prisons have been attributed to the supposed “private” nature of these firms. In particular, it’s posited that the introduction of profit to the prison system and the injection of the “free market” has led to inevitable negative outcomes (Ibid: 7). Private prisons are not the only part of the criminal justice system where “profit” has been labeled a problem. In recent years, the practice of civil asset forfeiture, programs that allow law enforcement agencies to seize and retain a portion or all of an asset involved in criminal activity, has come under increased scrutiny. Once again, the introduction of profit and the market into the criminal justice system (CJS) appears to lead to poor outcomes.

210  Abigail R. Hall and Veronica J. Mercier Although profit is the supposed villain in these cases, the problems observed with private prisons, asset forfeiture, and the CJS as a whole are not an intrinsic feature of private enterprise. Profit seeking is not inherently nefarious. In fact, economists have observed that profit seeking in the free market actually incentivizes an individual to serve his fellow man as opposed to taking advantage of him (Smith 1981: 25–30). The problems associated with “profit” in the CJS are the result of the institutional structure of the system itself. The CJS does not operate with the profit and loss and price signals that serve as checks and balances in the free market. Instead, all aspects of the CJS operate under a regime of overlapping bureaucracies. This creates and perpetuates a myriad of perverse incentives for both public officials and private corporations. In this chapter, we explore how the institutional structure of the CJS and the relationships between individual private actors and government bureaucracy has led to a variety of perverse outcomes over time. Beginning with the War on Drugs, we examine how asset forfeiture as a part of police operations has led to an erosion of private property. Moreover, we analyze how this practice has weakened individual rights and created an arena in which police face incentives to undertake more frequent and more aggressive operations. Likewise, we discuss how the incorporation of private facilities into the US prison system has formed a powerful special interest and fostered an ever-expanding prison population. We proceed as follows. Section 2 provides a discussion of the War on Drugs and how these policies worked to change and shape the criminal justice system. Within this context, we provide a brief overview of the economics of bureaucracy and highlight important differences between profit in the setting of a free market and “profit” under bureaucracy. Section 3 examines in more detail the use of “private” or contracted prisons throughout the United States. Section 4 analyzes the use of civil asset forfeiture by law enforcement. Section 5 concludes.

2 The War on Drugs and the Criminal Justice System in the United States In order to understand the current problems facing the criminal justice system, it is imperative to first gain an appreciation for US drug policy. In particular, it is necessary to understand how the War on Drugs, beginning more than forty years ago, has altered the criminal justice system. Drug policy in the United States has a long history dating back more than a century (see Boettke, Coyne, and Hall 2013; Reford and Powell 2016). Beginning in the early 1900s, the US federal government began programs of prohibition, banning the sale and use of particular substances like opium, marijuana, and cocaine, among others. For our purposes, the most important changes in US drug policy come from the “War on Drugs” begun under the Nixon administration in the early 1970s. This War on Drugs, which continues in full force today, is fundamentally different from other

Policing for “Profit”  211 wars throughout US history. Looking at engagements like WWI, WWII, the Korean War, and Vietnam, the enemy of the United States is not only clearly defined, but wholly external to the country. The enemies of the War on Drugs, however, have many different faces. Although the War on Drugs features some external enemies (i.e., Latin American drug cartels), other targets are internal to the United States—and they include US citizens. Domestic drug manufacturers, dealers, and users—both real and potential—are just as much targets as a foreign adversary. This change in enemy is particular important from the perspective of nearly all aspects of the criminal justice system. For example, police, historically tasked with upholding domestic laws, are now on the “front lines” of a larger war waged by the federal government (see Hall and Coyne 2013: 486–7). This not only impresses on law enforcement to undertake new activities related to drug interdiction policy, but has potentially massive implications for the budgets of law enforcement, as discussed below. In addition to impacting state and local police, the War on Drugs has also drastically altered the prison system and ushered in an era of mass incarceration. Prior to the launch of the drug war and the creation of the Drug Enforcement Agency (DEA) in 1972, approximately 196,000 Americans were incarcerated (Bureau of Justice Statistics 1982). Since the beginning of the War on Drugs, the prison population has exploded. By 1981, the prison population had more than doubled to over 353,000 (Ibid). In 2014, more than 2.2 million people in the United States were in either jail or prison. Another 4.7 million were under some sort of supervision, placing the total correctional population above 6.8 million (Bureau of Justice Statistics 2016: 2). This has led to a scenario in which many jails and prisons, built for much smaller populations, are not only overcrowded, but also persistently inundated with inmates. In addition to understanding how the War on Drugs has had lasting effects on the CJS, it is likewise essential to understand the institutional structure in which agents working in policing and the prison system operate. Each part of the criminal justice system, from police departments, to prisons, to prosecutors and district attorneys, operates as a bureaucracy. This structure has a variety of important implications, including understanding how the aforementioned “profit” works within the current CJS. In a free market system, profit serves as a powerful signal and incentive to producers. Positive profits indicate that a firm is providing valuable goods to society, as indicated by consumers’ willingness and ability to pay. Negative profits (losses), meanwhile, illustrate to producers that their outputs are failing to meet the wants of consumers. In the case of the free market, these profit and loss signals entice producers to make those goods and services most desired by consumers. The term “profit” in the context of bureaucratic institutions is inaccurate and misleading. Government agents in the CJS do not compete for profit or the business of consumers in the free market. Instead, resources are instead

212  Abigail R. Hall and Veronica J. Mercier allocated via politics. Existing literature on the economics of bureaucracy indicates that, in the absence of profit and loss signals found in the market, success in a bureaucracy is measured by the number of subordinates within the organization and the size of the bureau’s discretionary budget (Niskanen 1971, 1975; Migué and Bélanger 1974). As a result, bureaus within the CJS do not compete against other firms for genuine profits, but instead compete with other bureaucratic agencies for government resources. This has important consequences. Bureaus, facing the possibility of securing greater gains, in the form of larger budgets, more personnel, etc., are incentivized to engage in two distinct but important activities. First, bureaus face an incentive to undertake intense rent-seeking behavior, in which agents work to obtain favorable rules for their organizations and as much government funding as possible. Second, government bureaus are more likely to engage in “mission creep,” whereby bureaucratic organizations continuously expand their portfolio of activities in order to signal their relevance (and subsequently why they should be offered additional funds and personnel). Taken together, the institutional and other changes brought about by the War on Drugs, combined with the bureaucratic nature of the CJS offer a clear and powerful framework for understanding a variety of problems within the system and help to illustrate the modern expansion and perpetuation of these problems.

3  Private Prisons In 1968, the prison population and incarceration rate in the United States were the lowest they had been since the 1920s. Some 188,000 Americans were in prison that year and the incarceration rate was 93 per 100,000 individuals (Bureau of Justice Statistics 1982). The following year, Richard Nixon presented a special message to the Congress on the “Control of Narcotics and Dangerous Drugs,” arguing that the use of drugs was a threat to national health and the “safety of millions of Americans” (Nixon 1969). Nixon suggested that the urgency of the “drug problem” required a new national policy that was tough on drug manufacturing, sales, and use. That policy would come but two years later in 1970 when he signed the Controlled Substances Act (CSA) into law. The Drug Enforcement Administration (DEA) was created in 1973. The modern “War on Drugs” had begun. The War on Drugs that began under Nixon is critical for understanding the rise in the prison population in the United States, as well as the rise of private prisons. After the creation of the DEA, for example, the prison population increased by 7.1 percent, or 150,000 inmates, in the span of eight years. Compare this to the rate of expansion over the entire period of 1925–1970, in which the population increased by 2.4 percent (Bureau of Justice Statistics 1982: 2). By the end of 1981, the incarceration rate was 153 per 100,000, the highest ever recorded at the time (Ibid). Between 1970

Policing for “Profit”  213 and 2005, the US prison population grew by 700 percent (American Civil Liberties Union 2011: 11). Ronald Reagan continued Nixon’s anti-drug and “tough on crime” policies, ushering in even stricter counter drug programs. The Comprehensive Crime Control Act of 1984, for example, largely revised the US criminal code. The Act included the Armed Career Criminal Act, which provides for “sentence enhancements,” or harsher sentences for those convicted of committing particular crimes with firearms. As part of the Act, the United States Sentencing Commission was established. This new independent agency of the judicial branch of government is responsible for establishing sentencing guidelines. As a result of these changes and other changes during the period, the federal penalties for the cultivation, sale, and possession of marijuana and other drugs all increased (Harrison, Backenheimer, and Inciardi 1996). Other laws established mandatory minimum sentences, meaning those who committed certain crimes are guaranteed a certain amount of jail time. “Truth in sentencing” laws were also enacted. These laws sharply curtail probation and parole eligibility, even for those who are nonviolent and show “good behavior.” The period also saw the implementation of “three strikes” laws, mandates that greatly extend the sentences of those individuals with prior felonies. Taken together, these changes contributed to the unprecedented expansion in the prison population. The use of prisoners for private gain has a long history in the US, but privately contracted prisons did not exist in the United States until the reforms implemented by Nixon and Reagan as part of the War on Drugs took effect. With such a rapid increase in the number of individuals housed within the criminal justice system as a result of the new drug policies, jails and prisons became largely overcrowded. In an effort to manage these swelling numbers and prison overcrowding, the government began contracting with privately operated correctional facilities to house federal inmates. The number of private prisons in the United States quickly expanded. Private corporations, recognizing the opportunity to gain from the government’s drug policy, worked to grab a share of the government’s budget. By 1990, some 67 contracted facilities were operational, housing an average of nearly 7,000 prisoners. Between 1990 and 2009, the number of individuals incarcerated in private facilities increased by more than 1,600 percent, from approximately 7,000 to 129,000 inmates (ACLU 2011: 11). As of 2015, about 12 percent of inmates are housed in these facilities (Denvir 2016). Private prisons were supposed to solve or alleviate many of the problems of the prison system. Private facilities were intended to reduce costs while providing safety and offering rehabilitation programs to inmates. Their supposed “free market” nature was likewise intended to increase accountability, as private firms could be fined or fired for poor performance. Despite these intentions, however, the outcomes have been less than admirable. Various studies conducted by local, state, and federal authorities have found that private facilities may not only fail to provide a cost advantage

214  Abigail R. Hall and Veronica J. Mercier to taxpayers, but may be more expensive than their state-run counterparts (ACLU 2011: 19). Several studies have also found private prisons may experience more violence. According to the Bureau of Justice Assistance, the rate of inmate-on-inmate assaults in private prisons is higher than in state-run facilities (Bureau of Justice Assistance 2001: 46). Rates of assault on staff, prison riots, fires, and other violent acts are comparable across facility types (Ibid). The “profit motive” is frequently blamed for these failures. It’s argued that private prisons fail because, unlike government institutions, their goal is profit. Such sentiments are easily observable. The US Catholic Bishops, for example, stated that, We . . . question whether private, for-profit corporations can effectively run prisons. The profit motive may lead to reduced efforts to change behaviors, treat substance abuse, and offer skills necessary for reintegration into the community. (quoted in Glenmary Home Missioners 2016, emphasis added) The Presbyterian Church (2003) offered similar opposition to the “profit motive.” Since the goal of for-profit prisons is earning a profit for their shareholders, there is a basic and fundamental conflict with the concept of rehabilitation as the ultimate goal of the prison system. While the controversy surrounding private prisons is understandable, it is not the profit motive of the free market that creates the problem, but the intermingling of special interests with the bureaucratic structure of the US prison system. The private prison industry does not align with a traditional understanding of private enterprise, in which private actors interact with one another in hopes of making a profit. In a truly private system, firms compete within the marketplace for customers. The current system of “private prisons” is a true misnomer. The firms operating these facilities do not compete out in the market with other private actors. In a truly private system, firms operating prisons would compete with one another, facing the checks and balances inherent in the marketplace. Those who would do their job well would earn a profit. Those who would not perform well would earn a loss, facing eventual bankruptcy if changes were not made. No one or few organizations would enjoy any sort of state-conferred monopoly privilege. Today’s “private” prisons face no such competition or the checks and balances such competition provides. Instead, these institutions operate within the realm of government bureaucracy, competing, not for profit and customers in the free market, but for pieces of government budgets. Moreover,

Policing for “Profit”  215 the entities operating privately contracted facilities operate as monopolies. This is not because of some inherent feature of the market. Instead, the monopoly privilege enjoyed by these firms results from the interaction between government and special interest groups. Three companies—Corrections Corporation of America (CCA), GEO Group, and Management and Training Corporation of America (MGT), dominate the private prison industry. The two largest corporations, CCA and GEO, bring in more than $3 billion in revenue annually (Cohen 2015). To illustrate that these firms compete for government funds as opposed to traditional profit, consider the following statement from CCA, submitted in a recent filing with the SEC (United States Securities and Exchange Commission 2014: 31, emphasis original). We depend on a limited number of governmental customers for a significant portion of our revenues. We currently derive and expect to continue to derive, a significant portion of our revenues from a limited number of government agencies. The loss of, or a significant decrease in, business from the BOP [Federal Bureau of Prisons], ICE [Immigration and Customs Enforcement], and USMS [United States Marshall Services], or various state agencies could seriously harm our financial condition and results of operations. . . . We expect to continue to depend upon these federal agencies and a relatively small group of other governmental customers for a significant percentage of our revenue. These contracts effectively insulate these firms from accountability and sterilize any profit-loss incentives that would be present in a free market. Consider that, in a free market, no company would have its income guaranteed. Fluctuations in the demand for its products would be a normal part of doing business, something with which successful enterprises would have to contend. This is not the case for privately contracted prisons, however. One study, for example, found some 65 percent of prison contracts include what’s referred to as a “low-crime tax” or occupancy rate quotas. That is, government agencies agree to keep a certain number of prisoners in a given facility, or pay a penalty to the company (The Public Interest 2013: 3). Such clauses typically insure that 80–100 percent of beds are filled, thus protecting prison companies from fluctuations in the prison population (Ibid). Moreover, such contractual mechanisms disincentivizes these corporations from doing what would be required in a free market—discovering and implementing new ways to provide their service with changing market conditions. Even further, these contracts mean that contracted prisons will be paid regardless of their performance. In essence, a contracted prison may be wholly inept at providing its services, but the bureaucratic regime in which it operates means payment will still be made, regardless of the service provided. Indeed, these contracts have cost taxpayers millions of dollars. In 2013, for example, occupancy requirements in Colorado forced taxpayers

216  Abigail R. Hall and Veronica J. Mercier to hand over some $2 million to cover the empty cells in the state’s prisons (Hill 2013). Further evidence that contracted prisons’ success turns not on pleasing customers in the market but on their ability to work within the bureaucratic structures of government comes from data on lobbying. As noted, in a free market, no one or few firms would have special privileges in a market. Under a bureaucratic regime, however, lobbying becomes particularly important, as lawmakers have the ability to award contracts and favors to some groups over others. Private prisons have worked continuously to support those public officials and legal mandates that would help their businesses. Since 2010, for example, CCA and GEO have collectively spent nearly $9.7 million lobbying Congress with the goals of influencing the rules surrounding the Department of Homeland Security (DHS), the Federal Bureau of Prisons (FBP), and Immigration and Customs Enforcement (ICE), among others (Baccellieri 2016). In addition to lobbying, these firms have likewise poured thousands of dollars into campaign contributions, in an effort to sway various elected officials to enact laws that favor their businesses. To give but two examples, GEO has a history of providing campaign contributions to Marco Rubio and Jeb Bush. In 2016, Rubio received $40,000 in contributions directly from GEO. Further financial support was given indirectly though GEO’s donations to the Super PAC Right to Rise, providing Rubio and Bush with $2 million and $83 million, respectively, during the 2016 Presidential Campaign (Center for Responsive Politics 2016). Although Rubio has refused to provide voters with positions on key issues covered by the Political Courage test, his voting record aligns with the interests of private prisons such as GEO Group. Most recently, he voted in favor of Kate’s Law, which would increase the maximum prison term for undocumented immigrants (Vote Smart 2016). The issue of government utilizing private firms to run and maintain prisons remains a topic of controversy. While the DOJ announced its plans to discontinue the use of such facilities, private prisons derive most of their business from the Department of Homeland Security. As of late 2016, the DHS has not announced any intention to discontinue their contracts. Regardless of which government agency is responsible for providing the contract, the economic reasoning applied above remains the same. As opposed to acting as competitors in the marketplace, contracted prisons are “private” in name only. Operating in the same realm as any other bureaucratic agency, those who operate such facilities face no incentives to compete in the market, but instead work to secure their income from various government channels. As such, the problems associated with contracted prisons are likely to continue.

4  Asset Forfeiture The intersection of private interest and bureaucracy in the criminal justice system is once again put on display with the use of civil asset forfeiture

Policing for “Profit”  217 programs. Civil asset forfeiture is a “legal tool that allows law enforcement officials to seize property that they assert has been involved in certain criminal activity” (The Heritage Foundation 2016). Proponents of such programs contend these initiatives act as a deterrent and are a way for law enforcement to disrupt and dismantle organized crime by seizing the assets of criminals. Moreover, they argue the proceeds from such forfeitures may help relieve the budgetary pressures faced by many police departments, and that recovered and seized property may be used to compensate victims. Critics of such activities, however, contend that civil forfeiture is subject to rampant abuse, and violates even the most fundamental private property rights. Indeed, the use of civil asset forfeiture has expanded remarkably over time, with the Department of Justice’s Assets Forfeiture Fund taking in $93.7 million in revenue in 1986 and increasing to $4.5 billion in 2014—a 4,667 percent increase (Carpenter et al. 2015: 5). In the period from 2001 to 2014, the forfeiture funds of the DOJ and Treasury Department together took in nearly $29 billion, and combined annual revenue grew 1,000 percent (Ibid: 9). Seized assets include real estate, vehicles, commercial businesses, cash, financial instruments, jewelry, art, antiques, collectibles, vessels and aircraft, among other items (United States Marshals Service 2016). The dollar amount of seized assets alone is a source of controversy. What property is subjected to forfeiture under current laws is likewise a source of heated debate. Consider, for instance, that civil forfeiture does not require the owner of the property to be guilty of a crime. Civil forfeiture cases charge the property with a crime, meaning that homes, cars, cash, etc., may be seized without criminal charges ever being filed against the property owner! This has led to many high profile cases in which innocent individuals have had their property taken by authorities. Consider, for example, the case of Joseph Rivers. The 22-year-old had saved $16,000 and purchased a one-way train ticket from Michigan to Los Angeles with the intention of pursuing a music career. Government agents boarded the train in New Mexico, seizing his money, leaving him penniless and without a way to get to L.A. or back to Michigan. When asked about the case, Sean Waite, the agent in charge of the DEA in Albuquerque stated that, “We don’t have to prove that the person is guilty. It’s that the money is presumed to be guilty” (Friedersdorf 2015). In another case, a man driving from Virginia to Delaware was pulled over by an officer and “looked like a drug dealer” (Wolfe 2015). Despite having no drugs in the vehicle and finding no evidence of illegal activity, the officer seized the driver’s cash. He was not charged with any crime (Ibid). In another incident, Dale Agostini, a restaurant owner, was pulled over for driving in the left lane without passing. After using an untrained police dog to search the vehicle, police seized the $50,000 in cash Agostini was carrying to buy equipment for his business. The district attorney who arrived to the scene informed Agostini and his passengers that they would be charged with money laundering and that his one-year-old child, who was in the car, would be turned over to Child Protective Services. He was never charged with any crime (Ibid).

218  Abigail R. Hall and Veronica J. Mercier As with private prisons, private enterprise is often viewed as the culprit. Police and district attorneys, looking to pad their personal or department coffers, prey on often-innocent civilians. This is seen clearly in a variety of publications and public discourse, which refer to the practice as “policing for profit.” However, just as private prisons compete for “profit” only in the sense that they are seeking financial gain, a similar logic applies to asset forfeiture programs. Once again, it is not the problems of the “free market” generating the harms, but the intersection of private interests and the bureaucratic apparatuses of government. Just as the War on Drugs provides important context for understanding the marked expansion of private prisons, so too does US drug policy prove integral in understanding the rise of civil asset forfeiture. Although civil asset forfeiture has deep roots in common law, it was not until the Comprehensive Drug Abuse Prevention and Control Act of 1970 that government agents were allowed to seize drugs and drug paraphernalia. Section 881 of the Act allows for civil asset forfeiture explicitly (Comprehensive Drug Abuse Prevention and Control Act of 1970). In 1978 and 1984, Congress expanded civil asset forfeiture to include money, real estate, and other proceeds linked directly to drug operations (Psychotropic Substances Act 1978 and the Comprehensive Crime Control Act of 1984). The aforementioned Comprehensive Crime Control Act of 1984 allowed federal law enforcement agencies to retain the proceeds from forfeited assets and place them in the Department of Justice Assets Forfeiture Fund, for use by law enforcement. Moreover, Congress implemented an “equitable sharing” program that authorizes state and local law enforcement to turn over seized assets to the Justice Department and receive up to 80 percent of the assets’ value in return. Any state or local law enforcement agency that participates in an investigation that results in federal forfeiture may request to participate in the equitable sharing program, making a multitude of state and local agencies eligible to receive forfeiture funds (Weber 2009: 3). Other laws have continued to expand the use of civil forfeiture. In fact, some 400 distinct federal statutes allow for the use of forfeiture and nearly every state allows for the seizure of personal property (Malcolm 2015). Taken together, these laws induced police to engage in more anti-drug policing by providing a financial incentive for state and local law enforcement to enforce federal drug laws (Baicker and Jacobson 2004: 1). The incentives created by such a program are straightforward. Recall that police, originally tasked with upholding domestic laws, were placed squarely on the “front lines” of the federal drug war. Overnight, the scale and scope of police operations expanded immensely. Using the logic of the economics of bureaucracy, we can understand easily how this shift in policing operations, combined with changing forfeiture laws, provided a clear avenue for police to expand their reach and operations. Through active participation in the federally driven War on Drugs and the seizure of assets, police departments found an easy way to increase their discretionary budgets and

Policing for “Profit”  219 number of subordinate personnel. Through asset forfeiture, many police departments could altogether neglect traditional means of increasing their budgets—having relevant government officials allocate more resources to policing. Instead, departments could make up shortfalls, purchase desired equipment, and in some cases pay wages, through seized funds. This has a pernicious, self-perpetuating effect. As police departments become more dependent on forfeiture funds, the incentive to seize more assets increases. In turn, this leads to police actively seeking additional assets to seize. Given that an individual does not have to be guilty of a crime in order for his property to be seized, police can seize assets with little restriction. This idea is captured clearly in the words of a variety of individuals linked to forfeiture programs. In a warning sent out to all US attorneys, for example, one former Attorney General said that, “We must significantly increase forfeiture production to reach our budget target. . . . Every effort must be made to increase forfeiture income” (Benavie 2016: 7). The windfall experienced by some police departments, and the lack of constraints they face, is expressed by Columbia, Missouri, Police Chief Kenneth M. Burton. It’s usually based on a need—well, I take that back. There’s some limitations on it. . . . Actually, there’s not really on the forfeiture stuff. We just usually base it on something that would be nice to have that we can’t get in the budget, for instance. We try not to use it for things that we need to depend on because we need to have those purchased. It’s kind of like pennies from heaven—it gets you a toy or something that you need is the way that we typically look at it to be perfectly honest. (quoted in Carpenter et al. 2015: 19) Funds from seized assets have indeed increased and police spending has increased with it. In 1986, just two years after the adoption of the Comprehensive Crime Control Act, the DOJ’s Assets Forfeiture Fund took in some $93.7 million in revenue. In 2014, the same program took in some $4.5 billion—an increase of 4,667 percent (Carpenter et al: 5). Since 2008, police agencies have spent some $2.5 billion in proceeds from cash and property seized under forfeiture operations. Perhaps the clearest illustration of these mechanisms at work is the asset forfeiture programs in place in Philadelphia, Pennsylvania. Pennsylvania state law enables agencies to retain 100 percent of the value of forfeited property, giving police departments and district attorneys major financial incentives to seize as much property as possible. In almost all cases, individuals who have had their property seized fail to have it returned. Of the more than 8,000 asset forfeiture cases filed against cash in 2010, 96 percent of decisions favored the District Attorney’s Office (Sibilla 2014). Between 2002 and 2012, the city seized some 1,172 real estate properties (Ibid). Over that same period, the city took in $64 million in forfeiture funds, nearly $6 million per year (Ibid). Of this $64 million,

220  Abigail R. Hall and Veronica J. Mercier some $25 million was used to pay the salaries of law enforcement. This includes the salaries of prosecutors (Ibid)! Persons who have their property seized in Philadelphia and elsewhere may have a difficult time getting their assets returned to them. The “burden of proof” required to seize assets is remarkably weak. In criminal trials, government must show that a person is guilty beyond a reasonable doubt. In civil forfeiture cases, however, government must only show “probable cause” in order to seize property. The burden of proof then falls on the owner to produce evidence that his or her property has not been used in the commission of a crime, or that he or she took steps to try to prevent the crime (Hyde 1995: 7). In criminal trials, defendants are entitled to legal representation. No such requirement exists for civil forfeiture cases. If a person has their property seized, they must hire their own lawyer or fight the forfeiture on their own. Taken together, the quantity of assets seized, combined with the controversy regarding what assets can be seized has generated calls for reform. While some states have passed laws making it more difficult for law enforcement to seize assets or restricts how forfeited property may be used, the vast majority of states still allow for the broad use of such initiatives (Snead 2016). Moreover, the federal equitable sharing program allows many law enforcement agencies to bypass state and local protocols all together.

5 Conclusion Criminal justice reform has fast become a national issue, with civil asset forfeiture and private prisons representing key issues. Former Presidential candidate Hillary Clinton, for example, stated, “We need to end private prisons. Protecting public safety . . . should never be outsourced or left to unaccountable corporations” (quoted in la Monica 2016). Discussing civil asset forfeiture, Kentucky Senator and former Presidential candidate Rand Paul wrote that, “In our country, you should be presumed innocent until proven found guilty. Civil asset forfeiture doesn’t allow for the presumption of innocence. The government can take your stuff without being convicted of a crime” (quoted in Pye 2015). Without a doubt, problems at all levels of the CJS are beginning to be recognized. In addition to the problems associated with contracted prisons and civil asset forfeiture, issues like the militarization of domestic police have been brought to light by events in Ferguson, Missouri, Baltimore, Maryland, and New York City. The use of equipment once reserved exclusively for the military has many individuals questioning the trajectory of current police operations, as well as the relationship between police and those they are intended to protect and serve (Szoldra 2014). Advocates argue that changes to these and other areas are necessary in order to protect individual freedoms and ensure that the criminal justice system is impartial and fair to all citizens.

Policing for “Profit”  221 In proposing these changes, however, the villain is often portrayed as the “capitalistic” motives of private prisons, police departments, and others. As we have discussed, such sentiments miss the mark in correctly identifying the root cause of the problem. The issue is not profit in and of itself. As economists have long demonstrated, the profit motive, when combined with the checks and balances of free enterprise, leads to increased wealth and individuals serving their fellow man in order to serve themselves. The problems observed within the CJS, particularly as they relate to private prisons and civil forfeiture, are the products of perverse incentives created by the bureaucratic structure of all facets of the CJS. Just as the economic way of thinking can shed light on why these problems occur, it can likewise allow us to see why the road to reform is difficult. In order to reform civil asset forfeiture, for example, legislatures at the state and national levels would have to be willing to pass laws that would limit or prohibit law enforcement from benefitting from seized assets. It is not difficult to understand why police departments would fight against such legislation. They face strong incentives to keep policies in place that protect and expand their budgets. In a similar way, other special interest groups like police and prison guard unions and firms operating contracted prisons face strong incentives to keep policies in place that help their bottom lines. This idea is captured clearly in the Corrections Corporation of America’s (CCA) 2011 10-K report filed with the Securities and Exchange Commission (SEC). The demand for our facilities and services could be adversely affected by the relaxation of enforcement efforts, leniency in conviction or parole standards and sentencing practices or through the decriminalization of certain activities that are currently proscribed by our criminal laws. For instance, any changes with respect to drugs and controlled substances or illegal immigration could affect the number of persons arrested, convicted, and sentenced, thereby reducing demand for correctional facilities to house them. . . . Legislation has been proposed in numerous jurisdictions that could lower minimum sentences for some nonviolent crimes and make more inmates eligible for early release based on good behavior. (quoted in Cohen 2015) Without a doubt, the tide seems to be turning with regard to the need to reform the CJS. True changes, however, require first and foremost an understanding of the underlying incentives faced by the actors working within the system. Only then can lawmakers, scholars, and the public truly discuss and articulate possible solutions.

References American Civil Liberties Union (2011). “Banking on Bondage: Private Prisons and Mass Incarceration.” Available at www.aclu.org/files/assets/bankingonbondage_ 20111102.pdf

222  Abigail R. Hall and Veronica J. Mercier Baccellieri, Emma (2016). “Spotlight on Private Prisons.” Center for Responsive Politics, August 23. Available at www.opensecrets.org/news/2016/08/spotlighton-private-prisons-lobbying-and-otherwise/ Baicker, Katherine and Jacobson, Mireille (2004). “Finders Keepers: Forfeiture Laws, Policing Incentives, and Local Budgets.” NBER Working Paper 10484. Available at www.nber.org/papers/w10484.pdf Benavie, Arthur (2016). How the Drug War Ruins American Lives. Santa Barbara, CA: Praeger. Boettke, Peter J., Coyne, Christopher J. and Hall, Abigail R. (2013). “Keep Off the Grass: The Economics of Prohibition and U.S. Drug Policy.” Oregon Law Review 17: 1069–95. Bureau of Justice Assistance (2001). Emerging Issues of Privatized Prisons. Washington, DC: U.S. Department of Justice. Bureau of Justice Statistics (1982). “Prisoners 1925–1981.” Available at www.bjs. gov/content/pub/pdf/p2581.pdf Bureau of Justice Statistics (2016). “Correctional Populations in the United States.” Available at www.bjs.gov/content/pub/pdf/cpus14.pdf Carpenter, Dick M., Knepper, Lisa, Erickson, Angela C. and McDonald, Jennifer (2015). “Policing for Profit: The Abuse of Civil Asset Forfeiture, 2nd ed.” Institute for Justice. Available at https://ij.org/wp-content/uploads/2015/11/ policing-for-profit-2nd-edition.pdf Carson, E. Ann (2015). “Prisoners in 2014.” Bureau of Justice Statistics. Available at www.bjs.gov/content/pub/pdf/p14.pdf Center for Responsive Politics (2016). “Marco Rubio.” Available at www.opensecrets.org/politicians/contrib.php?cid=N00030612&cycle=2016 Cohen, Michael (2015). “How For-Profit Prisons Have Become the Biggest Lobby No One Is Talking About.” The Washington Post, April 28. Available at www. washingtonpost.com/posteverything/wp/2015/04/28/how-for-profit-prisonshave-become-the-biggest-lobby-no-one-is-talking-about/?utm_term=. ae8c95df3ebb Comprehensive Crime Control Act of 1984 (1984). Pub. L. No. 98-473, 98 Stat 2050 codified at 21 U.S.C. § 881(a)(7). Denvir, Daniel (2016). “Private Prisons Are Not the Problem: Why Mass Incarceration Is the Real Issue.” Salon, August 24. Available at www.salon.com/2016/08/24/ private-prisons-are-not-the-problem-why-mass-incarceration-is-the-real-issue/ Dyer, Joel (2000). The Perpetual Prisoner Machine: How America Profits from Crime. Boulder, CO: Westview Press. Friedersdorf, Conor (2015). “The Injustice of Civil-Asset Forfeiture.” The Atlantic, May 12. Available at www.theatlantic.com/politics/archive/2015/05/the-glaringinjustice-of-civil-asset-forfeiture/392999/ Glenmary Home Missioners (2016). “Wardens from Wall Street: Prison Privatization.” Available at www.glenmary.org/site/epage/109420_919.htm Hall, Abigail R. and Coyne, Christopher J. (2013). “The Creeping Militarization of U.S. Domestic Policing.” The Independent Review 17: 485–504. Harrison, Lana D., Backenheimer, Michael and Inciardi, James A. (1996). “Cannabis Use in the United States: Implications for Policy.” In Cannabisbeleid in Duitsland, Frankrijk en de Verenigde Staten, ed. Perter Cohen and Arjan Sas. Amsterdam: Centrum voor Drugsonderzoek, Universiteit van Amsterdam, 237–47. Available at www.cedro-uva.org/lib/harrison.cannabis.05.html

Policing for “Profit”  223 Heritage Foundation (2016). “Civil Asset Forfeiture: 7 Things You Should Know.” Available at www.heritage.org/research/reports/2014/03/civil-asset-forfeiture-7things-you-should-know Hill, Chris (2013). “Colorado Is Giving $2 Million to a Private Prison Company for Nonexistent Inmates.” Policy Mic, September 25. Available at https://mic. com/articles/65223/colorado-is-giving-2-million-to-a-private-prison-companyfor-nonexistent-inmates#.JHrC0o2HK Hyde, Henry J. (1995). Forfeiting Our Property Rights: Is Your Property Safe from Seizure? Washington, DC: CATO Institute. la Monica (2016). “Clinton Call for End of Private Prisons Sinks Jail Stocks.” CNN Money, September 27. Available at http://money.cnn.com/2016/09/27/investing/ prison-stocks-hillary-clinton-debate-corrections-corporation-america-geo-group/ Malcolm, John G. (2015). “Civil Asset Forfeiture: When Good Intentions Go Awry.” Testimony Before the Pennsylvania State Senate Judiciary Committee. Available at www.heritage.org/research/testimony/civil-asset-forfeiture-when-good-intentionsgo-awry Mason, Cody (2012). “Too Good to Be True: Private Prisons in America.” The Sentencing Project. Available at http://sentencingproject.org/wp-content/uploads/ 2016/01/Too-Good-to-be-True-Private-Prisons-in-America.pdf Migué, Jean-Luc and Bélanger, Gérard (1974). “Toward a Theory of Managerial Discretion.” Public Choice 28: 24–8. Niskanen, William N. (1971). Bureaucracy and Representative Government. Chicago and New York City: Aldine, Atherton. Niskanen, William N. (1975). “Bureaucrats and Politicians.” Journal of Law and Economics 18: 617–43. Nixon, Richard (1969). “Special Message to the Congress on Control of Narcotics and Dangerous Drugs.” The American Presidency Project, July 14. Available at www.presidency.ucsb.edu/ws/?pid=2126. Presbyterian Church (2003). “Resolution Calling for the Abolition of Private Prisons.” Available at www.presbyterianmission.org/wp-content/uploads/7-abolitionof-for-profit-prisons-2003.pdf Psychotropic Substances Act (1978). Pub. L. No. 95-633, Title III, 301(a), 92 Stat. 3777 codified at 21 U.S.C. § 881(a)(6). The Public Interest (2013). “How Lockup Quotas and ‘Low-Crime Taxes’ Guarantee Profits for Private Prison Corporations.” Available at www.njjn.org/uploads/ digital-library/Criminal-Lockup-Quota,-In-the-Public-Interest,-9.13.pdf Pye, Jason (2015). “Rand Paul to Barack Obama: Do You Support Civil Asset Forfeiture and Loretta Lynch’s Unconstitutional Theft?” FreedomWorks, April 20. Available at www.freedomworks.org/content/rand-paul-barack-obama-do-yousupport-civil-asset-forfeiture-and-loretta-lynchs Redford, Audrey and Powell, Benjamin (2016). “The Dynamics of Intervention in the War on Drugs: The Build-Up to the Harrison Act of 1914.” The Independent Review 20: 509–30. Sibilla, Nick (2014). “Philadelphia Earns Millions by Seizing Cash and Homes from People Never Charged with a Crime.” Forbes, August 26. Available at www.forbes.com/ sites/instituteforjustice/2014/08/26/philadelphia-civil-forfeiture-class-actionlawsuit/#117fe6435dda. Smith, Adam (1981). An Inquiry into the Nature and Causes of the Wealth of Nations. Indianapolis, IN: Liberty Fund.

224  Abigail R. Hall and Veronica J. Mercier Snead, Jason (2016). “An Overview of Recent State-Level Forfeiture Reforms.” The Heritage Foundation. Available at www.heritage.org/research/reports/2016/08/ an-overview-of-recent-state-level-forfeiture-reforms Szoldra, Paul (2014). “This Is the Terrifying Result of the Militarization of Police.” Business Insider, August 12. Available at www.businessinsider.com/police-militari zation-ferguson-2014-8 United States Marshals Service (2016). “Asset Forfeiture Program.” Available at www.usmarshals.gov/assets/index.html United States Securities and Exchange Commission (2014). “Form 10-K: Corrections Corporation of America.” Available at www.sec.gov/Archives/edgar/ data/1070985/000119312515061839/d853180d10k.htm U.S. Department of Justice Office of the Attorney General (2016). Available at www. justice.gov/opa/file/886311/download. Vote Smart (2016). “Marco Rubio’s Voting Records.” Available at http://votesmart. org/candidate/key-votes/1601/marco-rubio#.WDZkYyMrJcw Weber, Richard (2009). “Guide to Equitable Sharing for State and Local Law Enforcement Agencies.” Department of Justice. Available at www.justice.gov/ criminal-afmls/file/794696/download Wolfe, J. Francis (2015). “10 Egregious Abuses of Civil Asset Forfeiture.” Listverse, June 29. Available at http://listverse.com/2015/06/29/10-egregious-abuses-of-civilasset-forfeiture/ Yates, Sally (2016). “Memorandum for the Acting Director Federal Bureau of Prisons.”

13 Why Paternalists and Social Welfarists Should Oppose Criminal Drug Laws Andrew Cohen and William Glod1

Introduction The principle of legal paternalism allows that it is a good reason for a law that the adult person affected by the law is herself benefited or at least prevented from harming herself, often despite her occurrent desires and choices. Laws prohibiting the sale, purchase, or use of drugs, and laws requiring the usage of seat belts in automobiles or helmets on bikes, are examples of paternalist laws. One of us has written elsewhere about why legal paternalism toward competent adults disrespectfully substitutes unjustified judgments for the target agent’s own judgment, regardless of any success it might have in motivating targets not to harm themselves (Glod 2013). He has also argued that paternalists face daunting informational constraints regardless of whether we should be concerned about autonomy as merely a capacity for self-direction (Glod 2015). The other of us has written against paternalism and in favor of harm to others being the only justification for laws (Cohen 2014). Here, we argue that legal paternalism involving incarceration for nonviolent drug crimes will largely fail to achieve its own objectives. We don’t consider—what also worries us—the unfortunate limits paternalism puts on autonomy. Along the way, we also treat non-paternalistic “social welfarist” arguments that justify incarceration in terms of preventing harm to others. We focus on drug addiction, as its criminal prosecution is a notable source of over-incarceration in the United States.2 Our thesis is simple: we ought to have no criminal laws justified by paternalism; ridding our society of such laws would reduce the problem of over-incarceration, and paternalists and social welfarists should support their repeal. Some critics of drug criminalization are skeptical that paternalism or protection of society against allegedly dangerous drug addicts is the main motive behind drug laws. Consider “broken windows policing,” in which police arrest people who commit relatively minor crimes like vandalism and drug possession because that group of people overlaps with the group of criminals who do significant harm to others.3 Incarcerating them for minor crimes, it is claimed, has the effect of reducing the serious harm some of them would otherwise do. Others maintain that rent-seeking explains the

226  Andrew Cohen and William Glod rise and prevalence of criminal laws. Be this as it may, our focus is normative and critical rather than descriptive—we focus on justifications for treating drug use as a criminal matter. The above explanations may be correct; nonetheless, concerns about drug addiction’s harmfulness remain a key justification for the criminalization of drugs in the minds of many people beyond law enforcement, and this is an audience we are keen to reach. As such, we look at the real-life effects that incarceration due to criminal drug laws tends to generate. It would be odd for defenders of such laws to claim that their main concern is with the harm of drug addiction and not consider the ancillary negative effects of those laws. We proceed as follows. In the first section we explain paternalism and social welfarism. In the second section we consider costs and benefits of criminal drug laws to those who violate them. We continue discussing costs and benefits, but to loved ones and others closely connected with violators and with society at large, in sections 3 and 4. In section 5 we consider some objections and in section 6 we discuss alternatives to incarceration.

1.  Paternalism and Social Welfarism Explained Joel Feinberg writes that, according to the principle of legal paternalism, “It is always a good reason in support of a [proposed] prohibition that it is probably necessary to prevent harm (physical, psychological, or economic) to the actor himself” (Feinberg, 26–27). He distinguishes this from what he calls moralistic legal paternalism, according to which “It is always a good reason in support of a proposed prohibition that it is probably necessary to prevent moral harm (as opposed to physical, psychological, or economic) to the actor himself,” as well as a type of perfectionism, according to which “It is always a good reason in support of a proposed prohibition that it is probably necessary for the improvement (elevation, perfection) of the character . . . of the very person whose liberty is limited” (ibid, 27).4 In this chapter, we speak of paternalism in general terms as incorporating the three principles just discussed. We take justifications for laws to be paternalist if they claim that the law in question will tend to prevent physical, psychological, economic, or moral harm to the individuals such laws permit interference with or if such laws will tend to improve those individuals physically, psychologically, economically, or morally. This is an intentionally capacious understanding of paternalism, but one we think reasonable. The most straightforward example of paternalist laws in the US today comprise the so-called “war on drugs.” In 1997, fully 26.9% of those incarcerated by state and federal governments were imprisoned for violating a drug law. The number has fortunately decreased—to 20.5% by 2012, but still leaves 309,100 individuals whose lives—and those of their family, friends, and business associates—were worsened because of laws supposedly meant to help them.5 Our basic point is simple and, in our view, oddly overlooked: bringing people into the corrections system will set back their

Paternalists and Social Welfarists  227 interests, so doing so in order to prevent them from setting back their own interests via drug use is problematic. This is why the capacious understanding of paternalism we work with makes our case more difficult: while we can point out that arresting people sets back their interests, someone might claim that we are too narrowly considering only the things they actually recognize as in their interest and they may too easily ignore the desiderata of making would-be criminals more virtuous and better decision makers. If that claim were right, there might be good reason to continue to maintain paternalist criminal laws. We aim to show, though, that wanting to improve the character and decision-making skills of would-be criminals should lead us to reject those laws. While our primary concern is with paternalism, throughout our discussion we hope to show that social welfarists also ought to reject incarceration of those violating paternalistic laws. In brief, when we discuss social welfarists, we have in mind those who favor laws and legal actions that are aimed to promote the well-being of all individuals in society. (They may or may not also think society itself can be made better or worse.)

2. Costs and Benefits to Those Violating Laws Justified by Paternalism In one of our home states—Georgia—almost 6% of the population is under some form of correctional control.6 Nationally, more than one out of every fifty people is under some form of correctional control. As a fifth of our jail and prison populations are incarcerated for drug related offenses, it is reasonable to assume that a fifth of those under correctional control are in that situation because of drug related offenses. While being in prison is, we assume, worse than parole or probation, being under any form of correctional control is being subject to great limits on one’s freedom. For the sake of simplicity, we concentrate on incarceration. Let us explore some of the costs to those who face incarceration. First, those who encounter the possibility of being incarcerated typically face mounting legal and other fees to fight imprisonment. These fees are often enough to wipe out any savings they built up or to put them into debt. Public defenders are often overworked and ineffective at getting their clients absolved or getting them more lenient sentences. More effective attorneys are often expensive, sometimes prohibitively so. The poorest and worst-off are not well equipped to handle these fees or post bail while detained. Those who were relatively successful before incarceration face not only fees but also lost income once they lose their jobs. We should not overlook the fact that some of these costs are incurred by individuals who never face sentencing because found not guilty. Their costs are not typically reimbursed. While we are focusing on the costs to those incarcerated for violating paternalistic laws and the creep of those costs to those connected to them, the costs of those laws are clearly more widespread. Others have noted, of course, the

228  Andrew Cohen and William Glod simple costs of enforcement. We should also not be blind to the clear financial and emotional costs (etc.) to those exonerated. Second, the stress of prosecution and confinement can take a physical and psychological toll on prisoners, many of whom already suffer from mental health issues. Even hours or days in a local jail may suffice to lead a fearful or drug-withdrawn inmate to take rash action, including suicide.7 Recent studies indicate that separation from loved ones and loneliness are worse for one’s health than smoking or obesity, as intimacy is not a luxury but a vital human need (Cacioppo and Caciopp. 2014; Metzner and Fellner 2010).8 In many cases, visitation is highly restricted and subject to change at the whim of prison officials. People who have traveled a long way to visit might face last-minute cancellations due to (say) a lockdown. Parties may be separated by thick glass while visiting or only allowed to communicate through computer equipment,9 and even when they are not, affectionate exchanges are typically forbidden. Conjugal visits, even for married prisoners, are often not allowed for sentences under certain durations. Pets are typically not allowed at all, so someone imprisoned long enough may never get to see a beloved animal again. Prisoners often cannot be furloughed for funerals, so are unable to pay final respects to a friend or family member. Third, the purely physical hell of prisons can include abuse or humiliation from fellow inmates and prison guards; the constant fear (reasonable!) for one’s safety; low quality, unappetizing, and unhealthy food; uncomfortable bed and clothing; boredom; crowded and unsanitary conditions; and inadequate access to mental health resources, education, or vocational counseling. Joining a violent and racist gang is often one’s best means of protection from abuse if one has a lengthy sentence, and one is often beholden to the gang once released since failure to abide by its directives can leave one and one’s loved ones at risk of severe punishment.10 Fourth, confinement brings large opportunity costs with little chance to be productive or earn money. A typical day in prison often involves sitting in one’s cell watching TV, if one is lucky enough to have one. Literate prisoners are allowed to read only approved books and it is unclear how much effort is made to teach illiterate prisoners how to read since education is not a priority at most prisons.11 Fifth, life often fares no better for the prisoner after release, since a loss of social status and opportunities post-incarceration remains. It is often difficult to find a good job—or any job—with a criminal record. A person who committed nonviolent albeit morally objectionable crimes in exchange for drugs, such as distributing but not producing or viewing child porn, is a registered sex offender for life, seen as on a par with actual child rapists, and faces tremendous restrictions on where to live and work. Even ex-inmates who had been imprisoned for less serious crimes are treated as second-class citizens merely because they are seen as little more than untrustworthy felons. A diminution of the social bases of self-respect in light of the above is a

Paternalists and Social Welfarists  229 powder keg for recidivism, something that should concern paternalists who favor incarceration. Given all of these costs to the incarcerated, there must be some associated benefits—or there is no rationale for criminal drug laws. We take it that the benefit of drug laws to at least one sort of perpetrator—the individual who uses the illegal drug—is to help them not to use the drug.12 Assuming the drug use is bad for the user, the benefit of imprisonment should be clear: making the drug unavailable so that one has better opportunity to improve one’s health or decision-making abilities, or become morally better. It may also help foster stronger social bonds. Additionally, there are indirectly paternalistic effects when would-be users are deterred from use by the threat of prison. Finally, criminal drug laws are justified in terms of protecting society from the dangers of drug use, a “harm to others” rationale. If these benefits outweigh the costs of drug laws, that would be a strong if not decisive reason to maintain those laws. We submit, though, that the above-discussed costs—not to mention those discussed in the next section—far outweigh any benefits. Our hypothesis is that for those already motivated to curb their use of drugs, incarceration is unnecessary and that for those not so motivated, incarceration will not help since inmates can often get contraband in prison or return to their habits upon release. Incarceration is ineffective (at best) for the insufficiently motivated and unnecessary (at best) for the sufficiently motivated. That is, even if we could somehow avoid the costs discussed above, people who do not want to quit using the illegal drugs they were arrested for using will not tend to benefit from being forced to quit for the duration of their imprisonment, and those who are willing to quit do not need to be imprisoned in the first place. Consider those who are unmotivated to give up the self-harmful behavior. Perhaps addiction is always lamentable, but if defenders of incarceration are arguing that people have it within themselves to learn from incarceration, it’s unclear why they would not also believe people have it within themselves to learn from life’s challenges without the additional costs of incarceration. Paternalists can’t have it both ways. People can’t be so lost to addiction that they lack freedom or opportunities to improve, yet magically gain freedom and opportunities to improve through incarceration and its attendant costs. There are probably some people who will claim they were grateful for imprisonment because it forced them to take inventory and get their act together. However, they had to have some motivation to continue to stay clean, so imprisonment was not a necessary motivation to steer them in the “right” direction. Perhaps it happened to be the thing that spurred them to change at a given moment, but if so, that is clearly contingent—and there are likely more effective (and respectful) tactics worth exploring, such as those discussed in section 6.13 Of course, critics of our position can agree with us that the current corrections system is a massive failure at actually correcting behavior but insist that this does not mean we should do away with it or with criminal

230  Andrew Cohen and William Glod laws—including paternalist laws—that send people into the system. They can insist that we should change the system—which is really a penal system more than anything else—into an appropriately corrective rehabilitative system based on good data about what can change incentives. They could conclude, then, that we should not blame paternalistic laws for bad penal practices in the United States. We welcome their proposal of a better alternative. Our point is that in our current system, people found guilty of crimes suffer more from their punishment—especially incarceration—than is usually recognized and that the damage from punishing people for violating paternalist laws tends to outweigh the supposed damage they do to themselves. The same holds for harm to others. We suggest that whatever harm others suffer due to an addict’s habit will be outweighed by the harms they suffer by imprisoning the addict. We tend to think that should be the end of it, but if proponents of reformed carceral practices have a developed proposal to improve the US system so that it genuinely helps people without causing harm to others, we can discuss again whether we should punish people for violating paternalistic laws. We don’t think we should, but it is a topic for a later day—a day, we admit, we don’t think will ever come. Let us now look at whether criminal drug laws prevent harm to others on net.

3. The Initial Drift of Costs and Benefits: Family, Friends, and Neighbors It is clear to us that the costs to those subject to correctional control far outweigh the benefits. In this section, we suggest that the costs go far beyond those. The costs to family, friends, and business associates of those incarcerated for violating criminal drug laws are also tremendous. Richard Lippke (2016) nicely lays out many of the costs of incarceration that drift to loved ones of prisoners:14

• • • • • • •

lost intimacy lost emotional support lost income (and other financial benefits) lost assistance in raising children or caring for other loved ones incurred shame increased worry about the imprisoned loved one the burden of life with “offenders who are psychologically damaged by their imprisonment” when they are released (Lippke, 9)

To this list, we would add:

• financial costs of visiting loved one, often far away • financial costs of communication (prison phone calls are often very expensive)

• loss of role models

Paternalists and Social Welfarists  231

• incurred stress and concern about the imprisoned • incurred view, of child, that parent is a bad person • incurred view, of child, that they are themselves bad as child of a bad person

• decreased ability to secure a loan And these costs, to others in the neighborhood of the incarcerated:

• loss of employee or employer • loss of trusted neighbor And we ought not forget costs discussed in section 2:

• the financial costs incurred by the imprisoned, whom these people help • the opportunity costs of the imprisoned, which likely impacts loved ones. We agree that these “collateral consequences” of legal punishment come “a bit too close to punishment of the innocent to be casually tolerated” (ibid, 2). Our point, though, is not dependent on that claim. Our point is that these costs offset any potential gain. The point of drug offense incarceration is meant to be a net benefit but, as argued in the last section, that is unlikely to be the case. These added costs should further incline paternalists and social welfarists against incarceration. It might be objected that the imprisoned individual is unlikely to have been a good parent, spouse, friend, or employee, and may have even been abusive or otherwise brought more difficulties to family and friends with their presence. Certainly, spousal and child abusers do frequently have substance abuse problems that might leave them imprisoned or otherwise punished by law. “Fifty percent of batterers are believed to have had ‘addiction’ problems (Faller 1988)”; “A recent survey of public child welfare agencies conducted by the National Committee to Prevent Child Abuse found that as many as 80 percent of child abuse cases are associated with the use of alcohol and other drugs (McCurdy and Daro 1994)” (Substance Abuse and Mental Health Services Administration 1997: 3). It thus might be thought that maintaining incarceration as a form of punishment for violating paternalistic drug laws is well worthwhile. There are at least three problems with this view. First, while we are certainly in favor of policies to reduce child and spousal abuse, maintaining paternalistic laws against drug use is unlikely to help in doing so. It is clear that alcohol is more of a problem than currently illegal drugs. “Probably the largest contributing factor to domestic violence is alcohol. All major theorists point to the excessive use of alcohol as a key element in the dynamics of wife beating. However, it is not clear whether a man is violent because he is drunk or whether he drinks to reduce his inhibitions against his violent behavior” (Labell 1979: 264; cited in Substance

232  Andrew Cohen and William Glod Abuse and Mental Health Services Administration 1997: 4). Hence, little is to be gained by current drug prohibitions. Importantly, though, “[r]esearch also indicates that women who abuse alcohol and other drugs are more likely to become victims of domestic violence (Miller et al. 1989)” (ibid, 3). If this gives support to paternalistic legislation against drugs—and alcohol— as designed to help the women who would presumably be abused if they drink or do drugs, it also amounts to blaming the victim. It seeks to prevent women from being abused by stopping them from drinking alcohol and using drugs. Compare: we can prevent women from being raped by stopping them from wearing short skirts and blouses with revealing bust lines. This line is worrisome. It is not the woman who is to blame if she is raped and it is not the woman who is to blame if she suffers spousal abuse. The rapist and abuser are to blame. Criminal laws against rape and abuse are clearly appropriate. There is one final issue with the idea that the correlation between alcohol and drug use on the one hand and domestic violence on the other should be addressed with paternalistic laws against the former. Simply put, while “a substantial proportion of people who abuse their partners also abuse substances,” “[m]ost people who abuse substances don’t abuse their partners.”15 Hence, we can expect a criminal law preventing substance use to harm a great many people who would not have abused anyone. Our last claim is hardly novel. Any law can affect people that would have done no harm whether or not the law was instituted. Still, criminal drug laws are especially prone to this problem. Here, though, our point is that these laws have tremendous negative impact on many people, not just those directly cited for violating the laws. The effects so clearly drift to the family and friends of those punished for violating the laws that it seems unimaginable to think justice requires such penal law. It is also worth noting the racially disparate effects of drug laws. Though most people that violate drug laws are white, most people who are imprisoned for such violations are not (Chin 2002). Wealthy or middle class white neighborhoods have drug users though those users are seldom arrested.

4.  Costs and Benefits to Society as a Whole In section 3, we discussed how punishment of those that violate laws tends to drift on to their loved ones and neighbors. Of course, there are also broader societal costs involved that should be noted. To begin, there is the obvious cost of enforcement. Domestic spending alone on enforcement of drug laws in the US has been greater than nine billion dollars annually since 2014 (Sacco 2016).16 Interdiction and international costs add almost five and a half billion annually.17 Ending the so-called “war on drugs,” in other words, would provide an immediate $14.5 billion savings per year. That savings could translate into lower tax bills for the average taxpayer or additional dollars for rebuilding infrastructure or some other useful purpose. Everyone

Paternalists and Social Welfarists  233 should welcome that, particularly paternalists and social welfarists motivated by beneficence. As we already discussed, those incarcerated for violating paternalistic laws lose their employment when in prison and likely have a more difficult time getting hired when released. That means they have fewer opportunities to bring their productive abilities to bear on the economy. They may thus suffer significant financial difficulty or receive state aid. They are prevented from increasing their income (at least in legal ways) and so prevented from buying goods and services from their neighbors, whose businesses thus do less well than they might. These opportunity costs are significant, though we can’t provide good estimates.18 The enforcement of drug laws, notoriously, has clear and significant unintended consequences. Those who sell illegal drugs are driven “underground” and a black market emerges. Disputants in black market exchanges cannot seek the aid of police or a state court system and are thus more likely to carry weapons than they might if the market was legal. Violence is thus higher because of these laws. This also means the costs of drugs will be higher as availability is limited and traders must expend more effort to import and/or transport their goods. There are also interesting and unclear consequences for the quality of the drugs. On the one hand, some will point out that the drugs sold might be less pure or tainted as unscrupulous criminal types are involved. On the other hand, there is clear incentive to make the drugs purer and more efficient at delivering the desired effect. Someone risking legal prosecution for importing, transporting, buying, or using a drug, wants the drug to be as easy to carry in a concealed manner as possible and as potent as can be (Thornton 1998). If one can get the same reward using one gram of a drug as they can using 10 grams, there is a clear advantage. To take one example, while “the estimated adjusted prices [of powder cocaine] experienced a modest but continuously gradual decline . . . estimated purities increased to a much higher degree—increasing from the 55 to 65 percent purity level in 2001 to the 70 to 80 percent region in 2006” (Fried et al. 2008: 9). Societally, there seems less concern about the possibility of more potent drugs then the possibility of more tainted drugs,19 but those with a paternalistic desire to end drug use should be very concerned with increased purity and these other costs. It would be a mistake to think the problems discussed in this section concern only those involved in the drug trade. There are always spillovers when one encourages policies that incentivize the use of guns and violence. Innocent bystanders can be hurt. There are also clear beneficiaries of such laws: for example, police departments get more funding to combat drug use and police officers might get better salaries than they otherwise would. With increased prohibitions and the rise of a black market, there are also increased possibilities for officers to work with those in the black market for remuneration “under the table.” This does not help most police officers, who face increased risks in the more violent world created by prohibition. It helps those in the

234  Andrew Cohen and William Glod black market who bring in substantial personal incomes (untaxed), but not those with genuine medicinal uses for a drug (or who simply want to responsibly use it recreationally). In short, when we have policies that encourage black markets, we likely help the careers of gang leaders, opportunistic police officers, corrections officers, and district attorneys and other prosecutors who have more work on their hands. None of this helps the rest of us. The rest of this should be predictable: we get a rise in gangs or other organized crime groups, they dominate their crime-ridden and economically underdeveloped neighborhoods, leaving others with fewer opportunities for social and income mobility as young people (predominately minority men) see joining gangs, rather than applying for legal jobs or attending community college, as their best opportunity to rise or have social affirmation. This also likely leads to poor schools, ill-equipped to educate local children, as teachers may opt for lower salaries elsewhere rather than work in gang controlled environments—whether or not they are actually more dangerous. On top of all of the costs thus far discussed—which may seem limited to local communities—there are costs to the entire country. First, all of the increased criminal activity and economic costs just discussed would permeate the nation. Organized crime figures in New York might have dealings with others in Florida, for example. Loss of purchasing power in Georgia means loss of business for those in Silicon Valley, for another example. The problem is clear. But we want to discuss another, non-financial cost, given that some paternalists argue for criminal drug laws on the grounds of the moral harms they cause. There are tremendous psychic costs of being citizens of what Justice Sotomayor calls the “carceral state.” With those, typically poorer and minority people, “who are routinely targeted by police. . . [we should] recognize that unlawful police stops corrode all our civil liberties and threaten all our lives. Until their voices matter too, our justice system will continue to be anything but” (dissenting opinion in Utah v. Strieff). This cost is, on its own, too high. Increased surveillance reduces privacy and makes more people feel as if they are not trusted and as if they cannot trust their neighbors. Civil asset forfeiture and the militarization of police cause the same problem. These intrusions of liberty and privacy do not merely compromise the ability of the targeted to develop morally virtuous characters, let alone have basic self-respect as a free and equal member of society—they also compromise the characters of those who would intrude without a second thought.20 Some may think these problems only exist for those “routinely targeted by police”—and that we, overwhelmingly white male academic types, have nothing to worry about. Ignoring the obvious injustice present in such thinking, we would also note that a social environment with these problems is also likely to foster a dangerous status quo bias whereby we unreflectively become conformists, thinking it’s only “bad people,” “poor people,” or “people of color” that have a problem and that it’s a problem of their own devising, such that their lives would go better if only they would abide

Paternalists and Social Welfarists  235 by the law. We hear this sort of claim far too often—“Don’t do the crime if you can’t do the time!”—sometimes with a smugness, usually with a seeming lack of understanding that the likelihood of someone being subject to correctional interference is directly correlated with where they were raised. Such attitudes are morally problematic and represent a condescending dismissal of the less privileged as being incapable of moral equality. Those of us who grew up in relative affluence are too likely to have an attitude like this that is corrosive to liberal democracy. It’s also an attitude that simply assumes might makes right and implicitly accepts that what the state says is illegal is, in fact, immoral. As Frédéric Bastiat (1801–1850) observed, “We all have a strong disposition to consider what is legal to be legitimate. . . . many people mistakenly consider all forms of justice to be founded in law” (The Law, 111). This bias is unfounded, of course. It may or may not be that drug use or addiction is typically immoral—we doubt it, but won’t argue that here—but we know all too well that law can be used to criminalize activities that are clearly moral (miscegenation, for example) and can encourage activity that is clearly heinous (slavery, for example). Encouraging some to think of themselves as morally superior to others is always risky21 and never likely to improve a polity.

5.  Further Objections Critics may raise a couple of objections at this point. First, they might claim that fewer people today are imprisoned for drug possession than twenty years ago, so we’re presenting a straw man. Indeed, violent crime in the US has declined since the early and mid-1990s. If the tide is turning against large prison sentences, mandatory minimums, and three strikes laws that may have helped reduce violent crime but no longer serve a good purpose, then we are harping on social problems that are now largely getting solved through efforts to ease the burden on victimless criminals. We claim, though, that too many people are still imprisoned for mere drug possession. A system that continues to allow those people—likely disproportionately the most vulnerable and worst-off—to remain imprisoned needs reform. Moreover, many people who commit nonviolent crimes against others (e.g., theft, distribution of child pornography) and so (arguably) warrant incarceration only act criminally because they cannot easily access drugs made artificially costlier or more difficult to find due to illegality. That is, the artificially higher cost of the drugs they seek leaves them ready to commit real crimes. Legalized, the prices of those drugs would decrease and crimes committed to support a drug habit would decline. A second, almost opposite, objection is that drug addiction remains a serious problem affecting wide swaths of the population, and we need to incarcerate users not so much for their own purported benefit but in order to deter would-be users. This is the “indirect” paternalism mentioned above. We are skeptical of this objection for two reasons. First, evidence suggests

236  Andrew Cohen and William Glod that “swift and certain” punishment is a more effective deterrent than long and harsh drug sentences (e.g., Nagin 2013). Second, and more importantly, it seems clear from the discussion above that the harms of such a deterrent system would far outweigh whatever harms are associated with drugs themselves. The myth of widespread addiction may motivate the call for deterrence.22 According to the book Chasing the Scream (Hari 2015), only 10% of drug users are actually addicts, few of them are dangerous, and most people stop using drugs without any treatment. Of course, there are different models of addiction, but regardless of which (in particular, a disease or habit model) is more accurate, addiction is not so all-consuming that addicts completely lack the ability to improve themselves or seek help or treatment. Drug addicts do not lack choice altogether, but even if some of the worst were ciphers, it is hard to believe incarceration helps them. Someone who cannot function in the outside world will not magically start functioning in prison. Addicts in prisons suffer greatly from withdrawal or find ways to feed their habits. Prisoners in many facilities can get drugs and other contraband, or spend their limited prison capital trying to attain them. Either way, it’s not clear how incarceration would help someone so addicted to a drug that they were unable to choose against using it. Critics might argue that incarceration forces these addicts to dry out, something they might never have done on the outside. While this might help the addict stay clean in prison, it does not address how making him “go cold turkey” improves his ability or willingness to make better decisions once free again (nor recidivism upon release). His lack of drug use is not due to a reason he endorses or an incentive he values. Mere behavior modification is unlikely to be a recipe for turning the hardcore addict of the disease model into the self-motivated, responsible person paternalists—and all of us—prefer him to be. For similar reasons, penal laws are a poor choice for helping people be better qua moral beings. An individual that merely acts because of external constraints does not thereby become a better person.

6. Alternatives We argued above that incarceration would be ineffective for those lacking motivation to quit harmful addictions. Incarceration would also be unnecessary for those who do have sufficient motivation. For those people, providing alternative opportunities more attentive to their freedom and well-being would be sufficient. We now sketch a non-exhaustive list of alternatives to incarceration (from worse to better) as an invitation to skeptics to explain why they would be less effective than imprisonment. We suggest them with a caution that we should not assume without good evidence that imprisonment is most effective. We are not claiming any of these alternatives are panaceas. There are no panaceas, but we should not add harm in the name of speculation that we are thereby reducing it overall. Some of these are also paternalistic, though perhaps requiring less invasiveness in the life of

Paternalists and Social Welfarists  237 drug users. We are not endorsing these, but merely putting them forward as alternatives that would create less harm then incarceration. The first alternative is a temporary mandatory rehabilitation program coupled perhaps with work-release. The patient might be required to spend a few weeks at a drug treatment center where she would have access to the resources necessary for overcoming her addiction. As she shouldn’t have to lose her job over this, arrangements should be made accordingly. But objections remain. Perhaps she doesn’t want to discontinue using the drug and functions well with it.23 Perhaps she objects to the treatment methods at the rehab facility. For instance, some people think 12-step programs are ineffective or even counterproductive (e.g., Johnson 2011). And of course, there are some functioning addicts, so the assumption that they need “treatment” may be biased and lead to their resentment.24 A better alternative is to decriminalize drug possession and use with civil penalties such as fines rather than imprisonment. Perhaps (the threat of) getting hurt financially will motivate some people not to use drugs, and the fines paid by those caught could fund drug treatment programs. However, this would seem to work only for those people who don’t have strong preferences to use, or who have a strong motivation to quit. Those are not the people who most concern paternalists. If addicts have a fairly inelastic demand for drugs, they might be willing to pay even heavy fines, which just leaves them poorer. Moreover, for fines to be effective they must be significant and will thus hurt some people. People who don’t pay them can end up in . . . well, jail or prison. People who are well off may simply pay the fines and continue to use the drugs—thus not, on the paternalists’ view, being helped. Of course, with these latter sort of users, one wonders why it matters that the law is intended to help them—they continue to function on their own, even if only because they are wealthy. A better alternative yet is to legalize drugs but tax them, using the revenues for drug treatment programs for the willing. Poorer addicts who can’t afford the taxes might resort to crime, so the taxes might be progressive and indexed to a person’s reported wealth or income. Those below a certain financial level will pay little to no tax. Whether this approach could overcome corruption, paternalists may not find the alternative satisfying since many will pay the higher tax without curbing their drug use, which harms them more than if there were no tax. Paying people to stop self-harm is better still. Concerned parties can provide resources to those willing to accept the offer; this would arguably be cheaper, or at least more efficient, than the criminal drug laws assessed above. Drawbacks to this approach include the fact that people might bluff when they don’t really intend to use drugs, and some people might simply not be willing to accept even the maximum of what others are willing to pay. Again, however, it is unclear how any attempt to help the unwilling can work if they are unmotivated to end their addiction. For those who are willing, commitment devices such as stickK.com are an option. A person who

238  Andrew Cohen and William Glod wants to quit drugs can make a public bet with a referee. He gets rewards if he succeeds, including public acknowledgement of his success, but he has to pay a penalty (often a promised amount held by the referee to be given to a group he despises, such as the Republican National Committee) if he fails and also face the shame of publicly admitting his failure. Whatever the merit of the above approaches, we think the root issue comes down to bonding and civil society. Addiction is often a symptom of perceived isolation (Hari 2015), so friends, family, and community groups (churches, clubs, etc.) can try to look after those who need help through informal norm enforcement that is sometimes crowded out by legal institutions trying to perform these tasks. Critics might be skeptical that such mechanisms can be effective at helping those who want help, but the question to ask is what institutions can promise better results given incentive alignment and local knowledge of the addict’s situation and personality. In addition, voluntary treatments like cognitive-behavioral therapy can teach skills to those who are motivated to learn healthier ways to cope with stress, anxiety, or depression. Again, none of these approaches are guaranteed to work. But paternalist and social welfarist critics of these approaches are ill-placed to regard incarceration as a better option. At this point, skeptics might consider settling for letting people fail and living with that as a lamentable cost of freedom. It is a fallacy to assume that the best realizable liberal order will or should be problem free. Bad things will happen and people have to choose how to respond to bad things—and they won’t always respond in the healthiest of ways. Doing nothing may seem like an unsatisfying approach, but given the costs associated with incarceration, we should not be tempted by action biases. “We must do something” is no excuse for making people’s lives worse in the name of supposedly making their lives better.

Conclusion As liberals, we believe that demands on the worst-off to be free and equal, morally accountable members of the political community generate a corresponding obligation on society to at least not punish them for crimes that generate only self-regarding harms. In our highly imperfect world, this means rights to indigent legal defense should they nonetheless be charged with crimes. More ambitiously, we argue that nobody should be charged with crimes that are self-regarding. Legal punishments, and laws more generally, should not have paternalistic rationales in a liberal society, and one reason is because such laws tend not to benefit the very people paternalists intend the laws to benefit. Indeed, they tend to harm many others beyond those incarcerated.

Notes 1 We thank the other contributors to this volume, especially those in attendance at the 2016 Workshop in New Orleans, for assistance in thinking through these

Paternalists and Social Welfarists  239 issues and how best to formulate our view. We also thank Brendan Dooley and Daniel Shapiro for comments on an earlier draft and Chris Surprenant for organizing the workshop and editing this volume. 2 To be clear, we are speaking about laws regarding adults only. It is perfectly appropriate to have laws that prevent children from choosing to use drugs. 3 We believe Doug Husak similarly suggested, in a talk, that the real motivation of the War on Drugs is actually a “war on crime.” 4 This is what Feinberg calls Moralistic Benefit-Conferring Legal Paternalism; it is one of two parts of perfectionism on his view. The other part is what he calls the Moralistic Benefit-to-Others Principle. Put together, Feinberg’s perfectionism is about improving the character of everyone in the polity; we are concerned primarily with those with whom we would interfere. For more on perfectionism as a type of paternalism, see Dworkin 2005 and George 1993. 5 All numbers in this paragraph from http://felonvoting.procon.org/view.resource. php?resourceID=004339. 6 In Washington, DC, it’s almost half that, but 3% is still exorbitant. Georgia is a startling outlier; the national average is 2.111%. We’ll use 2% here for simplicity. For the numbers, see www.prisonpolicy.org/reports/50statepie.html. 7 See, for e.g., http://highline.huffingtonpost.com/articles/en/sandra-blandjail-deaths/ (accessed July 14, 2016). 8 This view is also defended in work taking a capabilities approach to political thinking. See, for example, Nussbaum 2003. 9 In the DeKalb County (GA) jail, for example, the only communication friends and family have with prisoners is via a video conferencing call—whether they do so from home or from the jail, where computer access is provided. 10 For more on how prison gangs function, see Skarbek 2011. 11 Some state prison systems do have programs for prisoners to earn GEDs or even associate degrees, though the latter is rare. 12 We realize that some form of moralism may underlie many actual drug laws—and other paternalist laws. Indeed, what motivates the creation of any law is often quite different from its usually touted justification. 13 We should mention the clear direct costs of being on probation or parole. Of course, all of the above remain relevant for those on parole (see www.hrw. org/sites/default/files/reports/us0214_ForUpload_0.pdf) Those on probation or parole may have difficulty finding employment, may have to find time for appointments with correctional officers, and may be confined to certain areas for travel. More attention can and should be spent on this matter given that 1.4% of the American population is on parole or probation. 14 See his first footnote for more sources. Lippke also responds to objections that punishment drift might be appropriate. We find his responses satisfactory. See also Uggen and Stewart 2015. 15 New York State Office for the Prevention of Domestic Violence (2016). “Information for Professionals.” www.opdv.ny.gov/professionals/abusers/excuse2. html, accessed November 23, 2016. 16 We originally found this data at www.whitehouse.gov//sites/default/files/ondcp/ policy-and-research/fy_2016_budget_summary.pdf, but it is no longer available at that web address. We appreciate that the Federation of American Scientists archived much of the data that the US Federal Government previously kept on its websites. While we are not certain that this document is the same as the one we used originally, the data is the same. See Table 1 on page 16. 17 We also spend more than 11.5 billion annually on “prevention and treatment” (ibid). One might think that number would increase if we decreased the amount spent on enforcement, but for reasons that will become clear below, we suspect this is mistaken.

240  Andrew Cohen and William Glod 18 The best available work here seems to be Anderson 1999. His 2011 follow-up puts the total annual cost of crime, net of transfers from victim to criminal, at $1.7 trillion (209). Thanks to Frank Stephenson for the source. 19 Those in drug enforcement, on the other hand, might be less worried about impure or tainted drugs. As one officer has indicated, those sold bad drugs have ways of making their displeasure known to their dealers. 20 As Kristen Bell suggested to us, an additional concern here is that the sort of system engendered may (sometimes does) lead to an inversion of values wherein those in communities where incarceration hits hardest come to see going to prison, for drug offenses or otherwise, as a badge of honor. Selling drugs that are not good for you may then be seen as honorable. More generally, acting badly is seen as good. 21 It also risks the creation of “moral panics” with people worked up over chimeras like roving gangs of dangerous drug addicts. Those who panic are complacent when states use the law to oppress people engaged in non-threatening (to others) activities, including the abuses of power mentioned above. 22 It’s also a perfect example of a moral panic. 23 If drugs are decriminalized, perhaps those who go through a mandatory rehab program but choose to continue using should then be free to do so without further confinement. 24 See Brecher 1973 about eminent heroin addicts. Also Kaplan 1983 and Zinberg 1986. Thanks to Daniel Shapiro.

References Anderson, David (1999). “The Aggregate Burden of Crime.” Journal of Law and Economics 42: 611–42. Anderson, David (2011). “The Cost of Crime.” Foundations and Trends in Microeconomics 7: 209–65. Brecher, Edward (1973). Licit and Illicit Drugs. New York: Little and Brown. Cacioppo, John and Cacioppo, Stephanie (2014). “Social Relationships and Health: The Toxic Effects of Perceived Social Isolation.” Social and Personality Psychology Compass 8: 58–72. Chin, Gabriel (2002). “Race, the War on Drugs, and the Collateral Consequences of Criminal Conviction.” The Journal of Gender, Race, and Justice 6: 255–78. Cohen, Andrew Jason (2014). Toleration. Cambridge: Polity Press. Dworkin, Gerald (2005). “Moral Paternalism.” Law and Philosophy 24: 305–19. Faller, Kathleen (1988). Child Sexual Abuse: An Interdisciplinary Manual for Diagnosis, Case Management, and Treatment. New York: Columbia University Press. Feinberg, Joel (1984). Harm to Others. New York: Oxford University Press. Fried, Arthur, Anthony, Robert, Cseko, Andrew, Gaither, Carl and Schulman, Eric (2008). The Price and Purity of Illicit Drugs: 1981–2007. Alexandria, VA: The Institute for Defense Analysis. Available at www.whitehouse.gov/sites/default/ files/ondcp/policy-and-research/bullet_1.pdf George, Robert (1993). Making Men Moral. Oxford: Oxford University Press. Glod, William (2013). “Against Two Modest Conceptions of Hard Paternalism.” Ethical Theory and Moral Practice 16: 409–22. Glod, William (2015). “How Nudges Often Fail to Treat People According to Their Own Preferences.” Social Theory and Practice 41: 599–617. Hari, Johann (2015). Chasing the Scream. New York: Bloomsbury Publishing.

Paternalists and Social Welfarists  241 Johnson, Bankole (2011). The Rehab Myth: New Medications that Conquer Alcoholism. Boston, MA: De Capo Press. Kaplan, John (1983). The Hardest Drug. Chicago: University of Chicago. Labell, L.S. (1979). “Wife Abuse: A Sociological Study of Battered Women and Their Mates.” Victimology 4: 258−267. Lippke, Richard (2016). “Punishment Drift: The Spread of Penal Harm and What We Should Do About It.” Forthcoming in Criminal Law and Philosophy. Available at http://link.springer.com/journal/11572/onlineFirst/page/1. Metzner, Jeffrey nd Fellner, Jamie (2010). “Solitary Confinement and Mental Illness in US Prisons: A Challenge for Medical Ethics.” Journal of the American Academy of Psychiatry and the Law Online 38: 104–8. Nagin, Daniel S. (2013). “Deterrence in the Twenty-First Century.” Crime and Justice 42: 199–263. Nussbaum, Martha (2003). “Capabilities as Fundamental Entitlements: Sen and Social Justice.” Feminist Economics 9: 33–59. Sacco, Lisa (2016). “Drug Enforcement in the United States: History, Policy, and Trends.” Available at https://fas.org/sgp/crs/misc/R43749.pdf. Skarbek, David (2011). “Governance and Prison Gangs.” American Political Science Review 105: 702–16. Substance Abuse and Mental Health Services Administration Center for Substance Abuse Treatment (1997). Treatment Improvement Protocol (TIP) Series, No. 25: Substance Abuse Treatment and Domestic Violence. Rockville, MD: Substance Abuse and Mental Health Services Administration. Thornton, Mark (1998). “The Potency of Illegal Drugs.” Journal of Drug Issues 28: 725–40. Uggen, Christopher and Stewart, Robert (2015). “Piling On: Collateral Consequences and Community Supervision.” Minnesota Law Review 99: 1871–912. Zinberg, Norman (1986). Drugs, Set, and Setting. New Haven, CT: Yale University Press.

14 The Need for Prosecutorial Guidelines John F. Pfaff

At this point, it’s no longer necessary to explain how unprecedented incarceration is in the United States today. The highest incarceration rate in the world (outside of tiny Seychelles); 5% of the world’s population but 25% of its prisoners (Walmsley 2015).1 Yet there has been some good news recently: after nearly forty years of almost unceasing growth, prison populations in the US dropped for four of the five years between 2010 and 2014. The decline wasn’t great—about 4%, and under 2% if we omit California—but incarceration growth was never going to stop on a dime. Any sort of drop after four decades of unrelenting rise is laudable.2 Yet there is much to be concerned with in this drop as well. I don’t think the relatively slow rates of decline are just indicators of a slow start; I think they reflect more fundamental problems with the way in which reformers are approaching decarceration (see, e.g., Pfaff 2017). For all the bipartisan cooperation, for all the speeches, for all the legislation state legislatures have pushed through, there is a remarkable blind spot at the heart of prison reform: no one is taking aim at the one actor most responsible for rising incarceration rates, especially over the past twenty years as prison populations rose while crime rates plummeted. The county-elected prosecutor is profoundly powerful—and seemingly invisible. If we are going to successfully scale back the US’s staggering incarceration rate, it is critical that we find ways to effectively regulate and constrain prosecutorial discretion. Here, I want to argue for one proposal: charging and plea bargain guidelines for prosecutors. Right now, the only real way to regulate prosecutorial behavior is through periodic elections of the chief prosecutor. Although elections have historically posed little threat to incumbent prosecutors (Wright 2009), the 2015–2016 election cycle saw several high-profile tough-oncrime district attorneys lose to reform-oriented challengers in primaries and general elections alike, even in states that voted for Donald Trump (Ciaramella 2016; Lerner 2016). This has led some reformers to argue that we need to focus more on these elections in future years. It has also encouraged reformers to insist that prosecutors provide more information, which in theory would help the electorate make wiser choices about who to vote for (Nugent-Borakove, Budzilowicz, and Rainvile 2007; Hubbard and

The Need for Prosecutorial Guidelines  243 Ifill 2015). And to be clear, electing smart-on-crime prosecutors and shedding more light on prosecutorial behavior and decision making are clearly advances that we should celebrate and advocate for. That said, I believe there are real limits on what prosecutorial elections can accomplish, both in terms of decarceration in particular and optimal prosecutorial behavior more broadly. Elections determine who runs the office at the very top—which certainly matters in determining whether it is more or less aggressive as a general matter—but they cannot directly control the day-to-day decisions of line prosecutors, who are often vested with significant discretion to decide what cases to pursue and how to pursue them. This is particularly true in the larger departments, which have more line prosecutors than any chief could personally oversee and which handle most of the cases nationwide. In 2007 (the last year of available data), 25% of all criminal cases were closed in the 2% of all offices that served districts with over one million people, and those offices had an average of about 200 line prosecutors each; the 11% of offices in districts with 250,000 to 999,999 people handled another 40% of all cases and had offices with an average of 43 line prosecutors (Perry and Banks 2011). Given the power and discretion prosecutors have, as well as the complicated and hard-to-quantify tasks we expect them to fulfill, I think it is essential to rely on something more comprehensive than periodic elections focused on just the top official. I think it is imperative that we start to develop publicly available, legally binding charging and plea bargain guidelines to restrict and channel prosecutors’ choices at every decision point they face (which are many). Such guidelines could not only provide effective means for pushing down prison populations, but if carefully designed they could help ensure that prosecutors better respond to other normative values such as “fairness” or “justice,” as well as address problems that are not necessarily directly linked to the size of prison populations, such as racial disparities in who is charged and how severely they are charged. My goals in this chapter, then, are three-fold. First, to explain why prosecutors are so central to prison growth and thus why their omission from reform efforts is so troubling. Second, to show that the current methods we have to regulate prosecutorial power are inadequate to the task at hand. And third, to demonstrate that we need to think much bigger and much more boldly than the current conventional approaches—but also that such ambitious approaches are politically and practically viable. Prison reform results so far are disappointing because reformers are playing it safe. It’s time to swing more for the fences.

The Power of the Prosecutor To understand why charging and plea bargain guidelines are so important, it may help to demonstrate just how central prosecutors are to prison growth. The basic story is as striking as it is simple: over the course of the

244  John F. Pfaff 1990s and 2000s, as crime fell and prison populations rose, it appears that the dominant policy shift that pushed the incarceration rate ever higher was an increased willingness on the part of prosecutors to charge arrestees with felonies (Pfaff 2011, 2012, 2017). Here are the central facts. Between 1994 and 2008, in a sample of 34 states, violent and property crime rates dropped by about 30%.3 The police did not really get any better at closing crimes during the time, and the total number of arrests for violent, property, non-marijuana drug, and public order offenses fell by almost 10%.4 Yet while arrests—the necessary first input into incarceration—fell, the number of cases filed in felony court rose, by about 40%. Fewer arrests, more felony cases: fewer people entering the criminal justice system, but more of them facing felony charges. All told, in probability that an arrest resulted in a felony case nearly doubled between 1994–2008, a change entirely within the control of local prosecutors. Beyond that, little else changed, with both the probability a felony case resulted in a prison admission and the time someone actually spent in prison if admitted holding steady. In short, as crime began to fall in the 1990s, prison populations continued to rise almost entirely due to increased punitiveness on the part of local county prosecutors. Given our general lack of data on prosecutors, it’s almost impossible to say what caused this widespread change. We don’t know how much of the increase in felony cases comes from “upcharging” cases that would have been treated as misdemeanors in the past versus continuing with cases that would have been completely dropped before. We also have almost no evidence on what caused attitudes to harden and turn more aggressive, although it is possible to point to a few theories. Part of it is that we simply have more assistant prosecutors: between 1970 and 1990, as the violent crime rate rose by 100% and the property crime rate by 40%, the number of prosecutors rose by 17%, from 17,000 to 20,000. Between 1990 and 2007, however, as violent and property crime rates both dropped by 35%, we added three times as many prosecutors as we did during the crime rise, ending up with 30,000 by 2007 (Pfaff 2017: 129). Relatedly, between 1974 and 2007, the number of districts with full-time prosecutors rose from 44% to 85% (Pfaff 2017: 129).5 All told, evidence suggests that prosecutorial productivity was relatively stable during the 1990s and 2000s: prosecutors weren’t working any harder, but there were more of them, and more of them were full-time— and these new prosecutors couldn’t just sit around doing nothing. There are many other speculative explanations we could bandy about, but there are two more that will end up being particularly relevant when it comes to charging and plea guidelines. First, while the criminal justice system is nominally adversarial, the prosecutor’s chief adversary operates at a severe disadvantage. Approximately 80% of all defendants facing prison or jail time count as “indigent” and qualify for a state-provided lawyer, yet public defenders and other indigent defenders face persistent funding and

The Need for Prosecutorial Guidelines  245 caseload crises (Pfaff 2016a, 2017). Compounding this problem is that we spend substantially more on prosecutors, giving them an even bigger edge over overwhelmed public defenders (N.C. Office of Indigent Defense Services 2011; Perry and Bank 2011: 1). As a result, indigent defense right now finds itself in a crisis, one that surely allows prosecutors to secure increasingly favorable deals (Pfaff 2016a, 2017). Second, a critical but underappreciated moral hazard problem runs through our criminal justice system. Prosecutors are county officials, but prisons are state facilities—so when a prosecutor sends someone to prison, the costs are borne by the state, not by that prosecutor’s county. Prison is effectively free, allowing the prosecutor to reap all the tough-on-crime political credit without having to pay for it. In fact, the problem is worse than that, because jail and probation—the sanctions of lower-level misdemeanors—are county expenses. Sending people to prison is thus both politically more popular and cheaper. And while this hazard problem didn’t cause mass incarceration, since it predated the rise in prison populations, it certainly facilitated it as prosecutors became more aggressive. Yet with perhaps one exception, no reform bill has confronted this moral hazard problem (or, to be honest, any other aspect of prosecutorial power).6 For those who aspire to see our prison populations shrink, leaving the prosecutor unregulated is thus a huge failing. But even those who are ambivalent about decarceration should find the unfettered discretion given to prosecutors troubling. Prosecutors are called on to make far more decisions than any other actor in the criminal justice system—whether to charge or dismiss, whether to charge or divert, whether to charge as a felony or as a misdemeanor, whether to invoke a statute with a mandatory minimum or without, what sentences to seek during a plea deal, and whether to seek incarceration or something less punitive—and they face almost no assistance in making these calls. As former Boston prosecutor Adam Foss (2016) pointed out in a gripping TED talk, young inexperienced prosecutors are forced to make these sorts of decisions every day, with little sense of or guidance about what normative goals they are trying to accomplish, and with little to no idea of how to make the right policy call, which often requires an understanding of psychology, criminology, and other social science topics that most prosecutors never study in law school. Decarceration requires guidelines, but so too do equity and justice. So far, however, such guidelines remain almost completely outside the realm of policy debates. It turns out that New Jersey actually has plea bargaining guidelines, but it is the lone state to use them, they apply to only six specific drug offenses, and no one raises them when talking about reform options. Right now, much more hope is being placed on prosecutorial elections. As I’ll show now, the changing political fortunes of punitive district attorneys is something to celebrate, but we should appreciate the limits to this approach as well.

246  John F. Pfaff

The Limits of Prosecutorial Elections Over the past few years, prosecutorial elections have suddenly begun to seem more relevant. Historically, these elections have not been seriously contested: one study of elections in ten states between 1996 and 2006 found that incumbent prosecutors won 95% of the primaries and 95% of the general elections in which they ran, in no small part because 85% of these races were uncontested (Wright 2009: 595).7 Recently, however, a spate of incumbents have lost high-profile primary and general-election races to challengers who framed their campaigns as pushbacks against excessively punitive policies. A strategy that would have seemed impossible even just in the early 2000s has worked in places such as Cook County, Illinois (Chicago); Caddo Parish, Louisiana; Mississippi’s 16th Judicial District; Florida’s Fourth Judicial Circuit (Jacksonville); Florida’s Ninth Judicial Circuit (metro Orlando); Harris County, Texas (Houston); Nueces County, Texas (Corpus Christi); and St. Louis, Missouri (Neyfakh 2015; Ciaramella 2016; Currier 2016; Lerner 2016; Smith 2016). Of course, the full story is somewhat more complicated. In most if not all of these races, the incumbent was hampered by at least one if not several scandals, like Alvarez’s failure in Chicago to aggressively pursue the Laquan McDonald killing; none of these races was simply about general severity. Nonetheless, each reflected a significant and important shift in attitude on the part of the electorate, and the successful challengers all made smarter approaches linchpins of their campaigns. Furthermore, George Soros pumped hundreds of thousands of dollars into several of these races, suggesting that some reformers are starting to view them as a key arena to effect reform going forward (Bland 2016). The attitude of the person the people elect as their chief prosecutor definitely matters, and a change in approach can have clear and immediate implications. For example, Ken Thompson, elected in 2013 on a reformist platform as the district attorney for Kings County, New York (Brooklyn), stated that his office would substantially scale back its prosecution of low-level marijuana arrests (Clifford and Goldstein 2014). More generally, New York State’s prison population started to drop from historic highs in 1999, even as the state legislature was adopting tougher sentencing laws for people convicted of violence. The main engine of the decline was a steady drop in people sent to prison for drugs from the five boroughs of New York Citya decision that predated by several years any reform of New York’s drug laws and that clearly reflected the policy preferences of the boroughs’ five district attorneys, since actual drug use and offending surely did not drop as precipitously (Austin, Jacobson, and Chettiar 2013; Pfaff 2015). Yet there are a few limits to elections that deserve notice. First, prosecutorial elections are vulnerable. One bad shock can undo decarcerative efforts, either by changing an election outcome or the behavior of those already elected. In New York City, for example, the murder of police officer

The Need for Prosecutorial Guidelines  247 Randolph Holder by someone freed through a diversion program led even liberal mayor Bill de Blasio to immediately adopt a tough-on-crime attitude (CBS 2015). Legislatively enacted guidelines, however, are naturally more immune to such pressures. They can, of course, be amended, but that takes time and political capital; they are more likely to weather a shocking crime-of-the-week storm, though they are certainly not invulnerable.8 The precariousness of prosecutorial elections is magnified by the relatively low turnout in most district attorney elections. In 2013, for example, Joe Hynes, the 5-term incumbent prosecutor in Brooklyn, faced a bitter primary challenge that ultimately led to his defeat. It was the first time in over one hundred years that a sitting Brooklyn district attorney ran for reelection and lost, but the turnout in that primary was only 20% (Pfaff 2017: 141). Similarly, in 2012 in Cuyahoga County, Ohio (Cleveland), 165,000 voters—or 34% of the 482,000 who turned out to vote—declined to cast a vote in the race for the open district attorney’s position, even though they were already in the polling booth (Reid 2015; Pfaff 2017: 142). Such low turnout means that a small, determined group can exert undue influence on election outcomes. Right now that turnout seems to favor reformers, but that need not remain so. If anything, the “politics of crime” are such that fear of rising crime is more likely to drive people to the polling stations than concerns about excessive punishment. Second, there is a serious informational problem. In his critique of prosecutorial elections, Wright (2009) points out that most contested elections turn on the outcome of a few salient (and likely non-representative) cases, not on the overall general performance of the office; this likely helps explain why even now almost all the incumbents who have been defeated have been hampered by scandals. Moreover, to the extent that performance is an issue in an election, the debate usually focuses on just one metric, the conviction rate. Prosecutors, however, are called on to do far more than just win convictions, even if convictions are what determine electoral victories. This has led some people to argue for the need to gather more data on what prosecutors do so voters can be more informed (Hubbard and Ifill 2015).9 There is no doubt that we should demand much more information from prosecutors. Unlike police and courts and prisons, prosecutors provide almost no data to the Bureau of Justice Statistics or the FBI, so we have almost no idea what exactly they do or how they do it. It’s almost a cliché by this point, but prosecutors really are the black box of the criminal justice system. Arrests go in, prison admissions come out, but we don’t know what happens in between. Prosecutors deserve to face the same levels of public and academic scrutiny as everyone else in the criminal justice system. That said, as a way of ensuring proper prosecutorial behavior, there’s a serious limitation to data-plus-elections: unlike police and courts, prosecutors are expected to perform a complex, multidimensional job that is hard to measure well.10 The National District Attorneys Association, for example, recently attempted to design a set of metrics to evaluate prosecutorial

248  John F. Pfaff performance (Nugent-Borakove, Budzilowicz, and Rainvile 2007). In the end, the set required 35 different measures, spread across three major domains and nine subdomains, with the metrics ranging from the easy to count, like trends in crime rates, to the much harder to quantify, such as district-level happiness with criminal justice outcomes. As a practical matter, such comprehensive data would be hard to gather—especially for smaller offices—and near impossible for voters to process. A smaller set of metrics could be more tractable for prosecutors and voters alike, but by now it should be clear that transparently measuring what prosecutors do will always be difficult. And a binary tool like elections may make it hard if not impossible to balance competing interests across all the various factors. A third limitation is that elections can’t fix punitiveness in counties that want to be punitive. Perhaps to localists, this isn’t a problem: harsh counties should be free to be harsh, “smart” counties to be smart. Fair enough, so long as counties are forced to pay for the state-provided prison beds that they use (which brings us back to the moral hazard problem). Yet there may be concerns about allowing this sort of local choice. In economically or racially diverse counties, for example, punitiveness may reflect the choices of one group while the impact of those policies is disproportionately borne by another; in urban counties, for example, we should be concerned that wealthier, whiter suburban voters will select harsh(er) prosecutors in part because these voters don’t feel the costs of aggressive enforcement in poorer and more-minority areas (Stuntz 2013; Pfaff 2016b, 2017). Elections can’t address such intralocal problems. Something like guidelines can. And fourth, elections and data are inadequate tools to address the myriad day-to-day challenges that prosecutors face, which arise from the unavoidably human failings of prosecutors, not the broad policy goals of those at the top of the office. We should, for example, be concerned that implicit racial bias on the part of line prosecutors shapes who gets charged, what they get charged with, whether they face a mandatory minimum or not, and so on (see, e.g., Rehavi and Starr 2014). There are also concerns that prosecutors may be aggressive less out of innate desire and more from a realization that the costs of being lenient are larger and more salient than those from being too harsh, to the prosecutor and public alike (Pfaff 2016c, 2017). Furthermore, as noted before, prosecutors are called on to make a host of decisions every day with little or no training or guidance in the psychology and criminology that should inform those choices; even the best intentioned prosecutor is sure to make consistent, systematic errors. Elections can install someone at the top more or sensitive to these concerns, and that matters. But direct regulation is needed to tackle these issues most effectively.

Direct Regulation: Plea Bargain Guidelines in New Jersey While we should continue to push to elect district attorneys who embrace smarter, less-punitive approaches to crime, it is also essential to develop

The Need for Prosecutorial Guidelines  249 tools that restrict and regulate prosecutorial discretion on a daily basis. We regulate every other actor in the criminal justice system; there is nothing special about prosecutors that absolves them from similar needs for guidance. The guidelines they require may look different than those used by judges or parole boards, and the limits on discretion they should face may not be the same. But the difficult choices they are called to make, the sorts of political pressure they face, and their ability to free-ride off state resources all argue for state-developed guidelines to regulate the vast charging and plea bargaining discretion they currently have. The idea here isn’t to encourage district attorneys to develop internal guidelines to regulate their own decision making. Some offices—perhaps many, perhaps only a few (we lack the data to really know)—already have guidelines like these. The goal is to adopt legislatively enacted, legally binding guidelines. It’s worth pointing out that nearly half of all states have adopted some form of formal sentencing guidelines for judges (see, e.g., Kauder and Ostrom 2008), and judges are elected in almost all of these states.11 This suggests that for judges, many states view the electoral check as insufficiently constraining. Why, then, would we think that the electoral check would be enough for prosecutors—who make far more decisions, and far more complicated ones, and do so with far less transparency? Prosecutors need guidelines as much as, if not more than, judges. It may help to start this discussion by looking at the one set of such guidelines that already actually exist: the Brimage Guidelines in New Jersey (New Jersey Attorney General 2004). These guidelines arose as the culmination of a long-running struggle between the New Jersey Supreme Court and the state legislature, with the Court clearly fearing that the legislature was giving too much sentencing power to prosecutors. As early as 1976, the state supreme court held that prosecutors had to develop written guidelines to determine who was eligible for certain pre-trial diversion programs (State v. Leonardis, 363 A.2d 321 [1976]). The struggle culminated in 1992, when the state supreme court held in State v. Vasquez, 609 A.2d 29 (1992), that the state Attorney General had to develop reviewable guidelines to determine when prosecutors could waive certain mandatory minimums under New Jersey’s drug laws. The Attorney General developed a set of model guidelines, but it issued them to local prosecutors as suggestions around which they could build their own jurisdiction-specific versions. Six years later, in State v. Brimage, 153 N.J. 1 (1998), the state supreme court held that county-specific guidelines allowed for too much inconsistency, and it instructed the Attorney General to draw up guidelines that applied to all prosecutors statewide. The resulting Brimage Guidelines are a complex, comprehensive set of rules that regulate plea bargaining decisions for six major categories of drug offenses in the New Jersey code. The guidelines look like the sentencing guidelines that restrict judges nationwide. For each offense there is a grid, with the severity of the offense on the left axis and the defendant’s prior

250  John F. Pfaff criminal history on the top axis. For each severity-history pairing, there is a recommended plea sentence.12 Each box on the grid also contains aggravated and mitigated options, which the prosecutor can offer only if he establishes that certain aggravating or mitigating factors exist. Fig. 14.1 provides one of the Brimage tables and Fig. 14.2 its list of aggravating factors, which should look familiar to anyone who has seen sentencing guidelines before. For all the power that prosecutors have, and for all the attention given lately to criminal justice reform, the Brimage Guidelines seem to have fallen through the analytic cracks. As far as I can tell, almost no one discusses them,13 and there appears to be no rigorous analysis of the guidelines or how well they work. The anecdotal evidence, however, is at least fairly positive. The AG’s office has already adopted plea bargaining guidelines for other offenses, ranging from shoplifting to aggravated sexual assault of a minor, suggesting that it finds them appealing—although none of the other guidelines are remotely as rigorous as the Brimage ones (Farmer, Jr. 2001; Hoffman 2014). And some informal discussions of the Guidelines certainly suggest that they have improved consistency in the cases they cover (Wright 2005b). People usually point to two immediately obvious criticisms when I raise the idea of adopting Brimage-style guidelines more widely. Neither poses a serious threat to my idea in the end, but both deserve attention. First, it may appear hard to export prosecutor-regulating laws from New Jersey to the rest of the country. New Jersey is one of only four states where prosecutors are not elected by county voters, but are instead appointed by the state Attorney General—which gives AG-designed guidelines more authority than they would elsewhere in the nation. This is not an irrelevant point, but it is more a red herring than it seems at first. It doesn’t mean that other states cannot adopt guidelines, only that they will likely have to follow a different, and slightly more difficult, path. It’s true that outside of a state like New Jersey, the AG could not compel county-elected prosecutors to follow plea guidelines. But that isn’t needed to make sure prosecutors adhere to them. What we need is for the courts to insist they be followed, since courts have to sign off on plea deals. If courts were open to challenges raised by defense attorneys about misuse of the guidelines, then prosecutors would adhere to them, regardless of who chose the prosecutors. There’s already substantial evidence that courts adhere to guidelines written by the legislature that restrain their own power. Why, necessarily, would they resist rules that constrain the prosecutor’s? The question, then, is what would make guidelines sufficiently legitimate in the eyes of the judiciary that it would enforce them. My instinct is that in states outside New Jersey the guidelines would have to come from the legislature, not the AG’s office; or, at the very least, the legislature would have to authorize the AG to draft the guidelines. If New York’s Attorney General simply wrote up some guidelines and told county prosecutors to follow them, it’s easy to imagine that courts and prosecutors alike would view them as illegitimate. If the legislature imposed them, however, the

Pre-Indictment Initial Post-Indictment Final Post-Indictment

Pre-Indictment Initial Post-Indictment Final Post-Indictment

Pre-Indictment Initial Post-Indictment Final Post-Indictment

Pre-Indictment Initial Post-Indictment Final Post-Indictment

Pre-Indictment Initial Post-Indictment Final Post-Indictment

2C:35–5 4th degree weapons

2C:35–5 3rd degree no weapons

2C:35–5 3rd degree weapons

2C:35–5 2nd degree no weapons

2C:35–5 2nd degree weapons

B.

C.

D.

E.

F.

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

I Minor

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

II Significant

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

n/a n/a n/a

III Serious

36 42 45

30 36 39

24 30 33

18 24 27

12 18 21

 9 12 15

12 18 21 18 24 27 24 30 33 30 36 39 36 42 45 42 48 51

18 24 24 30 27 33 24 30 30 36 33 39 30 36 36 42 39 45 36 42 42 48 45 51 42 48 48 54 51 57

48 54 57

42 48 51

36 42 45

30 36 39

24 30 33

18 24 27

54 60 63

48 54 57

42 48 51

36 42 45

30 36 39

24 30 33

V Enhanced Extended Term

12 18 18 24 21 27

IV Extended Term

*If the offense involves a first-degree amount of methamphetamine or marijuana, use Table #5. (See Special Offense Characteristic #3 in Section 7.3.)

Figure 14.1 The Brimage Guidelines

Pre-Indictment Initial Post-Indictment Final Post-Indictment

2C:35–5 4th degree no weapons

TIMING

A.

OFFENSE DESCRIPTION

252  John F. Pfaff

Figure 14.2  Aggravating Factors

guidelines would have greater democratic legitimacy and courts would be more likely to insist they be followed, as they generally do with legislatively established sentencing guidelines. In other words, that prosecutors in New Jersey are appointed may change who can draft the guidelines, not whether such guidelines can be drafted in the first place.

The Need for Prosecutorial Guidelines  253 The second issue has to do with the complexity of the Brimage Guidelines. They run to over a hundred pages, including six separate tables, just to regulate six offenses. This seems to argue against widespread expansion, not just to other offenses, but to other stages of the criminal trial process as well. But this, too, is mostly a red herring. Look again to state sentencing guidelines, which are able to handle the entire criminal code with one or two universal grids. Expanding the Brimage Guidelines to other offenses would likely not require much expansion of the mitigators, aggravators, prior history calculators, and so on, and it would not take much work to condense all the crimes into just a few grids. In some ways, developing guidelines is like making a new drug: creating the first pill (or the guideline for the first crime) requires tremendous work, but the second pill/crime comes far more cheaply. It’s true that the expansion of guidelines to earlier stages in the process—not just plea bargain outcomes, but charge selection, diversion decisions, and so on—would require the development of front-end risk-needs assessment tools. This would require time and effort, like that that has been spent designing back-end risk assessment tools for parole. While the returns on such an investment would more than justify it, such initial costs could prove daunting to many jurisdictions. The current emphasis on evidence-based decision making, however, gives me hope that at least some states would be willing to take the time to design such tools. So I want to argue that not only should (and can) states adopt Brimage-style plea bargain guidelines, but that they should extend them to cover all offenses, and that they should extend them to cover all major decision points that prosecutors face. Regardless of their impact on prison population size, such guidelines will inject necessary consistency and reviewability not only into the plea process, which produces almost all of our convictions, but into the charging process, which determines whether someone faces the risk of conviction at all. Furthermore, these guidelines provide a vital way to rein in prison growth by directly limiting and channeling the (often punitive) discretion of those with the greatest control over incarceration rates.

Expansive Charging and Plea Guidelines To understand the need for far-reaching guidelines, consider a story told by the former Massachusetts prosecutor, Adam Foss, about a case he faced when he was just a second-year prosecutor (Foss 2016). With no real guidance or oversight, he had to decide whether to crack down harshly on a young man who had been arrested for stealing several computers from a store. On his own, Foss worked out a solution with the store that allowed the young man to avoid a conviction; years later, Foss ran into the man at a party and discovered that he had turned things around after his encounter with Foss and was working in a managerial position in a bank. Though heartwarming, the story is in many ways unnerving. It shows just how much power Foss, with only a few years of experience, had over a

254  John F. Pfaff person’s life. This story had a happy ending, but it’s easy to see how it could have gone wrong either way: perhaps the young man would have gone on to commit even more serious crimes, or perhaps a different prosecutor would have cracked down harder on him and inadvertently thwarted those future successes. In many ways, to celebrate what happened here is to celebrate luck—and, to no small degree, capriciousness. Foss himself admits that he was never trained in law school or on the job to make these sorts of decisions about how to best approach a case like this young man’s. He acknowledges that he went on instinct. Here, he got it right. But how many times do prosecutors get it wrong, one way or the other? It’s hard to understate how many different decisions line prosecutors are called on to make: whether to press charges at all, whether to divert someone to an alternate program, whether to charge the offense as a felony or misdemeanor, whether to charge the felony that has a five year maximum or ten, whether to invoke the version that includes a mandatory minimum or not, whether to stack multiple charges or not. These decisions are subject to no external review; we don’t even have the data to examine them if we wanted to. Our only check right now is a once-every-four-year election of the person running the office, based on slivers of information, little of it reflective of the critical decisions being made daily in the office. Whether our goal is decarceration or just consistency, this sort of unfettered, unreviewable and unreviewed, untrained-for discretion should concern us deeply. The New Jersey Supreme Court insisted on guidelines out of relatively narrow concern over separation of powers, that the legislature was using mandatory minimums to effectively give judicial power to prosecutors. My argument for guidelines is much broader than that. Part of my argument is consequential: prosecutors need assistance with the sort of psychological and behavioral issues central to deciding how to optimally deter or incapacitate or rehabilitate. Part of my argument, however, is also normative, that guidelines can introduce regularity and transparency to how prosecutors balance various competing values, like deterrence and retribution, community safety and mercy.14 And in the process, guidelines will advance a third goal, namely helping prosecutors overcome their own biases, both implicit and explicit. In theory, it’s relatively easy to sketch out what these sorts of guidelines should look like:

• The decision to charge in the first place. The guidelines could use a

risk-needs tool to decide which defendants would be better served outside the criminal justice system.15 The guidelines could state, for example, that defendants arrested for minor offenses and who have sufficiently low risk scores should not be charged at all, unless certain aggravating factors exist. Conversely, the guidelines could state that some defendants cannot have their cases dismissed absent certain mitigators. These decisions can reflect not just utilitarian goals like optimal incapacitation, but normative ones like retribution.

The Need for Prosecutorial Guidelines  255

• If the case is not dismissed, the guidelines can then recommend whether

•

•

the defendant should see his case diverted to some sort of treatment or therapeutic option, like drug court, anger-management training, or so on. Like with the decision to charge at all, a risk-needs tool combined with aggravating and mitigating factors would determine what the guidelines instruct the prosecutor to do based in part on the defendant’s amenability for treatment. For those defendants who are to be charged and do not qualify for diversion, the guidelines would then provide a rule for what the charge should be. They could say, for example, that a given assault must be charged as a misdemeanor unless certain aggravators are present, or as a felony unless there are mitigating circumstances. The guidelines could even instruct prosecutors about when it is appropriate to charge an offense that carries a mandatory minimum when there is a similar charge available that does not. The guidelines would, again, balance competing values more transparently and consistently than relying on the personal calls of individual prosecutors. Finally, once the charges are selected, the guidelines, like those in New Jersey, would specify the appropriate range of sentences to offer during plea bargaining, or to seek at trial. And these too would be shaped by various mitigating and aggravating factors (and not necessarily the same ones as before, though possibly so).

At first blush, this list may make guidelines seem utterly impossible to create, but I think that’s incorrect. It should be possible to design guidelines that would be fairly easy to use, even when caseloads are high. When the case is handed over to the prosecutor’s office, the prosecutor would need to run a risk-assessment tool and a needs-assessment tool; those results, along with the defendant’s prior criminal history and the police case file, would provide enough information to make most of the decisions listed here. The risk score and prior history, along with the offense, may say that charges need to be filed; the needs score may say that the defendant is unlikely to succeed in the available diversion programs; the case file and prior history may point to a felony over a misdemeanor, but the risk score along with other evidence may then lead the guidelines to point to a lesser level of felony within the set of feasible charges. There are, of course, a host of implementation questions that immediately come to mind. Should the guidelines, for example, be presumptive or binding? Presumptive guidelines would take the form of “you must charge [conduct] as a misdemeanor, unless [mitigators] exist, in which case you may dismiss the charges,” while binding guidelines would state that the prosecutor must dismiss the charges if the mitigators are present. Almost all sentencing guidelines are presumptive, not binding. Here, it may make sense to make the charging guidelines binding for leniency (“must dismiss”) but presumptive for severity (“may charge the more-serious felony”), since

256  John F. Pfaff the politics of punishment generally push actors towards severity and away from leniency. It also may be necessary to figure out how to update results as the case progresses and new information becomes available. A tool that says “do not dismiss” when used at the start of the case may have yielded “do dismiss” if it had all the information that the prosecutors have later in the case. This could be easily managed, however, by simply requiring prosecutors to periodically update the model when new, relevant evidence arrives. Conceivably, if the guidelines are public even defense attorneys could see how the guideline’s instructions change with the introduction of new evidence that is favorable to their clients, giving the disadvantaged defender a bit more leverage.16 There are numerous other issues to consider as well. How and when, for example, could the defense challenge what it sees as misuse or misapplication of the standards? And what burden of proof should prosecutors have to meet when including evidence, particularly aggravating evidence, in the assessment (how sure do they have to be that, say, the victim was particularly vulnerable)? How can we ensure that guidelines do not reinforce structural biases in criminal justice instead of ameliorating them?17 And then there are the usual, more general concerns about the accuracy and biases of risk-need tools and other such actuarial models.18 But none of these issues is intractable or poses a serious threat in the abstract. Politics, of course, could make some of these issues complicated to resolve in practice, but none is conceptually impossible, which is where we need to start; even the political issues are plausibly surmountable. I want to conclude this section by just briefly highlighting three broad benefits that guidelines would introduce, if only to emphasize why overcoming the implementation costs should be worth the effort. First, guidelines, especially binding guidelines for leniency, could confront some of the political problems that bedevil criminal justice and that so far have received little attention from reformers (Pfaff 2016c, 2017). A prosecutor may wish to dismiss a case but fears the backlash if the defendant ends up recidivating. Since the prosecutor doesn’t pay much of the cost of excessive incarceration—nor, as we saw above, does the county that employs him—he’ll generally err on the side of severity. Binding guidelines may help push back against the political pressure by giving the prosecutor plausible deniability: “I wanted to go forward, but the model tied my hands.” Binding, here, is more effective than presumptive and helps counter asymmetric political pressures. Besides addressing decision-making problems faced by individual prosecutors, guidelines also have the ability to address at least two macro-level problems in criminal justice today. First, they may help level the playing field between prosecutors and underfunded public defenders. Publicly available, legally binding guidelines would restrict the threats that prosecutors could bring to bear during pleading, and they would provide a more concrete framework for the plea process, both of which would likely work to the public defender’s advantage. Second, guidelines directly confront the

The Need for Prosecutorial Guidelines  257 moral hazard problem that arises from giving county prosecutors unlimited access to state prisons, by allowing the state to regulate local prosecutors’ ability to send people to state prisons.19 Similarly, the guidelines can advance state decarcerative goals (where those exist) by enabling the state to make it increasingly hard for prosecutors to plead someone out to prison time, or by limiting the maximum time they face. States, in other words, can design guidelines to explicitly restrain or cut back prison populations.

Conclusion The current treatment of prosecutors by criminal justice reformers is frustrating. Despite clearly being the actors most responsible for driving up the rates of incarceration, especially throughout the crime decline that started in the early 1990s, they have been the subject of almost no reform efforts. And while increased attention to elections is promising and should be encouraged further, there are limitations to how much control they bring, especially to the myriad day-to-day decisions made by line prosecutors in busy offices. We already restrict judges, even elected judges, with detailed, binding, publicly debated guidelines. Although perhaps more complicated to design, such guidelines are all the more essential for prosecutors.

Notes 1 Note that these tables assume that all countries, including autocratic states such as Russia, Cuba, and China, are reporting honestly, and it does not include Eritrea, Somalia, and North Korea. That said, even if due to underreporting our rate is actually below that in countries like Russia or Cuba—among our closest neighbors in the international rankings—that should provide no comfort. The nations we view as our allies and peers are all so far below us that the differences are real and substantial. 2 National-level data on prison populations ends in 2014. Unless otherwise specified, all data on prison populations and other criminal justice statistics come from the online data tool provided by the Bureau of Justice Statistics at www. bjs.gov/index.cfm?ty=nps. 3 Both the states and the years reflect limitations in the data on felony filings in state courts, which is compiled by the National Center for State Courts. The NCSC revised how they gathered the data starting in 1994, making it hard to compare pre-1994 results with those from 1994 onwards. Not every state provides data to the NCSC, though the 34 states represent a reasonable cross-section of big and small, more-conservative and more-liberal states (Pfaff 2011, 2012). 4 Given that I am focused on the immediate drivers of prison growth, not criminal justice control more broadly (i.e., jail, probation, etc.), I exclude marijuana arrests since almost no one ends up in prison on a marijuana charge: only about 1% of all state inmates are serving time for marijuana. 5 There is surely a geographic distinction between these effects. Most of the assistant prosecutors were likely added in urban jurisdictions, since most non-urban districts have only two or three prosecutors total and thus little capacity to expand staffing significantly. Conversely, most of the growth in full-time prosecutor offices must have happened in smaller districts: Brooklyn or, say, Atlanta did not have the luxury of waiting until the 1980s to have a full-time prosecutor.

258  John F. Pfaff 6 The one exception is California, whose complicated “Realignment” law does attempt to push some of the costs of incarcerating lower-level offenders back onto the counties. 7 In contested races, incumbents fared slightly worse, as one would expect, but still generally won: they remained in power in 64% of the primaries and 69% of the general elections. In larger counties, incumbents fared even better, winning 75% to 80% of contested general elections (Wright 2009: 596). 8 Minnesota’s sentencing guidelines, for example, were heavily amended in 1989 in response to several violent rapes (two resulting in murder), and again in 1992 in the wake of two separate abduction-murders (Robina Institute 2016). 9 I fall among this number, as a signatory to Hubbard and Ifill (2015). While I still think it is essential we gather more data on prosecutorial performance, I am less sanguine about the impact such data-gathering on its own will have on prosecutorial behavior. 10 For more extensive discussions of the uniquely complex role prosecutors play in American criminal justice, see Sklansky (2016a, 2016b). 11 Judicial selection methods are listed here: www.judicialselection.us/judicial_ selection/methods/selection_of_judges.cfm?state 12 Three recommendations, actually, based on when in litigation the plea takes place: the earlier the defendant agrees to the plea, the more generous the offer the prosecutor can make. 13 Wright (2005a, 2005b) are among the few exceptions. 14 Guidelines could, for example, require incarceration for some crimes on retributivist grounds even if the consequentialist argument for that is hard to make, or permit leniency towards someone due to extenuating risk factors. 15 Out of every 100 felony arrests, about one-third see their cases dismissed (either by the judge or prosecutor) or diverted (Cohen and Kycklehahn 2010). 16 Of course, the use of the model can’t stop Brady violations, when the prosecutor fails to turn over exculpatory evidence to the defense, but the fact that the model can’t cure all ills is not a strong argument against using it still. 17 This was an issue in New Jersey: initially the Brimage guidelines made racial imbalances in punishment worse by limiting urban prosecutors’ ability to bargain around racially problematic laws involving drug dealing near schools (Greene, Pranis, and Ziedenberg 2006). 18 See, for example, Harcourt (2006). Of course, the question is never whether models are biased, but whether they are more biased than prosecutors, and whether they are harder to debias than prosecutors. 19 This is not the only way to solve the moral hazard problem—Eck, Jonson, and Cullen (2013) propose an innovative cap-and-trade system in which counties trade permits for bedspace in state prisons, just like we have markets in pollution permits—but guidelines build off an already-established approach for regulating criminal justice actors.

References Austin, James, Jacobson, Michael P. and Chettiar, Inimai (2013). How New York City Reduced Mass Incarceration: A Model for Change? New York: Brennan Center for Justice. Available at www.brennancenter.org/publication/how-new-yorkcity-reduced-mass-incarceration-model-change. Bland, Scott (2016). “George Soros’ Quiet Overhaul of the U.S. Justice System.” Politico. Available at www.politico.com/story/2016/08/george-soros-criminaljustice-reform-227519.

The Need for Prosecutorial Guidelines  259 CBS (2015). “De Blasio Calls for Changes in State Bail, Diversion Programs in Wake of Officer’s Death.” Available at http://newyork.cbslocal.com/2015/10/23/ de-blasio-diversion-program-changes. Ciaramella, C.J. (2016). “Problem Prosecutors Lost Big on Election Night.” Reason Hit & Run. Available http://reason.com/blog/2016/11/10/problem-prosecutorslost-big-on-election. Clifford, Stephanie and Goldstein, Joseph (2014). “Brooklyn Prosecutor Limits When He’ll Target Marijuana.” New York Times. Available at www.nytimes. com/2014/07/09/nyregion/brooklyn-district-attorney-to-stop-prosecutinglow-level-marijuana-cases.html Cohen, Thomas H. and Kycklehahn, Tracey (2010). Felony Defendants in Large Urban Counties, 2006. Washington, DC: Bureau of Justice Statistics. Available at www.bjs.gov/content/pub/pdf/fdluc06.pdf Currier, Joel (2016). “Former Prosecutor Turned State Rep Takes St. Louis Circuit Attorney Primary.” St. Louis Today. Available www.stltoday.com/news/local/crimeand-courts/former-prosecutor-turned-state-rep-takes-st-louis-circuit-attorney/ article_3f31a308-d84f-52bd-8d9e-f19e3dfb4ea3.html Eck, John, Jonson, Cheryl Lero and Cullen, Francis T. (2013). “The Small Prison.” In The American Prison: Imagining a Different Future, ed. Francis T. Cullen, Cheryl Lero Jonson and Mary K. Stohr. New York: Sage. Farmer, Jr., John J. (2001). “Attorney General Guidelines—Prosecution of Shoplifting Offenses.” Available at www.state.nj.us/lps/dcj/agguide/shoplift.pdf Foss, Adam (2016). “Ted Talk: A Prosecutor’s Vision for a Better Criminal Justice System.” Available at www.ted.com/talks/adam_foss_a_prosecutor_s_vision_for_ a_better_justice_system. Greene, Judith, Pranis, Kevin and Ziedenberg, Jason (2006). Disparity by Design. Washington, DC: Justice Policy Institute. Available at www.drugpolicy.org/doc Uploads/SchoolZonesReport06.pdf Harcourt, Bernard E. (2006). Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age. Chicago: University of Chicago Press. Hoffman, John J. (2014). “Uniform Plea Negotiation Guidelines to Implement the Jessica Lunsford Act, P.L. 2014, c. 7.” Available at www.state.nj.us/lps/dcj/agguide/ lumsford_act.pdf Hubbard, William C. and Ifill, Sherrilyn (2015). “ABA-LDF Joint Statement on Eliminating Bias in the Criminal Justice System.” Available at www.naacpldf.org/ press-release/joint-statement-eliminating-bias-criminal-justice-system. Kauder, Neal B. and Ostrom, Brian J. (2008). State Sentencing Guidelines: Profiles and Continuum. Williamsburg: National Center for State Courts. Available at www.ncsc.org/~/media/Microsites/Files/CSI/State_Sentencing_Guidelines.ashx. Lerner, Kira (2016). “Overzealous Prosecutors Are Losing Elections.” ThinkProgress. Available at https://thinkprogress.org/overzealous-prosecutorsare-losing-elections-6dad096118e1#.vvmmhq8on. NC Office of Indigent Defense Services (2011). North Carolina’s Criminal Justice System: A Comparison of Prosecution and Indigent Defense Resources. Durham, NC: Office of Indigent Defense Services. Available at www.ncids.org/Reports%20 &%20Data/Latest%20Releases/ProsecutionOfIndigentDefense.pdf New Jersey Attorney General (2004). “Revised Attorney General Guidelines for Negotiating Cases Under N.J.S.A.” Available at www.state.nj.us/oag/dcj/agguide/ directives/brimagerevision.htm

260  John F. Pfaff Neyfakh, Leon (2015). “How to Run Against a Tough-on-Crime DA—and Win.” Slate. Available at www.slate.com/articles/news_and_politics/crime/2015/11/dis trict_attorneys_scott_colom_proves_you_can_run_against_a_tough_on_crime. html Nugent-Borakove, M. Elaine, Budzilowicz, Lisa M. and Rainvile, Gerard (2007). Performance Measures for Prosecutors. Alexandria: National District Attorneys Association. Available at www.ndaa.org/pdf/performance_measures_findings_07.pdf Perry, Steven W. and Banks, Duren (2011). Prosecutors in State Courts, 2007—Statistical Tables. Washington, DC: Bureau of Justice Statistics. Available at www.bjs.gov/content/pub/pdf/psc07st.pdf Pfaff, John (2011). “The Micro and Macro Causes of Prison Growth.” Georgia State University Law Review 28: 1237–71. Available at http://readingroom.law. gsu.edu/cgi/viewcontent.cgi?article=2700&context=gsulr. Pfaff, John (2012). “The Centrality of Prosecutors to Prison Growth.” Available at https://papers.ssrn.com/sol3/papers.cfm?ab-stract_id=1990508. Pfaff, John (2015). “The War on Drugs and Prison Growth: Limited Importance, Limited Legislative Options.” Harvard Journal of Legislation 52: 173–220. Available at http://harvardjol.com/wp-content/uploads/2015/06/HLL104_crop1.pdf Pfaff, John (2016a). “A Mockery of Justice for the Poor.” New York Times. Available at www.nytimes.com/2016/04/30/opinion/a-mockery-of-justice-for-the-poor.html Pfaff, John (2016b). “Why Do Prosecutors Go After Innocent People?” Washington Post. Available at www.washingtonpost.com/news/in-theory/wp/2016/01/21/ why-do-prosecutors-go-after-innocent-people/?utm_term=.75cc049115fb. Pfaff, John (2016c). “The Complicated Economics of Prison Reform.” Michigan Law Review 114: 951–81. Available at http://michiganlawreview.org/wp-content/ uploads/2016/04/114MichLRev.951_Pfaff.pdf Pfaff, John (2017). Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform. New York: Hachette Books. Rehavi, M. Merit and Starr, Sonja (2014). “Racial Disparity in Federal Criminal Sentences.” Journal of Political Economy 122: 1320–54. Available at www.eco nomics.ubc.ca/files/2015/01/pdf_paper_marit-rehavi-racial_disparity.pdf Reid, Joy-Ann (2015). “Tamir Rice Prosecutor Indicted Innocent Men, But Not Killer Cops.” Daily Beast. Available at www.thedailybeast.com/articles/2015/12/29/ rice-prosecutor-indicted-innocent-men-but-not-killer-cops.html Robina Institute (2016). Twitter. Available at https://twitter.com/RobinaInstitute/ status/804034589224210432 Sklansky, David A. (2016a). “The Nature and Function of Prosecutorial Power.” Available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2770815. Sklansky, David A. (2016b). “The Changing Political Landscape for Elected Prosecutors.” Available at https://papers.ssrn.com/sol3/Papers.cfm?abstract_id=2828803. Smith, Jordan (2016). “Hard-Line Prosecutors Face Rejects from Voters in Elections Across the U.S.” The Intercept. Available at https://theintercept.com/2016/10/20/ hard-line-prosecutors-face-rejection-from-voters-in-elections-across-the-u-s/. Stuntz, William (2013). The Collapse of American Criminal Justice. Cambridge: Belknap. Walmsley, Roy (2015). World Prison Population List, 11th ed. London: Institute for Criminal Policy Research. Available at www.prisonstudies.org/sites/default/files/ resources/downloads/world_prison_population_list_11th_edition_0.pdf

The Need for Prosecutorial Guidelines  261 Wright, Ronald F. (2005a). “Prosecutorial Guidelines and the New Terrain in New Jersey.” Penn State Law Review 109: 1087–105. Wright, Ronald F. (2005b). “Sentencing Commissions as Provocateurs of Prosecutorial Self-Regulation.” Columbia Law Review 105: 1010–47. Wright, Ronald F. (2009). “How Prosecutor Elections Fail Us.” Ohio State Journal of Criminal Law 6: 581–610. Available at http://moritzlaw.osu.edu/osjcl/Articles/ Volume6_2/Wright-FinalPDF.pdf

15 Prison Tunnel Vision Josh Dohmen

Like many others, I share the belief that prisons should be radically reformed or even abolished. This statement, that there are those who believe prisons should be rethought or dismantled, is often met, in my experience, with shock, anger, or disbelief.1 This suggests, I fear, that fundamentally changing prisons is an aim far disconnected from mainstream thought in the United States. To counter this, there are at least two avenues that would-be reformers must pursue. On the one hand, they must articulate alternatives to the practices of imprisonment that seem unquestionable in the United States. On the other hand, they must question the unquestionable nature of imprisonment; they must ask, in other words, why it is so difficult to conceive a world without prisons as they currently exist. It is this second avenue that I aim to address in this chapter, and here I am guided by two principles. First, I believe there are a cluster of philosophical tools that can help us understand why prisons are, or have become, so obvious and inevitable for us. Specifically, I will discuss three concepts and methods that help us understand why there is an entrenched “prison tunnel vision”—that is, why there is a strong resistance to alternatives to incarceration, and why there is an aversion toward forming coalitions with those who are incarcerated and the communities most affected by mass incarceration. This resistance comes not from simple habit, and certainly not from calm reason, but from investments we all have in our own histories, identities, and knowledges. Questioning practices of imprisonment, then, must also be a practice of questioning ourselves, and it is this latter form of questioning that I seek to encourage here. Second, I believe that it is important to engage not only my fellow philosophers, or even my fellow academics, but to address anyone interested in discussing prisons, their reform, and potentially their abolition. For this reason, it is my intent to articulate these philosophical tools and their potential application to our thinking about prisons in ways that are more widely accessible than many academic articles. If there are ideas that remain unclear at the end of the chapter, I hope this will be understood as a fault of my writing, and not evidence of the difficulty of the texts or ideas that I am drawing from.

Prison Tunnel Vision  263 Before discussing the specific philosophical tools that will be the focus of the chapter, it will be helpful to discuss several terms that I will use throughout in ways that may not match more colloquial uses. First, throughout the chapter I will use the term “subject” as a noun. By using this term, I hope to capture three aspects of what might otherwise be called “persons” or “individuals.” A subject is one who (a) has agency, (b) has an identity or self-understanding, and (c) is subject to a range of norms and laws that both enable and constrain her or his agency and identity. Let me take each of these aspects in turn. By “agency” I mean the ability of subjects to initiate their own thoughts or actions. Subjects have “identities” insofar as they imagine or conceive of themselves in particular ways. So I may identify myself as a drummer, a philosopher, a spouse, or as white, male, or a citizen of the United States. There are a couple of things worth noting about this list. First, these are identities that are thought, imagined, and felt. Consider my self-identity as a drummer. I may think that I am a drummer because I grew up playing a drum set, or because to this day when I sit in front of a drum set it feels natural to play. Notice that here I am giving explanations for my self-conception or self-understanding as a drummer. I also imagine myself as a drummer. This identity is as much a part of my self-image as a part of my self-conception. When I listen to music, for example, it is most natural to imagine playing the drum part and it is usually the part that I focus on the most—I feel the motions and rhythms. Finally, I also have emotional investments in this identity. When I make mistakes drumming I feel disappointment. When people make jokes about drummers being the least important member of a band I feel angry or insecure. Second, these identities are all relational and are voluntary to varying degrees. I can only identify as a philosopher, for example, in relation to others who identify as philosophers and those who don’t, in relation to educational institutions in which philosophy is considered a distinct field, and in relation to an evolving history that has taken up philosophical questions such that there is a sort of tradition that self-identified philosophers can variously take up, respond to, or reject. Also, my identity as a philosopher is voluntary in ways that other identities are not. I made a choice to pursue philosophy; I never made a choice to be white or male. Because of this distinction, Iris Marion Young has called the latter types of collectives “social groups.” Members of social groups find themselves to be members of those groups (Young 2011: 46). When I started learning about racial issues in the United States, for example, there was not a point at which I thought to myself, “There are various racial groups I could be a part of, and I elect to be white.” Rather, I found whiteness and whites as a social group to preexist me, and I found that I already had that identity. But this is where it is important to be careful and note that social groups, and our identities as members of social groups, are not fixed. This is because we are subject to norms, laws, and practices that exist before us and outside

264  Josh Dohmen of us, but which are culturally and historically variable, and which are perpetuated through our performances of them. Had I been born in Europe prior to the period of colonization, for example, there would simply not have been a white race to be a part of; the meaning of practicing philosophy would have been very different because educational institutions, the questions philosophers asked, and the norms and methods for pursuing philosophical investigations have all undergone radical changes since then; and the norms and laws governing marriage would make the meaning of being a spouse quite different from how I am able to understand it today. How I understand myself, how others understand me, and the options available for me to pursue are all shaped by these norms, laws, and practices. One final choice that I should note before moving on is that I will occasionally use the plural “knowledges” and “histories,” rather than “knowledge” and “history.” I do this to flag the ways in which what we know is shaped by our context, such that it would be inaccurate to say that there is knowledge in general to which we all have access. Instead, what I know is importantly limited by and enabled by my social and historical situation. In other words, what I know, and what counts as knowledge, is much different for me as a white, male academic writing in 2016 than it would be given a different identity, cultural location, or point in history. Similarly, discussing history in general runs the risk of telling history as one grand narrative or trajectory. By using “histories” I intend to draw attention to the more local and specific narratives and pasts through which we understand ourselves.

Genealogy “Genealogy,” as I will use it here, is a method for investigating the conditions of possibility for historical phenomena. It is often said that genealogies give us “histories of the present,” because they do not aim to provide definitive, unifying narratives, but rather to help us understand how we got to where we are (Koopman 2013). So, given the guiding question of this paper, a genealogical question would be “What are the conditions of possibility for the unquestioned acceptance of prisons in the United States today, despite their continual failure?” Notice that asking this question requires certain assumptions. First, the current historical situation is contingent; it did not have to be this way. Instead, we can study the various factors that led to the mass incarceration practices of the contemporary United States, factors that might have been otherwise. Second, the question is attentive to the particular context. It does not attempt to unify practices of imprisonment across the globe, nor does it assume that prison practices today are the same as they always have been or always will be. Another important aspect of genealogy is that it investigates historical relationships between powers and knowledges. “Power” here should be broadly understood to mean relations of force through which knowledge, identity, and behavior is shaped. So, for example, power is functioning

Prison Tunnel Vision  265 when schools shape curricula, design methods of assessment that define what counts as a student having knowledge, rank students by their year or grades, keep a continual record on the student that is taken to be evidence of who the student is, and enforce attendance or hand-raising policies. But power is also functioning when students contest this use of force, perhaps by refusing to care about standardized tests or using excellent student records to argue for leniency or different policies. Of course, there remains an imbalance of power, here, but it would be a mistake to understand power as unidirectional. Importantly, power is always related to knowledge, not just in a repressive way, but in a productive way. So, to return to the school example, educational institutions do repress certain knowledge by, say, leaving it out of the curriculum. “Abstinence only” education, for example, actively seeks to keep students ignorant of sexual and birth control practices. But they also produce knowledge in various ways. They define what counts as knowledge, for example, through their curricula and assessment standards. And they produce knowledge of the students through practices like the “permanent record,” a practice which gives rise to knowledge of students’ identities, histories, or tendencies, not just students’ individual actions. This is a type of knowledge that would not exist apart from a systemic practice of educational record keeping. The final point to note about genealogy is that it is attentive to the ways in which we experience and understand ourselves as subjects given our particular situation in history, our particular knowledge practices, and our particular relations of power. Recalling the general definition of the subject above should make this clear. One’s possibilities for action and thought, how one understands and imagines oneself, and the norms that enable and limit oneself vary based on one’s social and historical setting. To take seriously local histories rather than history in general, then, requires that we also take seriously the different ways that we experience, understand, and imagine ourselves. What does this have to do with our contemporary relation to prisons? First, let me briefly discuss Foucault’s own genealogy of modern prisons in Discipline and Punish ([1975] 1995). I will focus on two of Foucault’s guiding questions. First, why was there a transition in Western Europe and the United States from public spectacles of violent punishment to hidden punishments, that is, to prisons? Second, why do we continually seek reform of prisons in the face of evidence that prisons do not achieve their stated aim of rehabilitation? The common answer to the first question is that governments, or those exercising power in general, became more humane, in part through the influence of humanist activists. By paying attention to the subtler workings of power and through archival research, however, Foucault argues that the birth of the modern prison is the most intense instantiation of a relatively new type of power, “disciplinary power.” Unlike the sovereign exercise of power in public torture which punishes occasional extreme acts

266  Josh Dohmen through excessive force to terrify others into submission, disciplinary power aims to create “docile bodies” through measured punishments for minute infractions and to approximate as much as possible a 1:1 correlation of infractions to punishment. This form of power works by distributing bodies in time and space, controlling even the most minute activities, coordinating the activities of individuals into greater forces, observing individuals’ activities as constantly as possible, and judging them against norms for performance. What becomes clear through this history is that these tactics are not at all specific to prisons. Rather, disciplinary power is manifest in schools, the military, workplaces, and myriad other institutions. The rise of modern prisons, then, is not (or at least not only) the result of humanist sentiments and initiative, it is one instance of a general form of power that aims to organize masses in an era of burgeoning capitalist production and population growth. In such a setting, the spectacle of torture, public punishment, and execution became dangerous not only to those punished, but also to those in power who could be subject to the violence of the masses. In their numbers, the crowds may have felt indignation toward the punishers and revolted instead of feeling only fear of their own punishment. This helps to answer the second question posed above. We continue to seek reform of prisons for two related reasons. First, the failure of prisons is actually productive for a certain type of power, according to Foucault. By creating the figure of the delinquent—the minor criminal likely or even destined to reoffend—the legal system, police, and social sciences like psychiatry and social work generate a network of individuals who circulate through prison and the margins of society where they can be observed (through parole, for example) and used in the surveillance of others. This sets the illegalities of those in poverty apart in an attempt to divide any social forces that may coalesce against the interests of the wealthy. Meanwhile, the illegalities of the wealthy are not considered to be evidence of delinquency. Moving beyond Foucault’s own explicit analysis, I believe his genealogy makes clear a second reason that we continue seeking modest reforms: we are embedded in a network of institutions that have disciplined us from very early in our lives so that it becomes difficult to imagine an outside, to conceive of an alternative to incarceration. Think, for example, of those who follow the most ideal path from school to college, earning good grades and few infractions, and then enter the workforce with the habits and history to work up a hierarchy in their field. Then consider those who follow a less ideal path, perhaps earning poor grades or student records, maybe being sent to juvenile detention, marked in a way that makes it difficult to find stable employment or work their way up any job hierarchy. The former people are likely to conceive and imagine themselves as successful in large part because of a history of hard work or specific personality. They are likely to understand themselves in strict contrast to those who have not, at least in their terms, succeeded. The latter are likely to conceive and imagine themselves as unsuccessful because of a personal failing. Even if a person does understand the role systems of

Prison Tunnel Vision  267 discipline play in their lives, their knowledge is likely to be discounted. Think of those who say, for example, “You can make excuses all you want, but at the end of the day you have to take responsibility for your choices.” Why then are we resistant—especially those who are successful by the standards of disciplinary institutions—to thinking of alternatives to incarceration? In part because how we understand and imagine ourselves is shaped so deeply by the ways in which we have been disciplined. If I succeed because of a related series of institutions that train some to be docile bodies for productivity and discard others to divide and quell any social movements from below, this would require a radical rethinking and reimagining of who I am. If the work of disciplinary power appears necessary to me for the orderly functioning of society, then the thought of dismantling prisons may well seem to be a threat to the very order of society. In short, genealogies like Foucault’s2 help us understand the ways in which we as subjects are shaped by particular, contingent histories and specific workings of power and knowledge. This can help us understand why our subjectivities, or identities, make it difficult to question prisons or our own complicities in the prison system.

The Open, or Vulnerable, Subject In this section, I will consider what it means for subjects to be “open,” specifically from the perspective of Judith Butler’s work on vulnerability. The central idea that I wish to highlight is that we are, as subjects, importantly constituted through and open to others in ways that shape our responses to incarceration. Who I am as a subject is the result of my specific history with others. I am a result of qualities taken, mimicked, or rejected from others. I disavow qualities that I have but do not want to have or admit to, qualities that may be sources of shame. And these qualities are often projected onto others. Importantly, those identifications and disavowals are largely (if not entirely) pre-intentional; that is, we are rarely conscious of the ways we mimic, reject, disavow, or project qualities.3 Also, these identifications are shaped by the social environment. By their very proximity but also because of social pressures to do so, it is likely that one will identify to a greater extent with one’s caregivers, with one race, with one gender, with one nationality, with one set of abilities, and so on. But identifications are also fluid; they can change over time. Aside from this developmental picture, Butler notes the way in which desire reveals our openness to others: “one is undone, in the face of the other, by the touch, by the scent, by the feel, by the prospect of the touch, by the memory of the feel” (2004: 24). In our desires for others it is clear that we are incomplete, that we are affected by others and often in ways that neither party consciously intends. In loss we are also undone: If I lose you under these conditions, then I not only mourn the loss, but I become inscrutable to myself. Who “am” I, without you? When we

268  Josh Dohmen lose some of these ties by which we are constituted, we do not know who we are or what to do. On one level, I think I have lost “you” only to discover that “I” have gone missing as well. At another level, perhaps what I have lost “in” you, that for which I have no ready vocabulary, is a relationality that is composed neither exclusively of myself nor you, but is to be conceived as the tie by which those terms are differentiated and related. (2004: 24) Loss of an other reveals our constitutive relationality. I am lost in losing you; I cannot name or pinpoint what exactly I have lost in myself in losing you, revealing my opacity to myself, the limit of my self-understanding; the loss of the tie between us is, importantly, a loss of myself because I exist, at least in part, only through that tie. Indeed, this is how Butler suggests we think about successful mourning: “Perhaps, rather, one mourns when one accepts that by the loss one undergoes one will be changed, possibly for ever [sic]. . . . There is losing, we know, but there is also the transformative effect of loss, and this latter cannot be charted or planned” (2004: 21). In these ways, it becomes clear that we as subjects are vulnerable to others in a fundamental and shared way. Importantly, this vulnerability is an openness to both positive and negative, desire and aversion, building relations and losing them, support and neglect or even violence. Such vulnerability is also revealed in our embodiment. As bodies, we are seen, touched, acted upon, cared for, and subject to violence and death, just as we see, touch, care, and act, sometimes violently, through our bodies. Indeed, Butler writes, “violence is, always, an exploitation of that primary tie, that primary way in which we are, as bodies, outside ourselves for one another” (27). Any conception or image of subjects as isolated individuals, as possessing autonomy (at least as it is conceived in certain liberal and libertarian traditions) will fail to account for this constitutive relation, this primary vulnerability of each subject. So far I have summarized Butler’s claims about our shared, primary vulnerability. Some have called this “ontological vulnerability” as it is the type of vulnerability that is at the root of the types of beings we are; we are constituted by this type of vulnerability.4 But this does not mean, as Butler is fully aware, that we all experience vulnerability in the same ways. Specifically, (a) some are more likely to be subject to violence than others, (b) our responses to this shared vulnerability varies, from acceptance to complete disavowal, and (c) some lives are represented as grievable, as being worth mourning, while others are not. First, while we all share a primary vulnerability, some are subject to greater exploitation of that vulnerability based on factors like class, race, gender, age, ability or disability, and geographical location. For example, those with great wealth are just as ontologically vulnerable as those in poverty or the working class. But because access to resources is different between the groups, the latter are much more likely

Prison Tunnel Vision  269 to be exploited precisely because of and through their primary vulnerability. Similarly, those whose gender or sexuality is considered normal share this ontological vulnerability with those whose gender or sexuality is considered, for one reason or another, non-normative. Yet LGBT persons are much more likely to be victims of hate crimes (and this discrepancy also varies by race).5 This point is closely related to the second, that our responses to vulnerability vary. First, we often think of those who are less subject to exploitation or violence as less vulnerable or even invulnerable. It may be illustrative to return to the example of the vulnerability of the wealthy. If greater access to resources were taken as evidence of the invulnerability of the wealthy, such a conclusion would ignore the ways in which wealthy persons, just as those in poverty, are open to being seen, touched, and acted upon; they require support; they are potentially subject to violence. If this were not the case, why would there be security, gated communities, efforts to break labor strikes? Where did the wealth and the desire to amass wealth come from if not from others? Have the wealthy no concerns for how they are viewed or treated, and have they no desires? Why do the wealthy depend upon others for services? These vulnerabilities become viewed as normative, however, and thus they remain unquestioned such that they are not even thought of as vulnerabilities. Those vulnerabilities viewed as non-normative, however, those that are understood to make subjects dependent or violable, become extra visible. “Vulnerability” thus comes to mean “negatively vulnerable,” or “subject to the exploitation of one’s primary vulnerability.” This relates to a second point: we often disavow our own vulnerability, or take invulnerability as an ideal. Because vulnerability is widely represented as a weakness, as fragility, rather than an openness to both the positive and negative ways in which others affect us, we often deny our vulnerability altogether or seek to become invulnerable to others. Shared social representations of invulnerability as ideal lead us to disavow our own vulnerability, and our disavowal of this primary vulnerability recreates the social ideal of invulnerability. These responses and the differential experiences of exploitation and violence discussed above are mutually reinforcing. Consider, for example, the representation of the Middle East as a “war-torn region,” a description that has taken on a sort of inevitability, compared to the representation of the United States as “the greatest nation on Earth.” The events of September 11, 2001, revealed the vulnerability of the United States in a way that few other events in history have, and this opened at least two possibilities. On the one hand, we in the United States could have taken this as an opportunity to recognize our shared vulnerability with those living in Middle Eastern nations; to reflect upon the ways in which our openness to others means that we are all subject to the exploitation of that openness through violence, but that we are also all open to support and care; to form coalitions around the mourning of those we have lost and will lose, rejecting further violence to fight instead for solutions that might keep us from losing so many, from

270  Josh Dohmen mourning so often. If one wants to assume that the US is a great nation, we could ask “What violence and exploitation has led to our greatness? What role have we played in creating the Middle East and others as ‘war-torn regions’? In what ways do we remain dependent upon, vulnerable to, those in other nations?” On the other hand, we could have denied—and for the most part did deny—our vulnerability, seeking a “fantasy of mastery” by responding with violence through wars that continue to inflict loss upon those regions thought of as vulnerable, and that continue to be thought of as vulnerable in part because of the violence we inflict. This example demonstrates the ways in which the lives of some, usually those thought to be less vulnerable, are grievable, while others are not. Butler suggests that we take the obituary as a performance of grieving, as a testament to which lives are mournable and which are not. Consider, then, the coverage of the deaths of CEOs or former politicians in national newspapers, compared to the coverage of deaths in local newspapers, compared to the deaths of the impoverished or homeless whose passing is never noted in print. Or consider the rituals of mourning for dead soldiers compared to the nameless and unrepresented deaths of civilians killed in war. Butler explains two ways in which this relates to the earlier reflections on vulnerability. First, the omission of some lives from public mourning itself is a form of violence. We make the life an ungrievable life, one not worth noting, precisely through not noting it. What is this other than an exploitation of the vulnerability that each of us shares: our openness to being connected to and thus mourned by others just as others are connected to us such that we mourn their loss. Second, the failure to represent some lives as grievable leads to physical violence. Why create programs to help the poor if they never live lives worth grieving? Why not indiscriminately bomb when the lives that are lost were never really lives? The functioning of prisons and policing in the United States is, I believe, helpfully understood through this lens. First, for those who come from communities where prison is seen as an alien experience, as an impossibility, policing and prisons may well be experienced as providing a sense of security. They deny their vulnerability to violence by living in “safe,” highly policed neighborhoods or homes with security systems, and by trusting in prisons to be secure, to be guarded, to prevent escape. They deny their positive vulnerability, the openness to being affected by others in supportive or helpful ways, by believing that they remain out of prison because of individual merit or habits or by ignoring the ways in which cheap prison labor saves them money or improves their dividends. In doing so, they ignore differential policing practices, bias that assumes the vulnerability of some to criminal behavior more than others, or the others with whom they identify that have provided models of socially approved behaviors. And importantly, these denials of vulnerability reveal a constant anxiety about vulnerability. Why have a security system if you live in a safe neighborhood? Why call the cops if there are unfamiliar people

Prison Tunnel Vision  271 on your street? Because efforts at denying vulnerability amplify the fear of that which is denied—especially outsiders who are working class and/ or not white—reinforcing the perceived need for greater efforts at denying vulnerability.6 Indeed, these denials can take on odd characteristics. Consider, for example, the support of this mayor, William Dean Young, for building a prison near Newport, Arkansas, as reported in the Arkansas Democrat-Gazette: ‘We think it will be great,’ Young said, adding that residents’ fear of having prisoners nearby has diminished. ‘There should be a greater fear of the criminals that haven’t been caught than of the ones who are behind bars.’ . . . There was initially ‘some opposition’ to the prisons, but that has been overshadowed by the economic benefits they bring, he said. (George 1998) Notice what this passage accomplishes. First, the prison is represented as a reassurance of security, but one that constantly reminds locals that there remain criminals that threaten their security. Much like the ever-present threat of terrorism calls forth continuous violence to reassure some nations of their invulnerability, the prison (along with media coverage of crime, representations of criminals in popular television shows, and so on) reminds those in the surrounding community that crime is always a threat, that the prison assures them of their invulnerability, and yet that the prison is called for precisely because of the continuous threat of negative vulnerability! Second, the prison provides an assurance not just of invulnerability to violence, but also a way to avoid economic vulnerability. In the same article, other officials call the prison “a shot in the arm for every business” and “a great asset to the community” (1998). But to do so, the justifications for building the prison avoid considering the lives of those who will be imprisoned as lives at all. They are threats to those in the local community, they are commodities helping to bring jobs and revenue to the area, but they are not grievable lives. To be sure, this attitude is not universally shared, especially in those (especially non-white or working class) communities whose members face higher rates of incarceration and police surveillance. Indeed, part of the work of the Black Lives Matter movement is to reassert that the lives lost to a system of white supremacy, and specifically to anti-Black racism, are grievable. Consider this statement from the network’s website: “Rooted in the experiences of Black people in this country who actively resist our dehumanization, #BlackLivesMatter is a call to action and a response to the virulent anti-Black racism that permeates our society” (Black Lives Matter Network 2016). The claim that black lives matter is (a) a rejection of the violence done by omitting the deaths and incarceration of black persons from collective recognition and mourning and (b) aimed at ending the physical violence enacted upon those lives. Permit me to quote at length another activist

272  Josh Dohmen fighting for prison reform, M Adams from Young Gifted & Black Coalition and Freedom Inc.: So, if there was no structural racism within the jail system or the criminal justice system, then you could expect that the jail population would mirror or match the population of the county or the city or wherever that jail is based. And so, in Dane County, there’s roughly a 5 percent black population, and so we should expect that county jail should be roughly 5 percent black. But instead, the county’s jail population is roughly 50 percent black. And so, at any given time there’s about 800 folks incarcerated, and close to 400 are black. The only way we’re going to end the racial disparity or the only way we’re going to end the disproportionate amount of black people incarcerated is to immediately release 350 black people, who we know are incarcerated due to crimes of poverty, to bring the jail population of black folks down to 50 instead of keeping it at 400. You either have to free the 350 to end the racial disparities or you have to lock up close to 6,000 white people, which we are not advocating for, but 6,000 feels like a really big number to the city, but that’s how big of an impact 350 are to our community. (Goodman and González 2015) Adams, here, explicitly asserts that the loss of 350 black lives from their communities constitutes a loss on par with the loss of 6,000 lives from predominately white communities, and that those that are incarcerated are imprisoned for “crimes of poverty.” It is often a shared ontological vulnerability, an openness to the support and exploitation of others, that leads some to commit crimes as a response to poverty and systemically disproportionate access to resources. And those lives are no less lives for responding to that exploitation. What would it be to imagine, in “safe,” white communities, that it would be difficult to avoid crime if non-criminal practices continually fail to provide the support our vulnerability calls for, or if one continually faces the exploitation of her or his vulnerability in ways that others do not? Might it be possible to recognize the mourning of the lives lost to prisons, to understand that grief is shared as we have all experienced loss, and to be motivated by one’s own mourning to prevent the loss of these lives so that others need not grieve, or at least not so soon?

Epistemologies of Ignorance In this final section, I will discuss a fairly recent development in philosophy referred to as “epistemologies of ignorance.” Epistemology is the study of knowledge, so it may seem strange to discuss an epistemology of ignorance. Central to those working in the epistemologies of ignorance, however, is the belief that ignorance is produced by our knowledge-producing

Prison Tunnel Vision  273 practices; it is not simply a lack of knowledge. To help analyze ignorance, Nancy Tuana has proposed a taxonomy of social ignorance. She considers: (a) “knowing that we do not know, but not caring to know”; (b) not even knowing that we do not know; (c) not knowing because “they do not want us to know”; (d) “willful ignorance”; (e) “ignorance produced by the construction of epistemically disadvantaged identities”; and (f) “loving ignorance” (2006). There is not space, here, to explain each type of ignorance. Instead, I will focus on the three (c, d, and f above) that will be applied to issues of imprisonment below. To better understand this taxonomy, it will be helpful to summarize an example of each. I will use Tuana’s discussion of the women’s health movement. Ignorance of the side effects of female oral contraceptives exemplifies the third type of ignorance. The financial interests of pharmaceutical companies for women to take the pill, and thus to protect their profits, led them to hide these potential side effects from women. Thus, women in this situation were ignorant because they (pharmaceutical companies) didn’t want them to know. Denial of the widespread occurrence of incest provides an example of willful ignorance. When in the 1890s it became easier to detect the presence of gonorrhea, it became clear that far more girls between the ages of five and nine were infected than previously thought. Rather than acknowledging that these rates were the result of widespread incest, healthcare professionals instead revised their knowledge of the transmission of gonorrhea, attributing it to, for example, toilet seats. (At least they did so for girls from middle- and upper-class white families.) Here we see what Tuana calls a “systematic process of self-deception, a willful embrace of ignorance that infects those who are in positions of privilege, an active ignoring of the oppression of others and one’s role in that exploitation” (2006: 11). Finally, loving ignorance is the acceptance that there will be experiences that others have that we cannot share, that there will be knowledge that others have that we cannot know. In the women’s health movement, the realization by some white women that they could not fully understand the experiences of women of color did not lead to a disengagement with them, but instead served as a reminder of the differences that remain despite their shared aims. How, then, can Tuana’s analysis of ignorance help us understand our own relationships to prisons and prisoners? Consider, for example, the erasure of the knowledge of how prison labor has contributed to the United States. Alex Lichtenstein explains, for example, that “Peachtree Street and the rest paved roads and modern transportation infrastructure, of Atlanta’s well­ which helped cement its place as the commercial hub of the modern South, were originally laid by convicts” (1996: 13).7 Developing upon this insight, Angela Davis notes that visitors to historic plantations are often well aware that slave labor built the plantation and allowed it to function. But having grown up in Birmingham, even she was unaware that Alabama’s mining industry widely used prison labor during the rise of convict leasing (Davis

274  Josh Dohmen 2003: 35–6). And it would be a mistake to believe that prison labor is no longer exploited, but it is not always easy to learn who uses this labor or for what.8 Ignorance of prison labor may fit into multiple categories of Tuana’s taxonomy. For some, it may be best characterized as “not knowing because they don’t want us to know.” The “they,” here, would refer to those who have an interest in keeping others ignorant: those who write histories that leave out the contributions of prison labor; the companies, cities, or states whose current success depended upon the use of convict leasing; or the companies and government bodies that continue to use prison labor. By keeping others ignorant, it becomes difficult to question the practice or to recognize how dependent we are, and have been, upon prison labor. In turn, it is easier for these companies and government agencies to continue to profit and benefit from this labor. For others, ignorance of prison labor is likely better understood as willful ignorance. It is simply easier, or more comforting, to ignore the history and continued practice of using prison labor than to confront the ways in which we are benefited by or complicit in it. Even those who have read an article about, or heard about, prison labor may well deny the ways in which they benefit from it, fail to learn more about the current or historical use of prisoners by industry, or remain convinced that it is an isolated practice. Willful ignorance is also often at play in discussions of prison abolition. Some, when confronted with the continual failure of prisons and prison reform respond rhetorically, “What’s the alternative?” Rather than confront the injustices of prisons, or one’s role in the perpetuation of prisons—through increased calls for policing, benefiting from prison labor, or simply understanding oneself as “not a criminal” in contrast with “criminals”—it is easier, it is more comfortable, to assume that there are no alternatives. The best we can do is call for modest reforms. What, then, would loving ignorance look like in relation to prisons? It may first involve combating willful ignorance. To share in the goals of the communities who are actively resisting prisons in the United States, one must first be open to learning what those communities already know: that prisons tear loved ones from communities, that they manifest rampant racial injustice, that others comfort themselves as innocent through ignoring the uncomfortable truth that there are innocent people behind bars, that some communities feel safer because of prisons but at the expense of losses to others, and so on. At the same time, those who have not lost a loved one or community member to prison should never assume they understand the experiences of those who have. Those who come from communities where prison is seen as an impossibility should never assume they understand the experiences of those from communities where imprisonment is viewed as probable or even inevitable. But these ineliminable differences need not be a barrier to forming coalitions, to working alongside those who are already resisting contemporary prisons. Instead, they may provide the opportunity to take others’ struggles for justice as one’s own.

Prison Tunnel Vision  275

Notes 1 Angela Davis powerfully details this point in the introduction to Are Prisons Obsolete? (2003). I hope this chapter can contribute to the same project by encouraging a critical self-examination of our investments in prison, especially for those who are from communities where ending up in prison seems unlikely and for those who are made to feel safer and more secure by practices of imprisonment. 2 Foucault’s is far from the only genealogy of prisons. See, for example, Andrew Dilts’s Punishment and Inclusion (2014) for its study of the relationship of punishment, citizenship, race, and disability in the United States, and Angela Davis’s Are Prisons Obsolete? (2003) for tracing a history of the gendered, racial, and industrial influences on contemporary prison practices. 3 Thanks to Jasmine Wallace for encouraging me to emphasize this point. 4 See, for example, Erinn C. Gilson’s The Ethics of Vulnerability (2014), which has been influential in my reading of Butler, and Vulnerability (2014), a collection of essays about vulnerability edited by Catriona Mackenzie, Wendy Rogers, and Susan Dodds. 5 See, for example, the 2016 New York Times article “L.G.B.T. People Are More Likely to Be Targets of Hate Crimes Than Any Other Minority Group,” written by Haeyoun Park and Iaryna Mykhyalyshyn. 6 Thanks to Jasmine Wallace for encouraging me to emphasize this point. 7 I learned of this text from Angela Davis’s Are Prisons Obsolete? (2003), in which she quotes the same passage. 8 Davis’s (2003) chapter on the prison industrial complex reports a variety of industries that benefit from prisons and prison labor. Also, Mother Jones (Winter 2008) and Salon (Henderson 2015) have published lists of industries and companies that use prison labor. Some states not only have their own prison industries, but mandate that other state agencies buy from those industries. See, for example, the Iowa state code which mandates the purchase of products from Iowa Prison Industries (Iowa Prison Industries 2016). This is not without controversy, however. See, for example, Wiser (2014).

References Black Lives Matter Network. (2016). “About the Black Lives Matter Network.” Available at http://blacklivesmatter.com/about/, accessed August 31 Butler, Judith (2004). Precarious Life: The Powers of Mourning and Violence. New York: Verso. Davis, Angela (2003). Are Prisons Obsolete? New York: Seven Stories. Dilts, Andrew (2014). Punishment and Inclusion: Race, Membership, and the Limits of American Liberalism. New York: Fordham University Press. Foucault, Michel ([1975] 1995). Discipline and Punish: The Birth of the Prison, trans. A. Sheridan. New York: Vintage. George, Emmett (1998). “Prisons Give Newport Shot in the Arm: Two New Facilities Dip Town’s Jobless Rate, Help Local Businesses.” Arkansas Democrat-Gazette, January 12. Gilson, Erinn (2014). The Ethics of Vulnerability: A Feminist Analysis of Social Life and Practice. New York: Routledge. Goodman, Amy and González, Juan (2015). “Wisconsin Activists to Continue Protests After Cop Avoids Charges in Killing of Tony Robinson.” Democracy Now! May 13. Available at www.democracynow.org/2015/5/13/wisconsin_activists_ to_continue_protests_after

276  Josh Dohmen Henderson, Alex (2015). “9 Surprising Industries Getting Filthy Rich from Mass Incarceration.” Salon, February 22. Available at www.salon.com/2015/02/22/9_ surprising_industries_getting_filthy_rich_from_mass_incarceration_partner/ Iowa Prison Industries (2016). “Waiver Request Process and Instructions.” Available at www.iaprisonind.com/store/pg/172-Waiver-Requests.aspx, accessed August 31 Koopman, Colin (2013). Genealogy as Critique: Foucault and the Problems of Modernity. Bloomington, IN: Indiana University Press. Lichtenstein, Alex (1996). Twice the Work of Free Labor: The Political Economy of Convict Labor in the New South. New York: Verso. Mackenzie, Catriona, Rogers, Wendy, and Dodds, Susan (eds.) (2014). Vulnerability: New Essays in Ethics and Feminist Philosophy. New York: Oxford University Press. Park, Haeyoun and Mykhyalyshyn, Iaryna (2016). “L.G.B.T. People Are More Likely to Be Targets of Hate Crimes Than Any Other Minority Group.” The New York Times, June 16. Available at www.nytimes.com/interactive/2016/06/16/us/ hate-crimes-against-lgbt.html?_r=0 Tuana, Nancy (2006). “The Speculum of Ignorance: The Women’s Health Movement and Epistemologies of Ignorance.” Hypatia 21: 1–19. Winter, Caroline (2008). “What Do Prisoners Make for Victoria’s Secret?” Mother Jones. Available at www.motherjones.com/politics/2008/07/what-do-prisonersmake-victorias-secret, accessed August 31, 2016 Wiser, Mike (2014). “Iowa’s Prison Industries and the Open Market.” The Cedar Rapids Gazette, April 1. Available at www.thegazette.com/2014/03/17/iowas-prisonindustries-the-open-market Young, Iris M. (2011). Justice and the Politics of Difference. Princeton, NJ: Princeton University Press.

16 Exile as an Alternative to Incarceration Briana McGinnis

While terms like “exile” and “banishment” evoke dramatic images of regimes distant in time or character from the modern liberal democracy, penal practices that rely on the geographic removal of offenders remain with such societies in the twenty-first century. The US state of Georgia, for instance, routinely imposes sentences of banishment, and the practice also exists in Mississippi.1 Formal banishment policies, however, remain relatively rare.2 Exile offers a tantalizing alternative to incarceration: it obviates the costs of feeding, housing, and providing medical care for prisoners, it offers a comforting cushion of distance between the offending party and the community or person against whom they transgressed, and a banished offender need not suffer the hardships that accompany imprisonment in most instances. On the whole, banishment appears to offer a humane, cost-effective alternative to incarceration. Given its benefits, it appears obvious that liberal democratic societies should consider adopting banishment as an alternative to imprisonment. This chapter will evaluate that intuition from a comparative perspective, considering the cases in which formal banishment policy has been employed recently in various societies and analyzing whether the features of those policies that accommodated exile are consistent with the norms, capacities, and commitments of mass liberal democracies. Exile is best thought of as “sending from,” in contrast to “sending to.” In this, it differs from imprisonment or transportation to a penal colony. People who are now imprisoned (or who were once transported to penal colonies) are subjected to close supervision by the state. Historically, these systems focused on correcting, punishing, or in some way changing the offender. Banishment, by contrast, simply aims to solve the problems their continued presence pose to their home community—once they have been separated from the banishing polity, the relationship of control and supervision ends or is greatly attenuated. While this quality of exile makes it an appealing option for a number of practical reasons, it must be recognized that it also makes exile a qualitatively different kind of penalty from those normally employed in modern liberal democracies: the good that banishment seeks to preserve is not that of the offender, but that of the greater community. This communitarian aspect of exile is particularly prominent in the societies

278  Briana McGinnis that currently employ it or have recently done so. It is also, however, one of the greatest normative obstacles for liberal democracies considering the adoption of banishment, as the practice’s community-first ethos is not easily adapted to societies in which individual rights have primacy. Banishment, codified and openly employed, is extremely rare in the current era, even outside of liberal democracies.3 Nonetheless, it tenaciously retains its place in a few legal codes. This chapter offers a brief comparative analysis of banishment policies in use during the late twentieth century (in Singapore), those used until recent years (in the Maldives), and of those still currently in use (in various Native American communities). While these societies differ greatly from one another in their particular features, the uses to which they put exile are often similar, protecting the community’s shared identity and values, as well as preserving social cohesion. Examining the relationship between exile, shared identity, and the conditionality of membership offers an indication of what purposes exile can serve in today’s world and of what kind of society the exiling society is. Ultimately, the particular features of these latter-day banishing societies that allow them to continue justifying the use of exile present serious justificatory problems to mass liberal democratic polities that might consider emulating their examples.

The Maldives Located off the southwest coast of India, The Republic of the Maldives is an officially Islamic nation made up of 1,192 tiny islands (the great majority of which are uninhabited) spread across 90,000 square kilometers. Until 2015, exile was an integral part of the Maldives’ legal code, primarily concentrated in the Penal Code and Rules Relating to the Conduct of Judicial Proceedings, which governed most criminal matters.4 While the Maldives’ integration of Shari’a into its legal system does not conflict with the use of exile, it also does not account for it. Rather, there is a tradition of exile particular to the Maldives that has adapted itself to Islamic customs (Pyrard de Laval 2000; Ibn Battuta 1996). That tradition, I argue, is fundamentally communitarian in nature, and also built around the particular geographic features of the Maldives. As such, the sort of regularized penal banishment that prevailed in the Maldives until 2015 was both feasible and easily justified in its own context, but is not suitable for emulation by most other societies. From the arrival of the banished founding monarch until the penultimate royal ruler was exiled in 1933, exile has been a regular feature of Maldivian politics.5 By the seventeenth century, non-elite transgressors of law or custom could also find themselves exiled. By the twenty-first century, exile was fully democratized: any Maldivian faced the possibility of penal exile upon the commission of certain offenses.6 Banishment, as prescribed by the 1961 Penal Code, was a strictly intranational affair, with those exiled sent to an island upon which they did not live. Twenty-first century

Exile as an Alternative to Incarceration  279 exile in the Maldives is best understood through the lens of community: offenses that threaten social cohesion—such as those against the shared national religion, against customs governing inter-family ties, and against the public peace—were most likely to draw exile. Before the adoption of the UN-sponsored penal code, penalties were not merely graded by severity, but were tailored to particular offenses: some offenses brought exile (as an alternative to, or in addition to, other penalties), others incurred fines, flogging, imprisonment, or house arrest. Before 2015, exile appeared in a number of Maldivian laws, both civil and criminal. Additionally, the Maldivian Constitution empowered courts to adjudicate moral and religious matters as “non-statutory crimes,” in addition to those recorded in the country’s many ad hoc laws.7 Exile was prescribed in response to a range of religious transgressions, from failing to observe proper Muslim burial customs to criticizing Islam.8 These religious offenses could bring periods of exile ranging from eight months to two years and also included illicit sex, consuming alcohol, and eating or drinking during the proscribed hours of Ramadan. Homosexuality, in addition to extraor pre-marital sex, also fell into this category. Violations of sexual morality, if made public, were likely to be met with exile, in addition to other penalties like flogging and house arrest. Homosexual conduct, for instance, was punishable by banishment from one to three years.9 Maldivian family law also made use of exile. Though the Family Act covered matters related to child custody, legitimacy, inheritance, and marriage, the Act prescribed exile for only seven infractions, almost all of which concerned marriage and divorce.10 These infractions included coercion to marry, divorcing without official approval, and giving false information for the purpose of contracting or registering a marriage.11 As in many cultures, marriage customs are vital to preserving the Maldivian social fabric. However, unlike many cultures, unusually high divorce and remarriage rates are one mechanism by which that social fabric is reinforced in the Maldives. What is often framed in Western reports and UN studies as a “crisis” of “skyrocketing” divorce rates is, in fact, an old custom in the scattered islands. Maldivians who threatened the system by bringing deception into it or failing to meet traditional obligations found themselves liable to exile under the pre-2015 law. Most codified exile-able offenses were classified as criminal, though civil offenses were also punishable by exile. The Penal Code imposed terms of exile for a variety of transgressions, including acts forbidden by the Quran, those categorized as “Acts against the State” and those labeled, simply, “Disharmony.”12 The code prescribed banishment for threats to public order (like assaulting a public figure or fomenting rebellion), for ordinary offenses like theft and assault, and also for a number of infractions related to refusing to aid authorities in the discipline of other citizens.13 Similarly, many infractions for which exile was imposed centered upon preserving trust within the community. The Penal Code prescribed exile among the

280  Briana McGinnis penalties for crimes of fraud or deception, including giving false evidence against another individual, counterfeiting, the sale of adulterated goods, “criminal breach of trust,” “cheating,” and defamation.14 Finally, transgressions against the cohesion of the community brought banishment. Perhaps the most striking example of exile being used to preserve social cohesion is found in the Maldives’ response to Islamic extremism. Traditionally, the Maldives has been a moderate Islamic nation. However, since tourism became the nation’s primary industry in the 1970s, tensions have arisen between the nation’s Islamic identity and its economic needs.15 The Religious Unity Act has been in effect since 1994, but the increasing popularity of fundamentalist preachers prompted the Islamic Ministry to add the new Regulation on Protection of Religious Unity Among Maldivian Citizens to the Act in 2011.16 Interestingly, neither regulation focuses primarily on matters of doctrine—in fact, they address no specific theological issues. While the regulations do limit the discretion of speakers to propagate unorthodox interpretations of sacred texts or represent them as widely accepted teachings, most of the provisions prohibit speech that is likely to incite public conflict, including one provision requiring that speakers “. . . Not engage in any talk that may create hatred and anger among the people, nor disseminate any information that incites to violence of any kind.”17 Failure to abide by the regulations was punishable with sentences of two to five years’ exile, in addition to possible imprisonment. This survey of the role of banishment in pre-2015 Maldivian law illustrates that it was not regarded simply as one penalty among many. Rather, Maldivian penal exile reinforced a thick sense of shared communal identity and values. The example of the Maldives suggests that exile is at home in a society characterized by a high degree of consensus about public morality and a strong sense of unity (or what Durkheim would term “mechanical solidarity”). In the Maldives, banishment was a special kind of penalty—one linked to crimes threatening to destabilize or undermine that public consensus. Creating distance between the offender and the small, tight-knit community against which they transgressed, in this context, offered a sensible response. In a pluralistic liberal democracy—characterized by deep disagreement about any possible common good—penal practices designed to preserve a single shared identity are unjustifiable. In sum, the case of Maldivian banishment policy before 2015 illustrates that banishment need not always be the tool of authoritarian or arbitrary government, but can serve legitimate communitarian ends. It does not, however, offer a model for large, pluralistic liberal democracies to emulate.18

Exile in “Indian Country”19 Native American nations in the US inhabit a unique position in that they are surrounded (both geographically and culturally) by a powerful, hegemonic liberal nation.20 Because of their peculiar status as “domestic dependent

Exile as an Alternative to Incarceration  281 nation[s],” American Indian communities retain only limited rights of sovereignty. Unfortunately, US Supreme Court rulings over the past three decades have further eroded what sovereignty remains in Indigenous hands. Currently, native communities in the United States wield broad control over their territories, yet are denied jurisdiction over non-Indians on Indian land.21 Additionally, the sentences that tribal courts are allowed to impose are severely limited by the Indian Civil Rights Act and the Major Crimes Act.22 Indigenous communities do, however, largely retain the right to control access to their territories, and while they have limited freedom to govern non-Indians on their lands, they do have the right to expel them.23 Since the US Supreme Court acknowledged the right to control access to Indian land as one of the inherent powers of sovereignty in 1982, it has become common for the laws of Indigenous communities to include provisions for the expulsion of non-members.24 The removal of non-members is not exactly “exile,” so this section will examine only laws that allow for the expulsion of existing members and cases in which those laws have been applied.25 Non-member exclusion ordinances have been widespread since the 1980s, but it is only relatively recently that laws providing for the expulsion of members have gained popularity. The US federal government recognizes more than 500 American Indian nations, each of which is entitled to draft its own constitution and laws, subject to limitation by the US Constitution and federal law.26 Not surprisingly, there is great variation among legal codes; each community that has introduced banishment provisions has its own particular practices. Terms of exile last from months to years, even in rare cases for life. A few exclusion provisions also strip membership from the exiled.27 The most extreme language actually pronounces the banished member to be “legally dead.”28 However, what each of today’s varied codes has in common is that they allow indigenous communities to expel members who fail to meet shared standards of behavior or who threaten the stability of the community. The text of today’s exclusion ordinances and the scholarly debate surrounding them often reference the history of exile in various nations’ traditions.29 There is, for instance, evidence that the nations of the Iroquois Confederacy employed exile centuries ago (Cousins 2004; Johansen 2007: 186–195; Kunesh 2007: 85–145; Wilkins 2004: 235–262). Additionally, the Cheyenne, Navajo, and Cherokee nations all have documented histories of exile dating back centuries, as do several others (Kunesh 2007: 91–100; Deloria and Lyttle 2010: 190, 113; Nickerson Llewellyn and Adamson Hoebel 1941: 85–87, 133–137, 143, 151, 161; Strickland 1982). By all accounts, banishment was very rare, employed when all other remedies were exhausted or in the face of an exceptionally reprehensible offense. Today, with their formal options for maintaining internal social control limited, several indigenous communities have turned to banishment to combat problems plaguing them. For some of the groups now using exile, the practice has deep historical roots; for others, it is new.30 Debates over the historicity

282  Briana McGinnis of banishment in American Indian communities approach exile as an atavism. However, exile need not have deep historical roots to claim legitimacy as a penal practice; it needs only to be consistent with the commitments of the community employing it. A small minority of member exclusion codes include banishment as one option, along with prison and fines, for virtually all crimes, but these are exceptions.31 Most banishment ordinances address a select few offenses, with the penalty tailored to the transgression. Exile provisions target antisocial behavior that threatens the stability of the community or displays of disrespect for the group’s shared culture and traditions. They also target individuals whose repeated offenses suggest incorrigibility or an unwillingness to make peace with the community. Often, exclusion codes employ language of collective self-defense.32 Scholars of American Indian law suggest that traditional notions of justice and crime often view breaches of tribal law not only as an injury to the immediate victim, but as a wrong done to the entire community (Zion 1985: 89–109, 1998: 58–74; Yazzie 1996: 157–173). Emphasis in many tribal justice systems has traditionally been upon restorative justice, generally entailing reparations, mediation, and community efforts to correct and reintegrate the offender (Archambeault 2009; Austin 2009; McNamara 1993; Meyer 2009: 176–190; Monture-Okanee and Turpel 1992; Riley 2007). While there are certainly expressive and disciplinary aspects to exile in Indigenous communities, the uses to which American Indian nations now put it are largely defensive. After centuries of fighting to preserve besieged identities, Indigenous polities now find both their members and the communities themselves threatened by social problems like drug addiction and gang violence. Their response, though painful, has been to excise the offending parts of the community in an effort to prevent further damage to the whole. In recent years, two of the greatest threats to Indigenous communities have been increases in drug use and trade and gang violence.33 Indeed, drug violations are among the most common enumerated grounds for expulsion. The Lummi, based in Washington, have suffered enormously from the illegal trade in methamphetamine and OxyContin, with tribal authorities estimating that nearly one quarter of the small community of 2,000 members was addicted to painkillers or other controlled substances in 2004. To reduce the damage, the Lummi nation adopted its own exclusion code in 2008, with the manufacture and distribution of illegal drugs listed among the grounds for exclusion.34 In the Midwest, the Lac du Flambeau Band of Lake Superior Chippewa, the Mille Lacs Band of Ojibwe, the Red Lake Band of Ojibwe, and the Upper Sioux Community have adopted similar codes to combat growing drug problems as well.35 Exclusion ordinances have become one of the few tools native communities have for combating the rapidly growing problem of gangs operating on their lands. Though few banishment laws target gang activity specifically, broad provisions in constitutions and legal codes give Indigenous communities’

Exile as an Alternative to Incarceration  283 authorities the power to exile members whose violent behavior or criminal actions strain community ties.36 American Indian communities located near metro areas, especially, suffer from the predations of rapidly growing gangs like “the Native Mob,” “Sovereign Nation Warriors,” and “Native Soldiers,” in addition to similar problems with nationally established gangs (Austin 2013; Steil 2006). The Mille Lacs Band of Ojibwe, located outside of the Minneapolis-St. Paul metro area, has taken to banishing gang members in an attempt to prevent further entrenchment in their territory.37 Those recently banished from Mille Lacs lands are repeat offenders who have not responded to imprisonment or the imposition of fines. The use of exile in these cases is not merely a matter of convenience. Rather, the imposition of exile in tightly knit indigenous communities is a method of protecting a particular shared way of life. These exiles do not merely damage themselves and their immediate victims; widespread drug use damages relationships in families and between neighbors, and the demands of gang membership compete with those of community membership. Casting drug dealers and gang members out is not intended merely as a punitive measure, nor one of instrumental convenience, but is a necessary step in protecting tightly knit communities. Actions that insult cultural or religious identity frequently appear in exclusion codes. This includes the abuse of tribal lands, damaging cultural or religious artifacts, or disrupting community ceremonies.38 Abuses of land access privileges, including environmental damage and unauthorized exploitation of natural resources, constitute grounds for exclusion in a number of codes. For instance, of the nine grounds for exclusion listed in the Yankton Sioux Exclusion and Removal Code, four of them concern the unauthorized taking of shared resources like timber or game from the land, or searching for natural resources within the Reservation without permission.39 The Ho-Chunk Nation includes “[d]isturbing or excavating items, sites, or locations of religious, historic, [or] cultural . . . significance without the authority of the Nation. . . .” among its grounds for exclusion, and the Siletz Tribal Code prescribes banishment in its “Cultural Resource Lands and Sacred Sites Ordinance,” though it does not for any other offense.40 Finally, American Indian communities use exile to permanently remove members who have, through an established pattern of bad behavior or for particularly egregious offenses, shown themselves unwilling to act as peaceful, cooperative members of their home communities. Permanent banishment is rare in the various codes, but when it does appear it is for offenses like arson resulting in death, premeditated murder, or especially reprehensible sex offenses.41 Sex offenders are often subject to exclusion, even in the codes of nations that do not otherwise employ exile. In addition to those who have committed especially harmful offenses, habitual troublemakers are common targets for indefinite exclusion.42 For example, the Osage Nation Code (one of the most comprehensive) prescribes terms of exile up to life for repeat offenders in a number of realms, including

284  Briana McGinnis malfeasance on the part of public officials or those who have committed various forms of fraud.43 Every instance of exile has two aspects: what the event means for the exiled, and what it means for the exiling community. The Exclusion Code of the Lummi Nation acknowledges both facets: (b) The fundamental purposes of this Exclusion Code are two-fold: the first is to protect the health, safety, and welfare of the Lummi Nation and the Lummi Reservation community; the second is to provide the person to be excluded with the motivation and means to seek treatment and rehabilitation so that their conduct may no longer be a threat to the health, safety, and welfare of the Lummi Nation.44 Tribal authorities routinely comment on how difficult and emotionally taxing the decisions are (Steil 2006; Shukovsky 2003; Kershaw and Davey 2004). Most often, Indigenous communities use banishment as a protective measure, limiting the damage a disruptive member might cause and creating the space between an offender and those he or she has victimized necessary for damaged relationships to heal. Banishment for specified terms holds out the hope, as expressed in the Lummi Exclusion Code, that those exiled will be able to shed their harmful ways and rejoin their fellows. Exiling members of gangs, drug addicts and dealers, and those who show disrespect for cultural artifacts or practices for limited terms suggests that these behaviors are viewed as remediable—but also incompatible with the character of the community. American Indian communities, as the Maldives did until recently, use penal banishment to preserve a tightly knit community predicated upon a shared vision of the common good. Indigenous communities resort to exile, in part, because their options for imposing other penalties are subject to constraints imposed by the US federal government. Generally, the many and varied Indigenous societies that use exile use it to preserve communities threatened from without by the encroachments of their “host” nation and from within by social problems that weaken the fabric of these traditionally cohesive communities. As in the case of the Maldives, American Indian communities employ exile as a response to particular conditions—relatively small and cohesive populations, legal access to the territory outside of the communities’ own lands, and thick cultural and moral identities. The conditions that make banishment a good “fit” for American Indian communities also raise doubts that similar models could be adopted by pluralistic mass democracies.

Singapore This chapter will only very briefly address Singapore’s preservation of its Banishment Act, as its employment has often been nakedly political rather than penal.45 While, admittedly, the line separating the two categories is not

Exile as an Alternative to Incarceration  285 always entirely clear, Singapore’s particular history as a “soft authoritarian” state and an “illiberal democracy” merits giving the nature of banishment in the small nation a careful examination.46 Singapore is, perhaps, best considered a cautionary example to any liberal democracy considering the adoption of banishment legislation—the temptation to use it to punish dissent or silence political opponents may be irresistible to some authorities. Nonetheless, this chapter examines recent uses of banishment, and Singaporean authorities made extensive use of banishment legislation over the course of the twentieth century. Singapore has long employed communitarian language in its public discourse, so it may seem unsurprising that it, like the other polities considered in this chapter, would also put exile to community-preserving ends (Chua 1995). Singapore has a long history of banishment law reaching back to its founding as a trading colony in 1822 (Raffles 1968). Banishing individuals for committing simple criminal offenses was not uncommon in Singapore’s history; however, over the course of the twentieth century, mass banishments of political protestors occurred with some regularity, along with targeted banishments of dissident leaders.47 For instance, the twentieth century opened with the banishment of some 400 “undesirables,“ including Chinese nationalist activists between 1912 and 1916 (Ching and MacKenna 1990: 56–57), and 1,500 were banished in the wake of upheaval caused by the global economic depression between 1928 and 1931 (Young and MacKenna 1990: 57). Mass banishments peaked in the wake of postwar privations, with a series of large-scale strikes occurring in 1946 and 1947. A state of emergency was declared in 1948, followed by the banishment of 13,000 trade unionists and activists (Tremewan 1994: 15). Political expulsions continued through the 1940s and into the 1960s, when Singapore became independent and the new People’s Action Party (effectively the sole party in Singaporean politics for decades) emerged. While the number of banishments declined, the 1960s saw some of the first high-profile PAP-backed political banishments. These included PAP dissidents, labor leaders, and would-be opposition party leaders.48 PAP co-founder Lee Kuan Yew was elected Singapore’s first Prime Minister in 1959, and later that year the most recent iteration of Singaporean banishment legislation was enacted: the 1959 Banishment Act, which remains in effect. The Banishment Act is no relic of an earlier age. Far from being declared obsolete and repealed, as Malaysia’s similar Act was in 2011, Singapore’s Banishment Act was amended in the same year to bring it in line with new due process requirements, reaffirming the city-state’s commitment to exile as a tool of governance. The Act empowers the Minister of Home Affairs to order the banishment of any non-citizen whose presence he deems not to be “conducive to the good of Singapore.”49 The language of the Act, notably, does not regard the security or sovereignty of the city-state, but the “good.” Its phrasing assumes that some self-evident “good of Singapore” exists, and also that banishment is suited to preserving it. The Banishment Act is seldom used at present. The

286  Briana McGinnis best-publicized case of banishment under the Act in recent years was that of Malaysian-born pilot Ryan Goh. Goh’s union activities within Singapore ultimately culminated in a confrontation with no less than Minister Mentor Lee himself in February of 2004 (Doran 2004). Yew accused Goh of “stirring up trouble” with his leadership of a divisive vote to dismiss members of the pilots’ union. Two days after his acrimonious meeting with Lee, Goh received an order requiring him to leave Singapore by May 1, 2004 (Doran 2004). Banishment in Singapore is illustrative of the dangers accompanying the adoption of exile as a penal practice. Despite its professed communitarian ethos, the government of Singapore has long used banishment not as an impartial penal practice, but as a tool of partisan (sometimes arbitrary) political power.

Conclusion The cases examined in this chapter demonstrate that there are legitimate, internally justifiable uses for banishment in some types of society, but they also illustrate the ease with which banishment can be abused and the close link between a non-liberal, communitarian orientation and the use of exile as a tool of social control. The aspects of these societies that allowed them to accommodate banishment even into the twenty-first century suggest that exile is not a feasible or justifiable option for mass liberal democracies. In addition to the problems that the territorially bounded nation-state faces in seeking to banish transgressors, the greater problem of justifying the expulsion of offenders remains. Liberal policies are characterized by a commitment to pluralism—the thick, unified identities and value codes that exile protects in the Maldives and in Indigenous communities are not only absent, but incompatible with the principles underlying liberal societies.

Notes 1 The state of Georgia produces an enormous number of banishment cases each year, but the US Supreme Court declined to hear a 2008 case that raised the question of intrastate banishment’s constitutionality. Brief of Respondent-Appellee at 24, Terry v. Hamrick (2008). A single Georgia county (Houston County) has banished more than 500 people since 1998. For a full account of banishment in Georgia, see Carpino 2011. 2 While an extended discussion of expulsion of the nation is beyond the scope of this chapter, there is a case to be made that targeted deportation and the criminalization of migrants constitutes a contemporary form of banishment. However, this chapter examines only regularized, formal versions of national banishment that are openly practiced and justified by the regimes employing them. 3 The officially acknowledged use of banishment remains rare. I give the examples of Singapore, the Maldives, and certain Indigenous communities because they all have codified banishment as one of the acceptable tools of government. Additionally, all have records of actually employing it with some regularity. Human Rights Watch has noted isolated incidents of banishment in Tunisia, the

Exile as an Alternative to Incarceration  287 Democratic Republic of Congo, and Turkmenistan (Human Rights Watch 2000, 2005, 2009). Under Kim Jong-Il, banishment was among the tools used by the executive to remove political opposition (Choe 2010). Similarly, Haiti and Uruguay both have legal provisions for banishment in their penal codes, but there is no evidence of their regular use. 4 When the Maldives finally adopted the Draft Maldivian Penal Code, drafted at the behest of the United Nations Development Programme in collaboration with a team from the University of Pennsylvania law school, banishment was ultimately omitted from the code, signaling the end of a tradition that had been with the Maldives for centuries. Until 2015, exile appeared in a number of Maldivian laws, both civil and criminal, including the Penal Code, Act No. 1 of 1961, 32, 34–5, 37, 39–41, 48–50, 53–4, 58, 63–6, 69–71, 73–4, 76–88, 91–9, 111–18, 120, 126–41, 143, 146, 163; Family Act, Act No. 4 of 2000, 63–4, 66–70; Prevention of Terrorism Act, Act No. 10 of 1990, 6(a-b); Law on Narcotic Drugs and Psychotropic Substances, Act No. 17 of 1977, 1(c), 3(d), 4(e-f), 5(b), 6(c), 7(b), 16(b); Prevention and Prohibition of Corruption Act, Act No. 2 of 2000, 2(b), 3(b), 4(b), 5(b), 6(b), 7(b), 8(b), 9(b), 12(b), 13(b), 14(d), 15(c), 16(d), 17(b), 18(c), 20(b), 21(b), 22(b); Act No. 6 of 1994. For a history of the use of exile in the Maldives, see Bell 1940, 2013: 38. 5 H.C.P. Bell even went so far as to compile an index of exiled rulers to help readers of Pyrard’s account sort out actors in the various intrigues, so common was the phenomenon in the early modern era (Pyrard de Laval 2000: 493–508). Mohamed Shamsuddine III was deposed in 1933. 6 The new Maldivian Penal Code folded these disparate acts into a single, codified criminal system, with graded penalties, rather than having penalties specifically tailored to each offense. 7 These unwritten crimes have been abolished upon the acceptance of the DMPC (Robinson et al. 2006: 13, Section 12). 8 Penal Code, 88; Law on Bathing, Shrouding, and Burying the Dead. 9 Rules Relating to the Conduct of Judicial Proceedings 173(8a). For examples, see Ahmed Naish 2009 and Lubna 2012. Rules Relating to the Conduct of Judicial Proceedings, 173. Possession of pornographic materials, sexual harassment, and conducting a sexual relationship with a minor were also among the reasons for which Maldivians have been exiled in recent years. See also Haveeru Daily, October 10, 2002; Lubna 2011; Nazeer 2013. 10 Family Act. Act No. 4 of 2000. 11 Family Act, sections 63, 64, 66–70. United Nations Department of Economic and Social Affairs, Population Division, “World Marriage Data 2008.” The divorce rate has been steadily dropping. As of 2005, the crude divorce rate stood at 5.9 divorces per 1000 people per year. 12 “Exile” was defined in the previous Maldives Penal Code as follows: “ ‘Exile’ shall mean confinement of a person to an inhabited island other than the island in which he is domiciled for a certain period of time while restricting his movement to any other inhabited place and his personal freedom or part thereof.” Receiving islands generally had few employment opportunities. The longest term of exile was set at twenty-five years, which was used as the equivalent of a “lifetime” term, according to section 16 of the Penal Code. 13 Threats to public order: 30–2, 34–5, 37–8, 40–1, 48–50, 53–4, 58–9; theft and assault: 126–34, 136–41, 143, 146; discipline of peers: 76–8, 83, 85–7. Nearly every offense listed in the Prevention and Prohibition of Corruption Act included exile among the penalties. 14 The code did not define the terms “criminal breach of trust” or “cheating.” 66–7, 91–6, 103–7, 131(a), 143–6, 163.

288  Briana McGinnis 15 Maloney 1976, 2013. For cases in which such extremists were banished, see Haveeru Daily, February 9, 2008, and November 8, 2009. See also Makan 2007(a) and 2007(b); BBC News, October 8, 2007. 16 Religious Unity Act, Act No. 6 of 1994. 17 Regulation No. R-40 of 2011, “Regulation on Protection of Religious Unity Among Maldivian Citizens,” 5(f). 18 Further, the availability of hundreds of uninhabited or sparsely inhabited islands within the Maldives made intranational banishment feasible in a way it could not be absent that particular geography. 19 18 USC § 1151 (2006) defines “Indian Country” as including reservations, communities, and allotments. 20 This chapter primarily addresses Indigenous communities in the United States. For a detailed analysis of the position of Indian nations in the US, see Aleinikoff 2002. Will Kymlicka notes similar struggles of First Nations within Canada, though with significantly different legal relations (Kymlicka 2002: 12–13,17, 22, 29–32, 39–40, 44, 46, 58–9, 65, 79, 116–20, 132–3, 142–5, 179, 190–1). 21 Jurisdiction in American Indian law is complicated. Title IX of the 2013 Reauthorization of the Violence Against Women Act authorizes “special domestic violence criminal jurisdiction” allowing tribal authorities to prosecute non-Indians on Indian land for certain defined acts of domestic violence, but generally, Indian nations have no jurisdiction over non-Indians. See Oliphant v. Suquamish Tribe, 435 U.S. 191, 212 (1978), Montana v. United States 450 U.S. 544 (1981), and Strate v. A-1 Contractors, 520 U.S. 438 (1997). They do, however, retain criminal jurisdiction over non-member Indians under 25 U.S.C. § 1301 (2006). 22 ICRA, 25 U.S.C. §1302. The limits are somewhat expanded under the 2010 Tribal Law and Order Act, H.R. 725. Under certain circumstances, tribal courts can now impose sentences of up to three years, rather than one. 23 There are some exceptions specified in federal law. 24 Merrion v. Jicarilla Apache Tribe, 455 U.S. 130, 144–5 (1982). Exclusion laws are numerous and varied, but this sampling gives a sense of a typical exclusion provision: Colorado River Indian Tribes’ Law and Order Code Article V and Constitution Article VI, Section 1J; Fort McDowell Yavapai Community’s Law and Order Code, Chapter 15 Exclusion Ordinance; Swinomish Exclusion and Expulsion Code Title 3.5, Makah Law and Order Code, Title 9, Title 6; Nisqually Tribal Code Title 46, Oglala Sioux Ordinance On Tribal Removal And Exclusion Of Nonmembers From the Pine Ridge Reservation, San Ildefonso Pueblo Code, Title VIII. 25 While disenrollment remains a live issue in American Indian membership debates, this chapter focuses not on the revocation of formal membership, but on spatial expulsion—a related, but separate, matter. As such, disenrollment is beyond this chapter’s purview. 26 Nations are entitled, but not required, to draft constitutions under the Indian Reorganization Act, 25 USC § 476 (1934). 27 The Absentee Shawnee Tribe of Indians of Oklahoma, the Cherokee Nation, and the Skokomish Nation remove exiles from their membership rolls for the time of their banishment, which may be up to life. Other nations, like the Osage, do not actually revoke membership from exiles, but they do suspend their financial benefits. Absentee Shawnee Tribe of Indians of Oklahoma Tribal Code Criminal Procedure, Section 404; Code of the Eastern Band of Cherokee Indians, 2–6; Osage Nation Code, 6 § 4–135. 28 The Absentee Shawnee Tribe of Indians Tribal Code, Chapter 4, Section 404 reads in part (emphasis added): “Sentence of Banishment: (a) Banishment Defined. Banishment is the traditional and customary sentence imposed by the Tribe for offenders who have been convicted of offenses which violate the basic rights to

Exile as an Alternative to Incarceration  289 life, liberty, and property of the community and whose violation is a gross violation of the peace and safety of the Tribe requiring the person to be totally expelled for the protection of the community. During the term of banishment, a person who is banished from the territory and association of the Tribe shall: (1) Be considered legally dead and a nonentity with no civil rights to engage in contracts or come before the courts of the Tribe for any reason not related to the original conviction, provided, that the banished person retains all rights of a criminal defendant during any prosecution for an offense during the term of banishment, and while attending or going directly to or from any Court, or a proceeding involving a criminal action to which he is a party including the appeal of his case. (2) Be expelled from the jurisdiction of the Tribe and not be allowed to return for any reason during the period of banishment except when required to attend court.” 29 For a sample of contributions to this debate, see Cousins 2004; Johansen 2007; Kunesh 2007; Wilkins 2004. 30 The question of whether banishment was a traditional practice became a matter of contention in the first highly publicized Indian banishment of recent decades, in which two Tlingit teenagers robbed and badly beat a pizza delivery driver in 1993 (Balzar 1994; Egan 1994; Johansen 2007: 186–95). Historicizing rhetoric also played a role in an important tribal banishment case, Poodry v. Tonawanda Band of Seneca Indians (2d Cir. 1996) 85 F.3d 874, 889. 31 The voluminous Osage Nation Code falls into this category, as does the Law and Order Code of the Pawnee Tribe of Oklahoma, Title VI. 32 See, for instance, the Sault Sainte Marie Tribe of Chippewa Indians Tribal Code, 61.101. 33 Methamphetamine, heroin, cocaine, and other drugs have ravaged many Indian communities (Shukovsky 2003; Kershaw and Davey 2004; Austin 2013; Brown 2008; Steil 2006). 34 Eugenia Phair and seven adult members of one family were banished from Lummi lands for drug trafficking after an incident that resulted in the death of an infant. Kay Commodore and John Jefferson were not only banished for selling illegal drugs, but were also disenrolled (Johansen 2007: 186–95; Shukovsky 2003; Kershaw and Davey 2004; Austin 2013). Title 12, Lummi Nation Code of Laws. 35 Red Lake Nation Prohibited Drug Crimes Code, 2004; see also Johansen 2007 186–95; Kershaw and Davey 2004; Austin 2013; Brown 2008; Steil 2006. 36 One exception to this generalization is the Lac du Flambeau Band of Lake Superior Chippewa. Their Law and Order Code does include explicit prohibitions of gang activity, though it does not name banishment among the penalties (Tribal Code 70, Title VI). The recent exile of dangerous gang members has been carried out under broad provisions in the Band’s constitution (Constitution and Bylaws of the Lac du Flambeau Band of Lake Superior Chippewa Indians of Wisconsin, Article VI, l,n,q,u.). Many codes include the power to implement exclusion ordinances among the enumerated powers of tribal councils or tribal courts, or have vaguely worded provisions that give tribal authorities broad powers, including those to draft and impose exclusion ordinances. Among those tribes are the Confederated Salish and Kootenai Tribes (Laws of the Confederated Salish and Kootenai Tribes, Codified. Article VI), the Skokomish Nation (Civil Exclusion and Removal Ordinance, Article III, Section 8(c) and Article V, Section 3 “reserve to the Skokomish General Council the power to establish procedures for the exclusion of enrolled tribal members from the Reservation”), the Hoopa Valley Tribe (Constitution, Article 9, 1.1), The Yankton Sioux (Exclusion and Removal Code, section H forbids, “Doing or threatening to do any act upon the Reservation which seriously threatens the peace, health, safety, morals or general welfare of the Tribe, its members, or other persons living on the Reservation. . .”), and

290  Briana McGinnis the Lower Sioux Community of Mdewakanton Indians, (Constitution, Article V, 1i,j). The same provisions, if not in conflict with constitutional or other explicit legal protections, also give tribal authorities the discretion to impose disenrollment along with banishment, even if loss of membership is not part of the banishment ordinance. 37 Mannix 2011. Benjamin Garbow, Patrick Provo Jr., (prosecuted for two assaults prior to being banished), Zachary Nayquonabe (previously prosecuted for DWI, theft, assault with a deadly weapon), Darrick Williams, (who had prior convictions for multiple counts of assault), and Blaine Beaulieu, (previously convicted for two DWI’s, and eight assaults) are a few of the gang members recently exiled from Mille Lacs land. On exclusion for drug crimes, see Mille Lacs Band Statutes, Annotated. Title 23, 2 § 201. 38 See, for instance, the Yankton Sioux Tribal Criminal Code, 10-1-3(E). 39 Ibid. 40 Ho-Chunk Nation Code, 3 HCC § 10; Cultural Resource Lands and Sacred Sites Ordinance, Siletz Tribal Code § 9.001; The Confederated Tribes of the Colville Reservation Tribal Code, Chapter 9–4 “Interference With Tribal And Indian Property.” Constitution, Article V (d) includes similar provisions. 41 Pawnee Tribe of Oklahoma Indians, Law and Order Code VI.1.101; Osage Nation Code: 6 ONC § 1-165C, § 2-111B, § 2-120B, § 2-130B, § 2-135A(4), § 2-137B, Lummi Nation Exclusion Code, 12.02.020b,c. 42 See, for instance, Ordinance No. 2010-002 Ordinance Of The Tribal Business Council 43 6 ONC §3-104C, § 4-107A(12), §4-111A(2), §4-112A(3), §4-151B. Ysleta Del Sur Pueblo Tribal Resolution TC-99-99, Pertaining to Amendments of Article 4 of the tribe’s Code of Laws Entitled “Peace Code,” also includes possible lifetime banishment for repeated serious, or “Class A” civil infractions, Section 4.2.40. 44 Lummi Nation Code of Laws, 12.01.010. 45 Banishment Act of 1959 (Chapter 18). The most relevant passage (5.1) reads as follows: “Where the Minister is satisfied after such inquiry or on such written information as he may consider necessary or sufficient that the banishment from Singapore of any person not being a citizen of Singapore or an exempted person would be conducive to the good of Singapore, the Minister may make an order that the person be banished from Singapore either for the term of his natural life or for such other term as may be specified in such order.” Continuing a tradition established in nineteenth-century banishment legislation, citizens of Singapore cannot be banished. In fact, Singaporean citizens enjoy explicit constitutional protection against banishment, although denationalizing a citizen is a relatively easy matter. Constitution of the Republic of Singapore, Part IV, 13.1. 46 On Singapore’s authoritarianism, Lee deputy Dr. Goh Keng Swee opines, “It is widely believed among Western observers that ours is an authoritarian political system. No one likes to be called authoritarian but it is no use pretending that this assessment does not exist.” Quoted in Mutalib 2000: 319. Originally from Goh’s speech delivered at seminar on “Role of Residents’ Committees in promoting courtesy,” Tanjong Community Center, Singapore. June 28th, 1980. 47 Banishment for simple criminal offenses was not uncommon in Singapore’s history prior to independence. Some cases include: Rex v. Adam Sing (02 Ky 12, 1822); Regina v. Tan Ah Tong (02 Ky 117, 1884); In re: Tan Tua Kow (04 KY 291, 1887); Regina v. Lin Chee (IV SLJ 69, 1891); Ex Parte Tan Kheng Long (1963 03 MC 205); (Yap Tan Lim v. Rex Singapore Criminal Appeal No. 22 of 1930). 48 Labor union organizers Datuk Dominic Puthucheary and Lim Chin Siong were among the founding members of the PAP, but their leftist commitments ultimately led them to found an opposition party, Barisan Sosialis. Both were arrested in

Exile as an Alternative to Incarceration  291 the PAP’s “Operation Coldstore” anti-communist raids. Puthucheary served ten months in prison prior to his banishment. Lim was even less fortunate; after six years’ imprisonment, Lim was finally exiled amid accusations of communism in 1969 (Hong and Huang 2008: 137–9). 49 The most relevant passage (5.1) reads as follows: “Where the Minister is satisfied after such inquiry or on such written information as he may consider necessary or sufficient that the banishment from Singapore of any person not being a citizen of Singapore or an exempted person would be conducive to the good of Singapore, the Minister may make an order that the person be banished from Singapore either for the term of his natural life or for such other term as may be specified in such order.”

References Aleinikoff, T. Alexander (2002). Semblances of Sovereignty: The Constitution, the State, and American Citizenship. Cambridge, MA: Harvard University Press. Archambeault, William (2009). “The Search for the Silver Arrow: Assessing Tribal-Based Healing Traditions and Ceremonies in Indian Country Corrections.” In Retrieving the Balance: Crime and Justice for Native Americans, ed. M. Nielson and R. Silverman. Tucson, AZ: University of Arizona Press, 143–60. Austin, Brenda (2013). “Synthetic Pot, Bath Salts and Gangs Have Native Community in Crisis.” Indian Country News, August 21. Austin, Raymond (2009). Navajo Courts and Common Law. Minneapolis, MN: University of Minnesota Press. Balzar, John (1994). “Two Alaska Indian Youths Banished to Islands for Robbery.” Los Angeles Times, July 15. BBC News (2007). “Militants’ Arrested in Maldives.” October 8. Bell, H.C.P. (1940). Monograph on the History, Archaeology and Epigraphy of Maldives. Colombo: Ceylon Government Press. Bell, H.C.P. (1998). Excerpta Maldiviana. New Delhi: Asian Educational Services Press. Brown, C. (2008). “Age-Old Tribal Practice Battles a Modern Threat.” Minneapolis Star-Tribune, October 5. Carpino, Curt (2011). “Banishment in Georgia: A New Approach to Domestic Violence.” Georgia State Law Review 27: 802–28. Carpino, Cameron (2011). “Banishment in Georgia: A New Approach to Domestic Violence.” Georgia State Law Review 27: 802–28. Choe, Sang-Hun (2010). “North Korea Reinstates Official Banished for Market-Oriented Reforms.” New York Times, August 24. Chua, Being Hurt (1995). Communitarian Ideology and Democracy in Singapore. London: Routledge. Cousins, Michael (2004). “Aboriginal Justice: A Haudenosaunee Approach.” Native Law Centre 9: 141–59. Deloria, Vine Junior and Lyttle, Clifford (2010). American Indians, American Justice. Austin, TX: University of Texas Press. Doran, D’Arcy. (2004). “Banished Pilot Who Enraged Singapore’s Lee Bids Farewell.” Malaysia Star, May 1. Egan, Timothy (1994). “Indian Boys’ Exile Turns Out to Be Hoax.” New York Times, August 31.

292  Briana McGinnis Haveeru Daily (2002). “Perhaps They Loved, Too? Film Stars Niuma and Seezan Convicted for Extramarital Affair on the Eve of the Premiere of Their New Film ‘Yes, I Loved, Too.’ ” October 10. Haveeru Daily (2008). “Man Receives Banishment Sentence for Teaching Religion Without Permission.” February 9. Haveeru Daily (2009). “Video Interview of Maldivian Jihadist Available on Internet.” November 8. Haveeru Daily (2011). “Judge and Wife Convicted for Sexual Misconduct Near Hulhumale Rubbish Dump.” Minivan News, December 21. Hawwa Lubna, H. (2012). “Maldivian Fishing Crew Arrested on Sodomy Charges.” Minivan News, January 10. Hong, Lisa and Huang, Jianli (2008). The Scripting of a National History: Singapore and Its Pasts. Hong Kong: Hong Kong University Press. Human Rights Watch (2000). “Congo: Government, Rebels Suppress Free Expression.” September 20. Human Rights Watch (2005). “Health and Human Rights Organizations Condemn Turkmenistan Plan to Close Hospitals.” March 22. Human Rights Watch (2009). “Tunisia: End Activist’s Banishment.” July 28. Johansen, Bruce E. (2007). “Cast Among the Living Dead: Banishment is Back.” In The Praeger Handbook on Contemporary Issues in Native America, Vol. I. Westport, CT: Praeger, 186–95. Kershaw, Ssrah and Davey, Monica (2004). “Plagued by Drugs, Tribes Revive Ancient Penalty.” New York Times, January 18. Kunesh, Patrice (2007). “Banishment as Cultural Justice in Contemporary Tribal Legal Systems.” New Mexico Law Review 37, no. 85: 85–145. Kymlicka, Will (1995). Multicultural Citizenship: A Liberal Theory of Minority Rights. New York: Oxford University Press. Makan, Ajay (2007b) “Operation Complete: Army Clears Himandhoo Mosque.” Minivan News, October 8. Makan, Ajay (2007a) “Fifty Arrested on Himandhoo.” Minivan News, October 8. Maloney, Clarence (2013). People of the Maldives Islands. New Delhi: Orient Blackswan/University Press of India. Mannix, Andy (2011). Mille Lacs Ojibwe Fighting Violent Offenders with Banishment. Minneapolis, MN: City Pages, November 9. McNamara, Luke (1993). Aboriginal Peoples, the Administration of Justice and the Autonomy Agenda. Winnipeg: Legal Research Institute of the University of Manitoba. Meyer, Jon’a (2009). “ ‘How Do We Get Rid of Crime? Restore It to Harmony:’ Tribal Peacemaking as an Alternative to Modern Courts.” In Criminal Justice in Native America, ed. M. Nielsen and R. Silverman. Tucson, AZ: University of Arizona Press, 176–90. Monture-Okanee, Patricia and Turpel, Mary Ellen (1992). “Aboriginal Peoples and Canadian Criminal Law: Rethinking Justice.” University of British Columbia Law Review 26: 366. Mutalib, Hussin (2000). “Illiberal Democracy and the Future of Opposition in Singapore.” Third World Quarterly 21, no. 2: 313–42. Naish, Ahmed (2009). “Imam Among Seven Men Arrested for Homosexual Activity.” Minivan News, December 7.

Exile as an Alternative to Incarceration  293 Nazeer, Ahmed (2013). “Judge’s Public Sexual Misconduct Charge Upheld.” Minivan News, August 28. Nickerson Llewellyn, Karl and Hoebel, Edward Adamson (1941). The Cheyenne Way: Conflict and Case Law in Primitive Jurisprudence. Buffalo, NY: William S. Hein Publishing. Raffles, Thomas Stamford (1968). “Raffles’ Singapore Regulations, 1823.” editor unnamed, Malaya Law Review 10, no. 2: 248–91. Riley, Angela (2007). “(Tribal) Sovereignty and Illiberalism.” California Law Review 95: 799–848. Robinson, Paul and the University of Pennsylvania Law School Criminal Law Research Group Commissioned by the Office of the Attorney General of the Maldives and the United Nations Development Programme (2006). “Final Report of the Maldivian Penal Law & Sentencing Codification Project: Volume II Official Commentary.” Shukovsky, Paul (2003). “Drug Dealers Face Banishment.” Seattle Post-Intelligencer, May 18. Steil, Mark (2006). “Upper Sioux Banishes Drug Criminals.” Minnesota Public Radio, May 31. Strickland, Rennard (1982). Fire and the Spirits: Cherokee Law from Clan to Court. Norman, OK: University of Oklahoma Press. Tremewan, Christopher (1994). The Political Economy of Social Control in Singapore. State College, PA: Pennsylvania State University Press. Wilkins, David E. (2004). “Exiling One’s Kin: Banishment and Disenrollment in Indian Country.” Western Legal History 17, no. 2 (Summer/Fall): 235–62. Yazzie, Robert (1996). “ ‘Hozho Nahasdlii’- We Are Now in Good Relations: Navajo Restorative Justice.” St. Thomas Law Review 9, no. 1: 157–73. Yong, Ching Fatt and McKenna, R.B. (1990). The Kuomintang Movement in Malaya 1912–1949. Singapore: Singapore University Press. Zion, James (1985). “The Navajo Peacemaker Court: Deference to the Old and Accommodation to the New.” American Indian Law Review 11, no. 89: 89–109. Zion, James (1998). “The Dynamics of Navajo Peacemaking.” Journal of Contemporary Criminal Justice 14, no. 1: 58–74.

17 Corporal Punishment as an Alternative to Incarceration Jason Brennan

On March 3, 1994, Singapore sentenced American student Michael Fay to four months in jail, a small monetary fine, and six strokes of the cane after Fay pled guilty for vandalism (Shenon 1994). The American reaction was revealing. Hardly anyone complained that Fay would be locked away and deprived of freedom for months on end, but many Americans regarded Singapore’s practice of caning offenders as alien, barbaric, and backward. In this essay, I will argue Americans have their values backward. What they should have complained about was Fay’s incarceration, not the caning. Incarceration is a far more barbaric and inhumane practice. Indeed, if given the choice, Americans should replace incarceration with caning. The so-called Land of the Free is the Land of the Imprisoned. Roughly 1 in 100 adults is behind bars in the United States; and the US houses roughly a quarter of the world’s prisoners (Warren 2008). The Brookings Institute (Kearney and Harris 2014) claims that as of 2010, the US federal government, 50 US States, and local jurisdictions spent approximately $80 billion on incarceration, including “expenditures [to] fund the supervision, confinement, and rehabilitation of adults and juveniles . . . and confinement of persons awaiting trial and adjudication.” The Justice Policy Institute (2011) claims that average length for all sentences in the US is about 63 months, a number that does not include pre-trial confinement. The approximately 2.22 million people behind bars in the US right now will spend a total of about 140 million months, or 11.7 million years, or 150,000 average US lifetimes in prison. Incarceration is tremendously expensive in terms of money and time lost. And, as we will see, it is more also more degrading and cruel than caning. Caning is a form of corporal punishment in which an officer lashes a criminal offender’s buttocks or feet with rattan cane. Highly trained professionals administer the lashes, making sure not to strike the offender in places that leave visible scars. Afterward, the punished person receives medical care for his wounds, and is set free. To contemporary Westerners, caning—still practiced in Singapore, Malaysia, and elsewhere—seems horrific. Many of their complaints are legitimate. Nevertheless, in this paper I argue that the American practice of

Corporal Punishment  295 imprisonment is even more cruel. Depriving someone of freedom, subjecting them to extended risk of physical and sexual abuse, forcing them to remain locked in a small space, and in some cases, keeping them isolated and without entertainment for extended periods, is more barbaric than simply lashing an offender and letting him go. I do not argue here that caning is the best possible response to the problem of criminal justice. However, I will argue that it is superior to incarceration in most cases.

Three Intuition Pumps We in the West are accustomed to incarceration, while caning seems alien and weird to us. If we were to, say, watch a video of a flogging, we would likely react with horror and disgust. It seems only natural to conclude that caning is unjustifiable. But I hope the next three sections will help alienate us from those intuitions. Because we are accustomed to incarceration, we fail to see that incarceration is the more horrifying practice.

Making Contact Imagine a twin Earth, like ours, in which corporal punishment had never been invented. Instead, suppose every country punishes its criminals the way the United States does. Now, suppose in the near future, this twin Earth makes contact with the Vulcans, the highly rational humanoid race from Star Trek. Suppose that shortly after making contact, we send ambassadors to planet Vulcan to learn about differences in cultural practices, institutions, and norms. We are surprised to learn that the Vulcans use corporal punishment in response to crime. In particular, when a criminal is convicted, he is immediately brought to a caning station. There, a trained medical doctor evaluates the criminal to ensure he is healthy enough to withstand the caning. If the criminal passes, he is placed on a special structure that holds him intact, which exposes his buttocks, and which ensures that his genitals or vital organs will not be struck. Then, a highly trained guard administers a set number of lashes, with more severe crimes leading to a higher number of lashes. For severe crimes, the criminal is struck with a cane that may break open his skin. For less severe crimes, the criminal either pays a monetary fee, or is subject to a far less severe form of flogging which does not break the skin. Lashes come at a set rate of one minute apart. After the convict is flogged, a medical doctor treats and dresses his wounds. He is kept in a hospital as he convalesces, and then is set free. The wounds are painful, but superficial, and impose no long-term ill effects other than scarring in areas that are always covered by clothing. However horrified we Earthlings are when we discover that the Vulcans use corporal punishment, they are even more horrified to discover what Earthlings do. We explain that when a prisoner is convicted, the prisoner is

296  Jason Brennan sent to live, for a long time, in a tiny, grey concrete cell with other criminals. In some cases, e.g., in Maricopa County, Arizona, the convict is made to live in cramped tents baking underneath the hot desert sun. The prisoner is deprived of all privacy and required to shower, micturate, and defecate in front of others. We require the prisoner to wear bright orange clothing and remove as many vestiges of individuality as we can. We deprive prisoners of sexual satisfaction, even forbidding them (though unsuccessfully) to masturbate, and forbidding them from having access to materials (such as pornography) to aid in achieving private sexual satisfaction. However, we also fail to protect prisoners from one another, and as a result, many prisoners are raped and abused by the others, and even those who escape rape live with constant fear of being raped. Prisoners are surrounded by metal bars. They face a set daily routine, and are told what to do, where to be, when to eat, and when to sleep by poorly educated, poorly paid guards who regard them as subhuman. Their conditions are so unsanitary that no blood bank will accept their blood for at least a year after their release. Most receive only a modicum of leisure time. If they misbehave, they are forced to spend weeks in even tinier isolated cells, with no form of entertainment. Further, the prisons are filled with violence, drugs, and alcohol, and many prisoners find they need to join violent gangs for some protection against violence (Skarbek 2014). Prisoners spend years of their lives behind bars; some even spend the rest of their lives there. Most receive at most a few visits from loved ones. Over time, their loved ones tend to abandon them and move on, as they have to, since the prisoners are shut off from society. When (and if) prisoners finally are released, most are not fully free. Instead, they are subject to a wide range of restrictive rules about how and where they live, work, and play. They must check in regularly with a police officer, who has the power to send the prisoner back to prison for even minor transgressions, such as having a glass of wine. Most released prisoners find it difficult to integrate back into normal, outside life, in part because they are stigmatized, in part because they have spent months or years living and making friends with other criminals, and in part because their lost time and lost skills make them unemployable. A large percentage become desperate and commit further crimes. Now imagine how the Vulcans might react: “Sure, we beat our convicts, but we let them go. You Earthlings enslave your convicts, place them under constant surveillance, remove their privacy, dehumanize them, bore them, psychologically torture them, and subject them to near constant threats of capricious physical and sexual abuse. You ruin their lives, and you make it so that these people provide nothing of value to society, either. And you do this all at tremendous monetary expense. We punish criminality with a moment’s brutality; you punish criminality with prolonged dehumanization punctuated by capricious, unexpected brutality.”

Corporal Punishment  297

What American Sentencing Guidelines Subtly Reveal Our own criminal justice system appears to reflect the judgment that striking someone with a cane on the buttocks is far less bad than imprisoning them. Though it varies from state to state, the criminal penalty we assign to battery is generally far less severe than the penalty we assign for prolonged false imprisonment or kidnapping. According to criminaldefenselawyer.com and similar websites, if you beat a person unlawfully with a wicker cane on the buttocks, you might receive a prison sentence of six months; if you unlawfully imprison someone in your basement for six months, you would go to jail for many years. The difference in the penalties attached to these crimes is meant to reflect the difference in the harshness or wrongness of treatment the criminal inflicted on the victim. By parity of reasoning, caning is less harsh than imprisonment.1

The Choice Peter Moskos (2011: 9) asks readers to consider, if they were convicted of a crime, which would they choose, incarceration or flogging? Before you’re lead out of the courtroom, the judge calls for order and offers you the flogging option. “Five years or ten lashes,” he says. If you choose flogging, an appointed state flogger will cane you immediately. Ten lashes, a little rubbing alcohol, a few bandages, and you’d be free to go home and sleep on your own bed. No holding cell. No lock-up. A quick and painful caning, and you’ll be on your way. Would you choose years in the joint over a brief punishment, however cruel? Keep in mind the description of American prisons from the previous section. Moskos expects most of his readers would choose flogging over incarceration. If he’s right—admittedly, he did not take a survey, but it is plausible most would agree—then that tells us something about how to compare incarceration to flogging. If someone would choose flogging over incarceration, then incarceration must be, all things considered, worse. Moskos suggest that instead of having US criminal courts mandate flogging, they should simply offer it as an option. I will follow his suggestion here. On Moskos’s scheme, convicts can choose incarceration if they so desire. Adding a new option simply expands the range of punishments, while letting the prisoner decide on the punishment means that the prisoner will likely receive the least bad punishment. If a person is free to choose from a range of options, the fact that a person chooses one option over the other is evidence that that is his best option. For instance, suppose Ivan is about to executed. Suppose the executioner offers him a choice of execution methods: guillotine, burning at the stake, or

298  Jason Brennan boiling. If the prisoner chooses the guillotine over the others, that is strong presumptive evidence that this is the least bad/best option from the prisoner’s point of view. This point holds even though the prisoner is not fully free or autonomous, and even though all the options are bad. Further, we do not generally harm the prisoner by simply adding a few other options to choose from (e.g., lethal injection, being eaten by lions, etc.) If an option is better, the prisoner will take it; if it’s worse, he will not. In Moskos’s scheme, prisoners are offered the choice between incarceration or flogging. He suspects most would choose flogging. He and I certainly would. What this means, then, is that to defend incarceration over flogging, we must in general take one of three stances. One stance is vindictive: it holds that in some way flogging goes too easy on prisoners, i.e., that the reasons they should be imprisoned is that flogging is not painful enough. A second stance is paternalistic: it holds that when prisoners choose flogging over incarceration, it must be because they are uninformed or irrational, and so we should choose incarceration on their behalf. A third stance holds that flogging will not work on some prisoners, as they are too dangerous to be set free. This point may be true but holds for only a minority of prisoners. I will discuss this further below.

Caning vs. Incarceration in Fulfilling the Purpose of Criminal Justice There is no consensus view in philosophy about what the purpose of criminal justice is (see Bedau and Kelly 2015). I will not here defend a particular view or attempt in vain to settle the matter. Still, we might consider each of the following putative purposes, and ask how caning fares compared to incarceration: 1. Retribution: A wrongdoer ought to suffer in proportion to the harm he has caused others. 2. Rehabilitation: Criminal justice should reform wrongdoers and turn them into productive, peaceful, functioning citizens. 3. Deterrence: Criminal justice should disincentivize potential offenders. 4. Communication: Criminal justice should communicate society’s anger or moral judgment of the wrongdoer. 5. Restitution: Criminal justice should require the wrongdoer to make his victims whole, to compensate them for the harms he has caused. 6. Removal/incapacitation: Criminal justice should remove dangerous wrongdoers from society in order to protect the innocent. Again, I take no stance on which of 1–6 (or which hybrid mix of 1–6) is correct. Each of these theories has some familiar problems. However, I will argue in this section that for most of 1–6, caning is probably superior to incarceration.

Corporal Punishment  299 Regarding retribution: The main idea behind retribution is that wrongdoers deserve to suffer the harm they have caused others. Of course, there are problems in interpreting just what that means and how one could implement it. A mass murderer has only one life to give. Few people seriously advocate having the courts punish rapists with rape, though many look sideways when such rapists are raped in prison. Putting people in jail and depriving them of freedom may seem like just retribution for kidnappers, or perhaps for murderers (who deprive their victims of freedom by killing them). But for most others, it seems out of sorts. If a person steals my car, it does not seem as though he “suffers the harm he caused me” by having to spend a few years in jail. (If anything, he suffers more greatly than I do.) Moskos (2011: 22–3) argues that flogging serves the desire for retribution far better than incarceration. He asks us to imagine that someone has developed a machine—call it the “reformatron”—which instantly rehabilitates any criminal placed inside it, curing that person of vice, and instilling in that person perfectly good character and desires. Moskos says that even if such a device were created and widely used, many people would feel something is missing. If you had been beaten by an attacker and that attacker were placed in the reformatron, you would feel that justice had not been done. You might be glad the attacker was reformed, but you would feel he got away with wrongdoing. Flogging, Moskos says, seems to satisfy the desire for retribution, while prison does not. “. . . [B]ecause prisons do not punish in a comprehensible manner, incarceration will never satisfy the public’s legitimate desire for punishment. . . . Without satisfactory punishment, the public brays for more punishment” (2011: 95–6). And so, he claims, the public demands ever longer sentences, but remains unsatisfied. Regarding rehabilitation and deterrence: According to the rehabilitation theory, the purpose of criminal justice is to reform criminals. They should be taught useful skills, taught to correct their impulsive behaviors, and somehow induced to have more pro-social attitudes. The goal is to take people whom society is better off without and turn them into people society is better off with. In contrast, deterrence theory claims that the point of punishment is to prevent criminals from committing crime in the first place by disincentivizing crime. The deterrence theory relies largely upon a “rational criminal” theory of crime, according to which criminals commit crime because the expected benefits of crime outweigh the expected costs. Punishment is supposed to change the calculus of the expected costs. For instance, suppose all crimes were punished by placing the criminal in a pain amplifier for eternity. In that case, it would never be rational to commit crime if there was even the slightest chance of being caught. In a pair-wise comparison, which does better from a rehabilitative or deterrence standpoint, incarceration or flogging? Since flogging is not widely

300  Jason Brennan practiced, it is not widely studied. No one thinks that receiving a lash somehow induces virtue. It’s also not clear that most criminals actively weigh the costs and benefits of crime before committing crime; on the contrary, they appear to be “impatient, myopic, or both,” and to act impulsively (Lee and McCrary 2005). The extant research indicates clearly that having a high probability of being caught and punished deters crime, but it does not appear that incarceration per se does much to deter criminals, and there is little evidence that adding extra length to sentences does much to deter criminals (Nagin 2013; Levitt 2004; Doob and Webster 2003; Langan and David 2002; Levitt 1998; Gendreau, Little, and Goggin 1996; Farrington, Langan, and Wikstrom 1994). In fact, some of the empirical work finds that increased sentences promote recidivism, as the longer a prisoner is in jail, the more he becomes institutionalized and the weaker his ties to his community become (Gendreau, Goggin, and Cullen 1999). Overall, rates of recidivism remain very high in the US and places with heavy incarceration. It makes sense: putting a criminal in a camp with other criminals, where he therefore makes friends with criminals or must adopt a tough, alpha male mindset to survive, all while breaking him off from civil society and the family and community ties that give him a stake in behaving well, is not an obvious strategy for rehabilitation. In short: incarceration per se does not seem to work as an effective means of rehabilitating or deterring criminals. What matters is not how severe the punishment is, but that criminals believe there is a high probability of being caught and punished. A priori, there is little reason to think that caning would fare even worse. However, if we give people lashes, let them suffer through healing for a few weeks, then let them go about their lives, we at least avoid some of the problems incarceration causes. We return people to their communities. We let them get back to work. We do not break off their ties to law-abiding citizens and friends. Further, since incarceration is very expensive but flogging is cheap, if the state replaces incarceration with flogging, it could use some of the leftover funds to A) increase policing in high crime areas or B) create job training programs, welfare programs, and the like. I take no stance here on which of these is preferable, but at the very least A is more effective than incarceration in preventing crime. Regarding Communication: Some theories of criminal justice hold that the point of criminal justice “is to communicate to offenders the censure or condemnation that they deserve for their crimes” (Duff 2013). While this theory of criminal justice is popular among philosophers, it faces a serious problem: which particular form of communication is best? If we want to communicate to a criminal that he deserves censure, instead of incarcerating him or requiring him to do hard labor, we might make him wear a scarlet letter, put him in the pillory, post a billboard condemning him, write a scathing poem about him, build a mocking statue, or have an insult comic dress him down in public each day for six months. As Duff

Corporal Punishment  301 (2013) points out, there is no obvious connection between communicating censure and incarceration. On its face, flogging a criminal seems just as good a means to communicate censure as incarceration. As a matter of fact, certain cultures have adopted semiotic codes of meaning by which incarceration is used to signal society’s disapproval. But this is at most a contingent, recently developed social construct that incarceration has that meaning. It is not written into the fabric of the universe. In general, we can judge such social constructs by their utility. For instance, consider that when I utter the sounds “I love you,” this signals concern and care. It really is a fact of the matter that the words express such. But this is only because of socially constructed patterns of meaning English speakers have attached to those sounds. In an alternate universe, “I love you” could be used by speakers to signal hatred or contempt. Now, suppose we discovered that uttering the sounds “I love you” had certain negative consequences. It turns out, thanks to bizarre laws of physics, that saying “I love you” to another person increases that person’s chance of developing cancer by half a percent. If we discovered that this form of communication had such negative consequences, we would not shrug our shoulders and continue talking that way. Instead, we would change the meaning of the English language, stop saying “I love you,” and find some other way to express concern and love. It would be monstrous of us to continue using such a dysfunctional and damaging method of communication. Similar remarks apply to incarceration. As others in this volume have attested, incarceration is replete with problems. It is dysfunctional. It does not “work.” It causes more problems than it solves. Thus, even if American culture has fixated on using incarceration as a method for expressing disapproval, this just seems to show that Americans should find some other way to communicate censure. Regarding Restitution: The restorative or restitution theory of criminal law holds that wrongdoers should compensate their victims or make their victims whole. Proponents of this theory (e.g., see Hasnas in this volume) sometimes reject the idea of criminal justice altogether and instead hold that all “crimes” should instead be treated as torts. Incarceration does not compensate victims or make them whole. On the contrary, it is tremendously expensive and must be paid for via taxes. When we throw a criminal in jail, we not only fail to compensate the victim, but in many cases make the victim partially pay (through taxes) for the criminal’s incarceration. On its face, flogging does not compensate the victim either. However, since flogging is far less expensive for the state than incarceration, Moskos (2013: 143) proposes an in-between solution: if the criminal agrees to a flogging, some portion (say half) of the money the state would have spent on incarcerating the prisoner will go to the victim. In this case, while the criminal himself does not compensate the victim, at least the victim receives some degree of compensation from the state.

302  Jason Brennan Regarding Removal: Removal and incapacitation theories hold that the purpose of criminal justice is simply to remove dangerous criminals from society and prevent them from committing further crimes. Pure versions of these theories face a problem in that it is unclear why we should wait until after a person commits a crime to remove him from society—after all, certain people (e.g., American black men who drop out of high school) who have not yet committed crimes are statistically very likely to commit crimes, while certain people who have committed crimes (e.g., upper class, white-collar criminals, or men who murder their cheating spouses in crimes of passion) are statistically unlikely to do so again. These problems aside, this may be one theory of criminal justice against which incarceration fares better than flogging. When we incarcerate a person for a long time, we deprive him of freedom and remove him from the outside world. He may continue to commit crimes, but at least his criminal activities will be directed against other criminals. When we flog a person, we let him go. If the criminal is dangerous, he is free to hurt more innocent people. Still, there are things to balance here. Incarceration for lengthy periods has a high recidivism rate in part because it disrupts a prisoner’s life, removes his contacts in civil society, ruins his résumé, institutionalizes him, and socializes him with other criminals. Incarceration prevents the criminal from hurting outsiders during his term, but it may lead to more crime once he is removed. That said, there may be some people so dangerous that they should not be allowed back in civil society at all. Flogging may not stop serial killers, unrepentant child rapists, and others. So, perhaps incarceration (or exile or even execution) is a better response to extremely dangerous criminals. But this applies only to a minority of criminals. To summarize, I have taken no stance here on what the purpose of criminal justice ought to be. However, it appears that for most of the theories of criminal justice, including retribution, deterrence, rehabilitation, communication, and restitution theories, caning either beats or ties incarceration. It only fares worse on removal theories, and even then, only when applied to the most dangerous criminals. Accordingly, at this point, we have a strong presumptive case in favor of replacing (or allowing prisoners the option to replace) incarceration with caning. In the remaining sections, I will address specific objections to caning.

The “Corporal Punishment Is Torture” Objection One possible objection to caning goes roughly as follows: 1. Caning is a form of torture. 2. Torture is either always wrong, or wrong except in bizarre and unique circumstances (such as ticking time bomb cases). 3. Therefore, caning is always wrong, or wrong except in bizarre and unique circumstances.

Corporal Punishment  303 The idea here is that we just need to show caning falls into the conceptual category of torture, and that is sufficient to show it is wrong. Matthew Kramer (2014) offers what is perhaps the most thorough and rigorous definition of torture. (I will not attempt to reconstruct his objections to torture here, as I find his arguments mostly impenetrable.) He claims it consists of six components: 1. Torture involves the infliction of severe pain and suffering. 2. Torture is conducted for a purpose, such as punishment, the enjoyment of the torturer, or some other purposes. 3. Torture generally continues until the purpose is achieved. 4. Torture shows indifference to the well-being of the victim. 5. Victims do not consent to the torture. 6. Victims lack control over the length of the torture. Now, one may certainly dispute whether this is a good characterization of torture (e.g., Lenta 2015). Indeed, “torture” is a natural kind. It is unlikely we will be able to provide necessary and sufficient conditions for the concept. Rather, it seems more likely to be an archetype concept. However, even if we accept this characterization, caning arguably escapes it. Judicial caning does meet criteria 1–3 and 6. However, we need not suppose that caning shows indifference to the well-being of the criminal. On the contrary, we might practice caning precisely because it is a form of punishment that shows greater concern for the victim’s well-being than incarceration. After all, as we discussed above, incarceration is generally worse than caning. If I could wave a magic wand and replace American-style prisons with caning, I would be a humanitarian. It is not just that caning could show more concern for a person’s well-being than incarceration. Rather, the kind of caning we are discussing here involves carefully monitoring a person’s health, carefully ensuring that he is not grievously or permanently injured, and then providing him with medical care afterward. These are not behaviors that indicate a lack of concern for the victim’s well-being. Further, as Koskos recommends, the most defensible form of caning does not require criminals to be caned. Instead, criminals are offered the choice between caning or incarceration (or other penalties). Since this is a forced choice between two unpleasant options, the criminal cannot be said to fully consent to the caning, in the way that perhaps a masochist might consent when he enters a torture dungeon. Nevertheless, when we offer a person a choice between two bad options, he can still freely choose between those two options. We may not want to edify his choice by claiming it was, all things considered free, and consensual. But if a person chooses caning over incarceration, or would choose caning if given the option, we certainly do not show respect for his capacity to consent by imposing incarceration anyway. At any rate, the torture objection is too simple. Moral philosophers tend to think the way to argue against a policy is simply to list its faults and/or

304  Jason Brennan to fit it into some objectionable category. This is one way in which philosophers tend to exhibit less intellectual seriousness about moral matters than supposedly amoral social scientists such as economists. An economist would not ask whether torture is absolutely right or wrong. Instead, she might ask, what are the available options in light of the technical and motivational constraints we face, and then try to determine which of these is best. Personally, I have little idea what an ideal criminal justice system would look like. However, I am not optimistic about American federal or state governments’ competence to implement anything like well-functioning, productive rehabilitation centers. We are competent to cane prisoners, and we would do it better than we do incarceration.

Is Caning Dehumanizing and Degrading? One objection to caning holds that caning is shameful and degrading. To cane someone lowers that person’s status and standing. It treats them in an inhumane way. The entire spectacle is embarrassing. It violates their bodily integrity. Since at the very least the criminal’s buttocks will be exposed, it also violates their privacy. David Benatar (1998: 243–4) has a powerful response to this objection: Here it is noteworthy that there are other forms of punishment that lower people’s standing even more than corporal punishment, and yet are not subject to similar condemnation. Consider, for example, various indignities attendant upon imprisonment, including severe invasions of privacy (such as strip-searches and ablution facilities that require relieving oneself in full view of others) as well as imposed subservience to prison wardens, guards, and even to more powerful fellow inmates. My intuitions suggest that this lowering of people’s standing surpasses that implicit in corporal punishment per se. . . . If corporal punishment is wrong because it involves violating the intimate zone of a person’s body, then surely the extreme invasions of prison inmates’ privacy, which seem worse, would also be wrong. When we cane someone, we intentionally inflict severe pain on that person for a short period of time. That is of course distinct from incarceration. But incarceration generally involves long-term severe restrictions on a person’s liberty, long-term constant surveillance, long-term exposure to risk of capricious abuse and physical harm, forced boredom, forced labor, deprivation of privacy, deprivation of sexual satisfaction, and separation from loved ones, friends, and civil society, and so on. As Benatar concludes, incarceration is even more degrading. So the objection that caning is degrading does not rule out caning unless it also rules out incarceration. Philosophers of a Kantian bent might in the future produce some newfangled theory of dignity by which corporal punishment will somehow qualify as undignified in a way that incarceration does not, despite all the horrors of incarceration I described above. I admit I am rarely impressed by appeals

Corporal Punishment  305 to dignity; I tend to side with Peter Singer (1989: 57) that such “fine phrases are the last resort of those who have run out of arguments.” Instead, I will just grant here, for the sake of argument, that there is some coherent conception of Kantian dignity under which it turns out that caning is inherently undignified in a way that incarceration, despite all its horrors, is not. Even if we grant this point, we can still meaningfully say, so what? If a person chooses caning over incarceration, that is strong evidence that caning, however undignified it might be, is, all things considered, less bad than incarceration, however more dignified it might be. It will not be enough to show that caning is in some way undignified. Rather, one would need to show that there can be no trade-offs between dignity of some sort and other moral values.

The Paternalistic Objection Koskos thinks hardly anyone would, if given the choice between caning and incarceration, choose incarceration. Admittedly, he did not conduct a survey, and neither have I, but that is because the choice to us seems so easy. He takes that as presumptive evidence that a year of incarceration is significantly worse than two lashes on the buttocks. As I discussed above, our own criminal code seems to concur, since we would punish a criminal more severely for A) falsely imprisoning a person for a year than for B) lashing a restrained victim twice with a cane and then letting that person go. Nevertheless, one might object that people would choose caning only because they do not really understand what they are signing up for. Perhaps caning is far worse. Perhaps the state should override criminals’ preferences for their own sake. But there seems to be an easy solution to this problem. Criminals could be supplied with educational materials (preferably not those created by the state, which may be under control of vested interests such as prison guard unions) describing the differences between incarceration and caning. After being required to read these materials, and perhaps even tested to ensure understanding, then the criminal would choose. Thus, paternalistic concerns are easily remedied without banning caning.

Better Incarceration Methods? One major objection to my thesis is that it problematically contrasts caning to American-style prisons. Sure, American-style prisons are horrific, and so caning is plausibly preferable to imprisonment. However, perhaps an alternative prison system would be better. Consider this description (Larson 2013) of the Scandinavian and Finnish model: Suomenlinna Island has hosted an “open” prison since 1971. The 95 male prisoners leave the prison grounds each day to do the township’s general maintenance or commute to the mainland for work or study.

306  Jason Brennan Serving time for theft, drug trafficking, assault, or murder, all the men here are on the verge of release. Cellblocks look like dorms at a state university. Though worse for wear, rooms feature flat-screen TVs, sound systems, and mini-refrigerators for the prisoners who can afford to rent them for prison-labor wages of 4.10 to 7.3 Euros per hour ($5.30 to $9.50). With electronic monitoring, prisoners are allowed to spend time with their families in Helsinki. Men here enjoy a screened barbecue pit, a gym, and a dining hall where prisoners and staff eat together. Prisoners throughout Scandinavia wear their own clothes. Officers wear navy slacks, powder-blue shirts, nametags and shoulder bars; but they carry no batons, handcuffs, Tasers or pepper-spray. In these prison systems, prisoners are only partially removed from society. They are made to live together, but in humane conditions. They are put to work or allowed to keep their jobs, and paid real wages. They retain their outside relationships. Their main “punishment” is a loss of freedom, but it is a loss of freedom closer to that of having strict parents than the kind of loss American prisoners face. Note that the Scandinavian countries and Finland also have closed, high security prisons in the American style. Conditions there may be somewhat better and more humane, but prisoners are still kept under more constant surveillance and are in greater danger from one another. My goal in this essay is not to argue that caning is the best possible practice. However, one might argue that “good” incarceration is better than caning, and that all I’ve done here is show that caning is better than “bad” American-style incarceration In some sense, I am willing to concede this kind of argument. However, there are a number of difficult questions we would have to answer in order to reach a judgment either way. First, is not clear that just because something works in Sweden it can work in the United States. (I admit, this applies to caning as well.) The Scandinavian countries have remarkably low crime, have small, relatively homogeneous populations, are very conscientious and have high trust, and so on. It may turn out, empirically, that such forms of incarceration only “work” in that kind of context. Second, there are cost trade-offs that need to be considered. The Scandinavian model, with its focus on humane treatment and rehabilitation, may be more expensive up front, though perhaps less expensive in the long run. That said, compared to caning and setting someone free, it will certainly be more expensive. However, even if we concede that these points come in favor of the Scandinavian model, there are still reasons to prefer caning. First, we might ask, if you were convicted of, say, grand theft auto, would you prefer A) to spend 4 years in a Scandinavian open prison, or B) be lashed eight times and then set free? Many people (including me) would find B preferable to A. Perhaps some would find A preferable to B. But that indicates that the Scandinavian model is arguably harsher than caning, though perhaps it should be

Corporal Punishment  307 preferred, all things considered, if it turns out to have a better rehabilitation rate. Second, consider that if I am convicted for A) kidnapping you and keeping you imprisoned in a Scandinavian-style open prison for a few years, I would receive a harsher sentence in most countries than I would if I simply B) lashed you eight times with a cane. So, while the Scandinavian model is better than the American, even it is not clearly better than caning.

Summary The problem of criminal justice is puzzling. A perfectly just society would not have institutions of criminal justice at all. The problem of criminal justice is thus concerned with asking what the best, most effective, and most just response is to certain forms of injustice. In this chapter, I take no stance on what the purpose of criminal justice ought to be, or what the ideal form of criminal justice is. Instead, I have defended a narrow point: for most crimes, caning criminals is better—cheaper, more effective, more humane—than incarcerating them. Incarceration is a sensible penalty only in the minority of cases where criminals are so dangerous that we may justly remove them from society in order to protect ourselves.

Note 1 Thanks to Alexei Marcoux for this wording.

References Bedau, Hugo Adam and Kelly, Erin (2015). “Punishment.” In Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta. Available at http://plato.stanford.edu/ entries/punishment/ Benatar, David. (1998). “Corporation Punishment.” Social Theory and Practice 24: 237–60. Doob, Anthony and Webster, Cheryl (2003). “Sentence Severity and Crime: Accepting the Null Hypotheses.” Crime and Justice 30: 143–95. Duff, Antony. (2013). “Legal Punishment.” In Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta. Available at http://plato.stanford.edu/entries/ legal-punishment/ Farrington, David, Langan, Patrick and Wikstrom, Per-Olaf. (1994). “Changes in Crime and Punishment in America, England, and Sweden Between the 1980s and 1990s.” Studies in Crime Prevention 3: 104–31. Gendreau, Paul, Goggin, Claire. and Cullen, Francis. (1999). The Effects of Prison Sentences on Recidivism. Ottawa: Public Works and Government Services Canada. Gendreau, Paul., Little, Tracy. and Goggin, Claire. (1996). “A Meta-Analysis of Adult Offender Recidivism: What Works!” Criminology 34: 575–607. Justice Policy Institute (2011). Finding Direction: Expanding Criminal Justice Options by Considering Policies of Other Nations. Washington, DC: Justice Policy

308  Jason Brennan Institute. Available at www.justicepolicy.org/uploads/justicepolicy/documents/senten cing.pdf Kearney, Melissa. and Harris, Benjamin. (2014). Ten Economic Facts About Crime and Incarceration in the United States. Washington, DC: Brookings Institute. Available at www.brookings.edu/research/ten-economic-facts-about-crime-andincarceration-in-the-united-states/ Kramer, Matthew. (2014). Torture and Moral Integrity: A Philosophical Enquiry. New York: Oxford University Press. Langan, Patrick. and Levin, David. (2002). “Recidivism of Prisoners Released in 1994.” U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics. Larson, Doran. (2013). “Why Scandinavian Prisons Are Superior.” The Atlantic, September 13. Available at www.theatlantic.com/international/archive/2013/09/ why-scandinavian-prisons-are-superior/279949/ Lee, David. and McCrary, Justin. (2005). “Crime, Punishment, and Myopia.” National Bureau of Economic Research working paper 11491, Cambridge, MA. Lenta, Patrick. (2015). “Is Corporal Punishment Torturous?” Journal of Applied Philosophy, online first edition. doi: 10.1111/japp.12134 Levitt, Steven. (1998). “Why Do Increased Arrest Rates Appear to Reduce Crime: Deterrence, Incapacitation, or Measurement Error?” Economic Inquiry 36: 353–72. Levitt, Steven (2004). “Understanding Why Crime Fell in 1990s: Four Factors that Explain the Decline and Six that Do Not.” Journal of Economic Perspectives 18: 163–90. Moskos, Peter. (2011). In Defense of Flogging. New York: Basic Books. Nagin, Daniel S. (2013). “Deterrence in the Twenty-First Century.” In Crime and Justice: A Review of Research, Vol. 42: Crime and Justice in America, 1975–2025, ed. Michael Tonry. Chicago: University of Chicago Press. Shenon, Philip. (1994). “A Flogging Sentence Brings a Cry of Pain in the U.S.” The New York Times, March 16. Singer, Peter. (1989). “All Animals Are Equal.” In Environmental Ethics: What Really Matters, What Really Works, 2nd ed., ed. David Schmidtz. Oxford: Oxford University Press. Skarbek, David. (2014). The Social Order of the Underworld: How Prison Gangs Govern the American Penal System. New York: Oxford University Press. Warren, Jenifer. (2008). “One in 100: Behind Bars in America 2008.” Pew Center on the States. Available at www.pewtrusts.org/~/media/legacy/uploadedfiles/ wwwpewtrustsorg/reports/sentencing_and_corrections/onein100pdf.pdf

18 The Potentials and Limitations of De-Incarceration Daniel J. D’Amico

I. Introduction In the wake of the fastest and largest accumulation of inmate populations in modern history, the American prison system is now seen as the archetypal case of “mass imprisonment.” Standard definitions reference prison growth’s supposed Western cultural attributes, its extremity, its rapid expansion, and its disparate impact upon poor minorities (Garland 2001). Embedded within this label, intended to distinguish the phenomena apart from ordinary rates of incarceration, is a normative assessment that prison populations are excessive, unjustified, and demanding of reform. This chapter investigates these implications by surveying the potentials and limitations of alternative punitive institutional arrangements. What’s striking about most philosophical punishment theory is how irrelevant it is to justifying actual punishments. While substantial philosophical attention has been dedicated to justifying criminal punishments, far less work has aimed at specifically justifying imprisonment apart from other techniques. Furthermore, most normative analyses of imprisonment stand apart from punitive philosophy more generally, as they often call for the radical abolition of prison space.1 I argue this tension stems largely from a dearth of research regarding the comparative potentials and limitations of non-imprisonment forms of punishment. This chapter first surveys the moral justifications of criminal punishment. Any normative assessment of contemporary imprisonment must first rely upon some justification(s) of criminal punishments more generally. If punishment is unjustified, then punishment by imprisonment is also unjustified. If punishment is justified, it remains uncertain if punishment by imprisonment is justified. If criminal punishments via imprisonment are unjustified, mass imprisonment is also obviously unjustified. If criminal punishments via imprisonment are justified, then what is to be said of mass imprisonment? As Kant (1781) would say succinctly, “ought implies can.” The normative claim that mass incarceration is demanding of reform, implies that some preferable alternatives are at least feasible. Insofar as all popular paradigms need to provide justified reasons for the imposed harms of criminal

310  Daniel J. D’Amico punishments, they also share a commitment to minimizing the social harms of punishments and punitive institutions. Hence, to justify incarceration apart from other punishment types requires some accurate knowledge of imprisonment’s unique costs and consequences. Such cost benefit calculations in turn depend upon some comparative assessment regarding the potentials and limitations of alternative schemes. This chapter surveys research from two sets of social environments in the hopes of better understanding the practicalities of punishment norms other than mass imprisonment. In section III, I survey the empirical research surrounding cross-country incarceration rates. Contemporary nation-states, historically founded by different varieties of the civil law tradition, have avoided mass incarceration relative to common law counterparts. Rather than imprisonment, civil law countries tend to leverage a variety of social monitoring and enforcement techniques more available therein due to larger governmental bureaucracies and more pervasive social welfare programs compared to under the common law. In section IV, I survey research of social environments wherein the relative absence of traditional governments has occurred in stride with a similar absence of punishments via imprisonment. Ancient societies, primitive tribes, and criminal organizations in the underground economy all manage to coordinate complex swaths of human behaviors by leveraging pre-emptive security technologies, reputational mechanisms, restitution payments, and cultural rituals of dispute resolution and apology. In section V, I ask if these non-incarceral strategies are viable for countries currently enduring mass imprisonment. Simply put, not really. More extensive bureaucratic infrastructures and social welfare programs under the civil law developed over long evolutionary swaths of legal, cultural, and economic history. Hence, developing such in contemporary common law settings would entail large start-up costs and/or radical changes in the institutional organization of legal and political processes therein. Such programs would also likely impede long run economic performance if imposed within common law jurisdictions. Though this type of institutional benchmarking or transplantation seems costly and/or ineffective, the available empirics do imply some potential(s) for genuine institutional innovations without disrupting economic performance. Second, punishment norms in conveniently stateless societies ought not to be heralded as just or ideal, as they are often viscerally severe relative to contemporary processes. Their strong dependence upon reputational mechanisms similarly limits their cross-viability in contexts with greater population sizes and degrees of ethnic diversity. But, I argue there are meaningful inferences to be garnered by viewing mass incarceration within a broader context that includes these two sets of social environments. Mass imprisonment appears to occur in stride with more hierarchically organized criminal justice systems. In contrast, social environments wherein criminal justice decision making is more localized and decentralized tend to harbor more

Potentials, Limitations of De-Incarceration  311 adaptive processes of criminal justice decision making and greater potentials for institutional innovation therein. Though apparently harsh, punishment norms in relatively stateless contexts do effectively tend to reduce and avoid social harms relative to the limited viable alternatives within their localities. To preserve and protect the feedback processes and evolutionary potential(s) of criminal justice institutions, I argue punitive philosophy would benefit from being viewed through a lens of constitutional political economy, wherein justifications and justificatory frameworks are not applied to particular institutional forms of governance or punishment ex post. Instead, government institutions, especially punitive enforcements, if necessary, should also be recognized as potentially threatening to social order when excessively imposed. Effective governance is thus a practical task of minimizing the size and scope of centralized authority. Hence, imprisonment is only justified as a form of criminal punishment insofar as it is embedded within a reliable and effective system of checks and balances. Hence, a potentially viable strategy for mass imprisonment reform would be to consistently extend designed checks and balances to preserve decentralized authority similar to those frequently found within the civic adjudication and tort procedures under the common law, into the spheres of criminal justice decision making therein.

II. Justifying Imprisonment Apart from other Punishments Insofar as criminal punishment entails the infliction of intentional harm, it requires justification. Several different perspectives aim to provide good reasons for why criminals should be punished. Consequentialists, retributivists, moral expressionists, and others provide different justificatory frameworks with different types of reasons for punishment (Berman 2008; Duff 2013; Bedau 2015). If an act meets the criteria laid out by the correct theory, then it is said to be justified. If punishments meet several criteria across alternative theories, even better (Rawls 1993). Conspicuously, no such framework seems to provide explicit justification(s) for imprisonment apart from other punitive techniques. All popular justifications for punishment share a common commitment to minimizing harm. Insofar as the need for justification is inspired by the inherently harmful nature of punishment, it also follows that any particular punishment type or institutional provision ought to minimize (or at least aim to minimize) the harms it imposes. If the harms imposed by punishment onto criminals demand justification, then any harms or costs imposed by punishment(s) upon other members of society demand justification as well. Perhaps even more so as non-criminal third parties lack desert. Hence, to provide specific justification for imprisonment or mass imprisonment, one must engage in some comparative assessments across mass imprisonment and other punishment practices regarding their associated social effects.

312  Daniel J. D’Amico A strict consequentialist would justify punishments with reference to their preferable effects.2 Hence, consequentialists tend to focus upon deterrence, incapacitation, and/or rehabilitation as desirable outcomes. Some complain consequentialism proves too much, as it appears to pay little or no attention to proportionality.3 For example, if hanging a man for petty theft carries enough deterrence, it can be seen as justified from a purely consequentialist standard. Rawls (1955: 9–13) suggests this is more a caricature of consequentialism than an accurate portrayal. In short, assessing the justness of apparently extreme punishments from a consequentialist vantage requires accounting for punishments’ full social consequences. Hangings for petty theft would deter behaviors beyond the crimes applied to and such disproportionality can arguably inspire more severe crimes and weaken trust in the rule of law.4 These indirect effects would need to be accounted for and thus detract from the consequentialist reasons undergirding justified punishments. So long as the beneficial effects outweigh the direct costs and indirect social harms, punishments are in part justified on consequentialist grounds. But, this does not provide a unique reason for incarceration over other types of punishment. For incarceration to be specifically justified, it must be perceived as “economically efficient” as opposed to merely beneficial. Economic efficiency is a broader concept than mere cost benefit accounting or “technical efficiency” (Farrell 1957). Whereas a punishment is technically efficient insofar as it reaps benefits beyond harms, economic efficiency demands that the practice represents the best available benefit to cost ratio. To be economically efficient, incarceration would have to minimize the social costs of punishment relative to other viable punishment strategies or institutional norms. If incarceration yields benefits above costs, but an alternative scheme yields greater benefits above costs, then the difference in benefits between the two is effectively an additional imposed cost of incarceration. Economic inefficiency weakens a practices justification on consequentialist grounds, as it imposes unnecessary harms upon undeserving individuals. Furthermore, economic efficiency can provide a justification for punishment even if the consequences of a marginal unit exceed benefits. In other words, a punishment may impose net harms over benefits, but economic efficiency can demand that punishment norms, if presumed necessary for the foundation of social order, at least seek to minimize these harms. While economic efficiency is an obvious component of consequentialism, all popular justificatory frameworks are committed to economic efficiency insofar as they must provide justifications for the harms imposed by punishment and therefore must aim to minimize the harms of punishment amidst feasible alternatives. Retributivists argue punishment is justified because criminals deserve to suffer in proportion to the harms they impose (see: Moore 1997; Murphy 2007 for recent surveys). Retributions seek to correct the imbalance(s) of justice caused by crimes. Incarceration, at first, appeals to a variety of

Potentials, Limitations of De-Incarceration  313 retributive standards (see: Mundle 1969; Kleinig 1973; Davis 1983). Prisons effectively remove a criminal’s liberty akin to how victims suffer from crime. Prison sentences can also be proportionately gauged to accord with the different severities of crimes. But, other forms of punishment could also meet these criteria, such as house arrest, parole, or even corporal punishments. Imprisonment’s application to nonviolent or victimless crimes seems harder to justify with retributive reasons. Selling drugs may impose social costs, but drug dealers do not forsake the liberties of others in a comparable fashion as prison imposes. Thus, if punishments are to be justified on retributive grounds, the particular forms of punishment and the unique types of harms they cause will matter. Imprisonment may be justified for some types of crimes, but not for others. Hence, mass imprisonment outcomes are likely indicative of unjustified or excessive applications of prison sentences. The essential metric of justification from a retributive perspective is desert. Punishments are justified in so far as criminals deserve to suffer in proportion to the suffering they impose upon others. Hence, if any punishment practice imposes harms upon non-criminal members of society, such effects would need to be grounded upon good reasons. As social harms increase, so too would the justificatory burden. Hence, for imprisonment to be specifically justified via retributive reasons apart from other forms of punishment, some assessment regarding the comparative social costs of different norms is needed. Similarly, communicative frameworks provide little attempt to justify incarceration specifically. Also known as “moral expressionists,”5 communicative reasons claim particularly harsh penalties are needed to convey society’s unique disapproval of crimes relative to other immoral acts. Offenders perceive a unique message regarding the seriousness of crime when subjected to punishments rather than merely informal shaming (Tasioulas 2006; Bennett 2008). But, there is no specific communicative reason to suspect that severe punishments must take the form of prison sentences over other tactics. Punishment via moral communication is inherently linked to severity. If the sanctions across different crimes do not comport to their relative severities and harms imposed, then the moral messaging from society to criminals will be warped (von Hirsch 1993: 14–15). Insofar as any punishment type imposes harms upon non-criminal citizens, it would thus also convey a warped message throughout society regarding punitive deserts. Hence, the justificatory standard is again higher for punishment practices with greater social costs. This chapter does not adjudicate between competing punishment paradigms. Punishments may be justified or not. If not, then incarceration is certainly not justified and mass imprisonment even less so. If punishments are justified they may be so for different types of reasons. What punishment types are most justified and in practical terms how they might be implemented remains uncertain. All popular paradigms share a commitment to

314  Daniel J. D’Amico providing justified reasons for the imposed harms of criminal punishment. Greater social costs of punishment types require higher justification standards. Hence, to justify imprisonment apart from other punishment types requires some working knowledge of the potentials and limitations of alternative punishment types.

III. Contemporary Nation-States Without Mass Incarceration Empirical research describes in limited detail the cross-country patterns of crime and punishment (Mauer 1995, 2003; Newman 1999; Walmsley 2003, 2011). Nations with shared political, economic, legal, and or cultural institutions tend to harbor similar criminal justice systems and punishment outcomes (see: Cavadino and Dignan 2006a: 3–30, 2006b; Brodeur 2007; Lacey 2008: 3–55, 2012 for thorough surveys). Debate persists regarding what particular institutions or combinations shape imprisonment most and how. Clearly, some nations have experienced mass imprisonment in greater proportions than others, and some nations have effectively avoided it. Hence, it may be useful to investigate institutional commonalities and or punishment norms in those countries that lack mass incarceration outcomes. Given empirical limitations, more is known about what does not sufficiently explain the global patterns of incarceration than what conclusively does. Casual observation has emphasized a link from Western, liberal, market economies (Rusche and Kirschheimer 1939; Foucault 1975; Wilkins, Newton, and Steer 1991; Wacquant 1999, 2001; De Giorgi 2006) and electoral democracies (Savelsberg 1994, 1999, 2008; Sutton 2000) with mass incarceration. In contrast, more sophisticated analyses show no strong relationship between economic institutions or performance to greater crime and punishment trends (Neapolitan 2001; Sutton 2004; Ruddell 2005). In fact, economic freedom is positively related to crime reporting (Soares 2004a, 2004b) and negatively related to homicide rates (Stringham and Levendis 2010). Others claim Western societies lack social safety nets necessary for curtailing crime and prison growth. Collective labor interests (Cavadino and Dignan 2006a, 2006b; Downes and Hansen 2006) and welfare redistributions (Sutton 2000, 2004) correlate with lower prison populations and are thus proposed to limit the effects of vengeful public opinions (Flanagan and Longmire 1996; Enns 2014). Empirical verification is mixed. Mocan (2008, 2013) shows, in contrast, vengeful attitudes are incubated in high crime, low individualist, and low wealth contexts. Experimentally, punitive biases transcend demographic identities and political environments (Greene 2007; Houser et al. 2008; Guala 2012; Butler, Conzo, and Leroch 2013). Econometrics show democratic quality does not correlate with incarceration, and most cultural variables such as social capital do not appear relevant either. Similarly, factors representing labor organization, welfare spending, and other cultural factors like ethnic heterogeneity and trust lose significance

Potentials, Limitations of De-Incarceration  315 when other institutional factors are included (D’Amico and Williamson 2015). Some weak evidence suggests countries with lower prison populations have more male labor market participation (Sutton 2000, 2004), more judges, fewer homicides, more ethnic homogeneity, higher rates of Catholicism, less corruption, and a higher perception of law and order. D’Amico and Williamson (2015) show nations with smaller prison population rates tend to have civil legal origins, no death penalties, and fewer years under communism (see also: Ruddell 2005; Greenberg and West 2008; DeMichele 2013, 2014). Such factors dominate other variables, even seemingly relevant trends like homicides.6 Whereas common law countries like the US host greater incarceration rates, civil law nations like France, Germany, and Scandinavian countries leverage the more pervasive bureaucratic infrastructures therein via “fines, seizure of property, closing down of establishments, and community service (France), day-fines based upon the offender’s income, instruction, declaration of guilt without imposition, community service and probation (Germany), psychiatric treatment, probation, community service, treatment within the social services, and fines (Sweden) (Newman, Bouloukos, and Cohen 2002),” to monitor citizens’ behaviors and impose punitive sanctions without constructing or managing as many prison facilities. Such empirics do not prove any specific normative implications regarding social optimality (ibid.: 607). They merely provide an explanation for the observed fact that common law countries host larger prison populations than civil law nations. But, because prison expansion entails lower social costs under the common law, no specific justification is evident regarding precisely how much more incarceration under the common law is needed or optimal. Spamann (2015) for example, shows that the nations at the highest end of the imprisonment spectrum retain population rates beyond what can be accounted for by efficient institutional diversity.

IV. Punishments Without Formal States or Imprisonment While contemporary civil law nations rely less upon incarceration and more upon bureaucratic infrastructures, such mechanisms are essentially complements to imprisonment in a broader basket of criminal punishment types. Incarceration remains the default and standard form of criminal penalty around the globe. In fact, many philosophical definitions of the criminal law relative to civil liability hinge upon the potential of imprisonment. What is criminal is defined by what is subjected to incarceration (Barnett 1996; Duff 2013). Knowing what types of punishments and institutional norms are feasible without incarceration requires looking at a broader sample of social contexts than conventional nation-states. Ancient legal regimes, primitive tribes, and criminal organizations in the underground economy all possess customary as opposed to formal legal processes; none of which typically entail monopolized governmental authority,

316  Daniel J. D’Amico prison facilities, or punishments via incarceration (Leeson 2014a). Despite the absence of a formal government and monopolized prison powers, these contexts host a substantial degree of personal security, property rights enforcement, and predictable conflict resolution procedures. Hence, to conclude that without imprisonment, punishment processes cannot maintain a base level of social order is inaccurate. Social order is at least feasible without leveraging the specific threat of imprisonment.7 Punishments in relatively stateless contexts can be more violent than formal legal processes. Ancient customs such as the Lex Talionis, or “an eye for an eye,” put in literal terms the harshness of pre-modern punishments, often applied across wide ranges of behaviors, including relatively mild violations. Exiles were de facto death sentences, as life outside the collective meant subjection to constant predation. Similarly, tribal customs typically leverage strict protocols of ostracism, shame, expulsion, and even physical torture against seemingly venial breaches of patriarchal authority (Hoebel 1954). Criminal organizations similarly use violence and coercion to maintain power within their ranks and monopoly profits within their markets and territories (Sobel and Osoba 2009). Last, violent feuds were culturally demanded during interpersonal conflicts across groups (Oppenheimer 1913; Schafer 1965, 1970). The levers of violent enforcement outside the boundaries of formal state legitimacy have sharper and more viscerally severe blades (Seabright 2004; Pinker 2012). But, simply dismissing customary norms as brutish or archaic obfuscates many of their operational features.8 While potent, these pre-modern sanctions were likely less frequent than many perceive. Across cases, death penalties, exile, outlawry, and blood feuding were commonly accepted, but served as secondary mechanisms—final threats to compel suspected violators to comply with preliminary legal processes (Pollock and Maitland 1898: 29–69; Plucknett 1929: 307–14; Parisi 2001). Such procedures often converged on milder outcomes in reality as participants were inclined to seek mutually agreeable terms. Early legal systems often lacked any meaningful distinction between criminal and contractual law (Maine 1861; Zane 1927; Berman 1983). Even violence was adjudicated via civic rituals and interpersonal, but often publically performed, rituals of recompense and apology. Claimants were motivated to demand and accept tangible benefits against defendants they could prove responsible for causing harm. Suspects, in turn, were motivated to offer tangible value for compensation and thus avoid deadly or painful sanctions. Customary legal processes across historical, cultural, and geographic contexts tend to converge upon third party standards of arbitration (Posner 1980; Benson 1988), restitution norms (Barnett 1977; Schafer 1970), and similarly ranked scales of penalties (Sellin and Wolfgang 1964; Friedman 1979). To deviate too far from commonly accepted punishment quantities thus meant sacrificing one’s reputation and/or evoking subsequent liability (Stringham and Caplan 2008).9

Potentials, Limitations of De-Incarceration  317 Many seemingly obscure legal customs can be shown to have mitigated against social costs by effectively avoiding more disorderly alternatives. Such norms demonstrate a sort of “bounded” or “ecological” rationality (Simon 1991; Smith 2009) insofar as they represent least bad or cost minimization strategies. Leeson (2014b, 2014c, 2015), for example, explains how even apparently superstitious protocols like ordeals (Leeson 2012) or magical rituals (Leeson 2014a) served functional social roles to resolve conflict, salvage resources, protect order, and maximize group profits amidst communities where said superstitions were taken seriously and technologically superior alternatives were absent. What is perhaps most striking about such cases is the degree of specialization and adaptability observed therein. Akin to the contextual complementarity observed in animal societies relative to their geographic conditions and genetic reproductive traits (De Waal 1990), or the proprietary innovations of common pool resource dilemmas across contexts (Ostrom 1990), legal processes across non-traditional social settings demonstrate adaptability and complementarity to their local conditions. Property rights are less present and formally defined in environments of relative abundance (Demsetz 1967). Innovative security technologies can deter thefts and verify ownership when transportation is costly and information unreliable (Ellickson 1994; Anderson and Hill 2004). Nonviolent strategies emerge amidst a scarcity of manpower and strong group dependency on collective action. For Inuit tribes, for example, the death or injury of an adult male would have jeopardized the entire community. Hence, conflict resolution occurred through song and dance rituals (Hoebel 1954: 67–99).10 Across cases, contextual adaptability seems high in contrast to the uniformity of punishment via imprisonment across nation-states today. The stability of these norms within their respective contexts ultimately depends upon adaptability and thus informational feedback within their respective decision-making processes. Such systems typically existed within relatively decentralized networks of authority and thus possessed a degree of evolutionary feedback via competitive pressure. Without monopolized prison power, law and security providers in customary settings have to maintain the relative consent of their communities. Akin to Olson’s (1993) model of public good provision by state powers or Tiebout’s (1956) model of competitive governance, such incentives promote the adoptions of preferred legal standards, provide effective constraint upon the excessive growth of violent authority, and serve as a deterrent against the application of excessive or inaccurately gauged punishments.

V. Are Alternative Punishment Norms Feasible? Some patterns and consistencies can be recognized once mass imprisonment is situated within the broader context of these two sets. Whereas the variance of imprisonment across contemporary nation-states is most strongly

318  Daniel J. D’Amico correlated with variables like legal origins and years under communism, such are essentially proxies for the organizational structure of the institution in which legal decision making takes place (Djankov et al. 2003a). In short, the common law has long been accredited as more economically efficient than civil law because of its decentralized properties (Hayek 1960). By conveniently evolving amidst historical conditions of dispersed authority, common law territories hosted competitive incentives for the innovation and adoption of legal norms that preserved and promoted trade and efficiently allocated resources (Glaeser and Shleifer 2002; Djankov et al. 2003b). Thus, economic performance tended to accumulate and thrive more in common law economic spheres (La Porta, Lopez-de-Silanes, and Shleifer 1997, 1998). Inversely, a similar relationship between organizational centralization and mass imprisonment can be recognized. In short, greater rates of incarceration occur in stride with more hierarchically organized criminal justice decision-making patterns. Communism was a sort of ultimate expression of centralized authority, wherein all matters of society were subjected to the coercive authority of state control (Kornai 1992). While civil law nations allowed greater rates of dictatorial power to inculcate economic decision making, similar cannot be said regarding criminal justice processes therein. While civil legal processes are more codified in both the economic and criminal legal sectors than under the common law, local authorities possess a stronger and more focused discretionary authority to dismiss criminal charges (Merryman 1969; Borricand 1993). The French adopted stronger protections of individual rights in the wake of their violent revolutionary experiences (Farrand 1901; Stuntz 2011: 74–9). And France continued an explicit campaign of decentralized criminal legal power in the wake of WWII (Donovon 2010). Similarly, the vast network of bureaucratic interest groups effectively decentralized criminal enforcement as jurisdictions fought to retain local power once in place (Roche 2007). In contrast, criminal justice procedures under the common law have concentrated in more hierarchical patterns with greater dominance of federal authority since the latter twentieth century (Murakawa 2014; Boettke, Lemke, and Palagashvili 2016). In short, decentralization was crucial for economic growth, while more hierarchically structured criminal justice systems fostered more mass imprisonment. If mass imprisonment is accurately perceived as a form of social inefficiency or institutional failure, these would seem consistent observations. Viewed in isolation, the punitive techniques of relatively stateless social orders are not particularly desirable for modern liberal societies. But it remains unclear as to how intrinsic these qualities are, or what particular feature such undesirability stems from. It is unclear if such norms would operate differently amidst alternative exogenous conditions, devoid of isolation, underdevelopment, or illicitness. In contrast, their particular features of adaptability seem similarly a byproduct of institutional feedback fostered

Potentials, Limitations of De-Incarceration  319 amidst relative decentralization and its associated degree of competitive incentives. Again, this seems a consistent observation. Are the punitive techniques leveraged by nation-states that avoid mass incarceration a viable reform strategy in those contexts currently experiencing mass incarceration? In short, not without costs. As La Porta, Lopez-de-Silanes, and Shleifer (2008: 309) explain, “courts or legislators in a country might bring into one domain a set of tools that has been used in another, based on either philosophical outlook or a desire for consistency, with adverse results.” The common law’s greater economic performance is partially accredited to the absence of more pervasive state bureaucracies and the rent-seeking interests they foster (La Porta, Lopez-de-Silanes, and Shleifer 1997, 1998). Though there is no direct empirical link between economic performance and prison population rates, the institutional resources leveraged within civil law countries, if mimicked in common law countries, would carry large start-up costs and likely sacrifice future economic performance. But, that is not to say that no feasible opportunities for reform are possible. In short, the greatest opportunities would seem to stem from the lack of observed correlation between economic performance and prison populations in general. In other words, common law countries with high economic performance could reduce their prison population rates without sacrificing prosperity via endogenous institutional innovation. Similarly, we should expect civil law countries to possibly improve their economic performances without necessarily growing their prison populations provided they maintain effective checks and balances against prison growth. One potentially feasible avenue for reform is to preserve, protect, and expand those institutional features that promote innovation, adaptability, and general evolutionary effectiveness into the social arenas where they are systemically lacking and thus correlated with mass incarceration. In the economic spheres across societies, decentralization goes hand in hand with good economic performance. In the criminal justice spheres across societies, mass imprisonment correlates with centralization. Hence, effective reform may depend upon the application and preservation of decentralization within common law criminal justice systems. How decentralization might be promoted and preserved in common law criminal justice spheres remains uncertain.

Conclusions To justify mass imprisonment outcomes requires some preliminary justification for punishment by imprisonment apart from other norms and techniques. All popular justificatory frameworks inherently concede that the justificatory burden of a criminal punishment or punitive institutional scheme increases if it imposes additional social harms. Hence, to justify imprisonment apart from other punitive types requires comparative cost

320  Daniel J. D’Amico benefit analysis. To inform such an analysis this paper investigated two sets of social contexts that conveniently lacked mass incarceration outcomes. The general observations from those contexts are straightforward. Mass incarceration occurs in stride with more hierarchically structured and centrally organized criminal justice institutions. Communist regimes are the most potent expression of criminal justice centralization, followed by contemporary common law nation-states, then civil law nation-states, and finally relatively stateless social orders. If mass imprisonment outcomes are to be interpreted as evidentiary examples of excessive applications of prison sentences, then punishment philosophy ought to take account of systemic relationships between institutional organization and prison outcomes. Few if any of the observed alternative norms apart from mass imprisonment seem viable as reform strategies within contemporary mass imprisonment contexts without substantial costs. Civil law bureaucracies cannot simply be imposed within common law countries. Hence, some degree of mass imprisonment outcomes may be understood as contextually efficient given a lack of viable alternatives. Some dynamics of institutional innovation and evolution are also revealed in the observation sets. Adaptability for institutions to comport to local conditions and constraints appears to be dependent upon the processes of institutional choice and exit fostered through local conditions of decentralization. Though punishment norms in relatively stateless social orders appear unjustifiable in visceral severity, they do succeed insofar as they represent institutional strategies of minimizing social costs within their unique environments of limited economic, technological, and scientific development. Interestingly, such insights stand in contrast to conventional perspectives of both punishment philosophy and political economy of crime and punishment. To close the gap between punishment theory and practice, we should not view incarceration as a default punitive technique in need of justification, nor as an inherently public good threatened by insufficient provision or management: the necessary and appropriate arena for state redistribution and subsidy. Instead, real historical examples of incarceration have tended to be expressions of concentrated state power. Hence, a constitutional approach aimed at promoting and preserving local level authority and avoiding the concentration of state power via punitive authority may be warranted and needed.

Notes 1 See: Bedau (1972, 2015) and Davis (1972). 2 Duff (2013) surveys Smart (1973) and Bagaric and Amarasekara (2000) as consequentialists. Bentham (1830) is a classic advocate. 3 On proportionality see: Duff (2001), Ryberg (2004) and von Hirsch and Ashworth (2005). 4 “If the punishment is the same for simple theft, as for theft and murder, you give the thieves a motive for committing murder (Bentham 1830: 36).”

Potentials, Limitations of De-Incarceration  321 5 Duff (2013) surveys Duff (2001), Bennett (2008), Markel (2011) as communicative theories. Greiff (2002) borrows Feinberg’s (1970) “expressionist” label, Hampton (1984) and Murphy (1985) use the term “moral education.” Though minor differences exist across these terms, they are functionally synonymous for our purposes. 6 In general and with a variety of control variables and robustness checks, “a one standard deviation increase in homicide rates (about 11 per 100,000) increases the number of prisoners by about 60 (per 100,000) (D’Amico and Williamson 2015: 600). . . French, German, and Scandinavian legal origins reduce non-drug related inmates by about 58, 107, and 113, per 100,000 respectively. Communism increases non-drug related inmates by about 279 per 100,000 (ibid. p. 607).” 7 Friedman (1979), Benson (1988), Leeson (2011, 2012) argue customary legal practices were more efficient than typically recognized. Boonin (2008) argues cases like these suffice to refute the necessary presumption of state sponsored criminal justice provision. 8 It also contributes to the difficulty of recognizing the failures of modern imprisonment. Foucault (1975) argued similarly that the modern shift from punishments via public spectacles and physically corporeal sanctions towards the private spaces of jails and penitentiaries, though motivated by humanitarian rhetoric, allowed for a capture and imposition of state power and interests. Rejali (2000) explains the modern rise of unique torture practices by advanced democracies in similar terms. 9 Similarly consistent incentive structures mitigate the escalation of violence amidst networks of criminal gangs. Within groups, gangs seek to avoid the costs of violent conflict by imposing internal governance norms and procedures (Leeson 2007; Skarbek 2010, 2011, 2012). Leeson (2009: 82–106) also highlights how credibly signaling extreme violence can be an effective mechanism to conserve violent effort (see also: Gambetta 1996, 2011; Kaminski 2004). Buchanan (1972) even argues that monopolization of criminal markets can reduce violent conflict and the net level of criminal behavior. 10 Further examples abound. Pirates, similarly, had strict rules to deter the use of firearms aboard ship, where a bullet to the hull risked the group’s livelihood (Leeson 2008). D’Amico (2010) notes that searching for stolen property in Ancient Greece demanded the accuser be naked to avoid the potentials of planted evidence. Kaminski (2004) explains, in prison confines social prestige is often afforded to individuals with productive value. Given resource scarcity humor can itself be a productive trait. Elaborate rule enforcements also exist to coordinate flatulence given the extreme scarcity of fresh air. And, Friedman (2005) outlines seemingly ingenious techniques devised to assure secure delivery of valuable assets in ancient China.

References Anderson, Terry and Hill, Peter (2004). The Not So Wild, Wild West: Property Rights on the Frontier. Redwood City, CA: Stanford University Press. Bagaric, Mirko and Amarasekara, Kumar (2000). “The Errors of Retributivism.” Melbourne University Law Review 24: 1–66. Barnett, Randy (1977). “Restitution: A New Paradigm of Criminal Justice.” Ethics 87: 279. Barnett, Randy (1996). “Getting Even: Restitution, Preventive Detention, and the Crime-Tort Distinction.” Boston University Law Review 76: 157.

322  Daniel J. D’Amico Bedau, Hugo (1972). “Penal Theory and Prison Reality Today.” Juris Doctor 2: 40–3. Bedau, Hugo (2015). “Punishment.” In Stanford Encyclopedia of Philosophy. Available: http://plato.stanford.edu/entries/punishment/ Bennett, Christopher (2008). The Apology Ritual: A Philosophical Theory of Punishment. Cambridge: Cambridge University Press. Benson, Bruce (1988). “Legal Evolution in Primitive Societies.” Journal of Institutional and Theoretical Economics 144: 772–88. Bentham, Jeremy (1830). The Rationale of Punishment. Honolulu, HI: University of Honolulu Press. Berman, Harold (1983). Law and Revolution: The Formation of the Western Legal Tradition. Cambridge, MA: Harvard University Press. Berman, Mitchell (2008). “Punishment and Justification.” Ethics 118: 258–90. Boettke, Peter, Lemke, Jayme and Palagashvili, Liya (2016). “Re-Evaluating Community Policing in a Polycentric System.” Journal of Institutional Economics 12: 305–25. Boonin, David (2008). The Problem of Punishment. Cambridge: Cambridge University Press. Borricand, Jacques (1993). “France.” World Factbook of Criminal Justice Systems. Bureau of Justice Statistics. Brodeur, Jean-Paul (2007). “Comparative Penology in Perspective.” Crime and Justice 36: 49–91. Buchanan, James (1972). “Defense of Organized Crime.” In The Economics of Crime and Punishment, ed. S. Rottenberg. Washington, DC: American Enterprise Institute. Butler, Jeffrey, Conzo, Pierluigi and Leroch, Martin (2013). “Social Identity and Punishment.” EIEF Working Paper 16/13. Cavadino, Michael and Dignan, James (2006a). Penal Systems: A Comparative Approach. Los Angeles, CA: Sage. Cavadino, Michael and Dignan, James (2006b). “Penal Policy and Political Economy.” Criminology & Criminal Justice 6: 435–56. D’Amico, Daniel (2010). “The Prison in Economics: Private and Public Incarceration in Ancient Greece.” Public Choice 145: 461–82. D’Amico, Daniel and Williamson, Claudia (2015). “Do Legal Origins Affect Cross-Country Incarceration Rates?” Journal of Comparative Economics 43: 595–612. Davis, Angela (1972 [2003]). Are Prisons Obsolete? New York City: Seven Stories Press. Davis, Michael (1983). “How to Make the Punishment Fit the Crime.” Ethics 93: 726–52. De Giorgi, A lessandro (2006). Rethinking the Political Economy of Punishment: Perspectives on Post-Fordism and Penal Politics. Aldershot: Routledge. De Waal, Frans (1990). Peacemaking Among Primates. Cambridge, MA: Harvard University Press. DeMichele, Matthew (2013). “Using Weber’s Rechtssoziologie to Explain Western Punishment: A Typological Framework.” European Journal of Crime and Criminal Law and Criminal Justice 21: 85–109. DeMichele, Matthew (2014). “A Panel Analysis of Legal Culture, Political Economics, and Punishment Among 15 Western Countries, 1960–2010.” International Criminal Justice Review 24(4): 1–17.

Potentials, Limitations of De-Incarceration  323 Demsetz, Harold (1967). “Toward a Theory of Property Rights.” The American Economic Review 57: 347–59. Djankov, Simeon, La Porta, Rafael., Lopez-de-Silanes, Florencio and Shleifer, Andrei (2003a). “Courts.” Quarterly Journal of Economics 118: 453–518. Djankov, Simeon, Glaeser, Edward, La Porta, Rafael, Lopez-de-Silanes, Florencio and Shleifer, Andrei. (2003b). “The New Comparative Economics.” Journal of Comparative Economics 31: 595–619. Donovon, James. (2010). Juries and the Transformation of criminal Justice in France in the Nineteenth and Twentieth Centuries. Chapel Hill, NC: University of North Carolina Press. Downes, David and Hansen, Kirstine (2006). “Welfare and Punishment in Comparative Perspective.” In Perspectives on Punishment: The Contours of Control, ed. Sarah Armstrong and Lesley McAra. Oxford: Oxford University Press, 133–54. Duff, Antony (2001). Punishment, Communication, and Community. Oxford: Oxford University Press. Duff, Antony (2013). “Legal Punishment.” In Stanford Encyclopedia of Philosophy. Available at http://plato.stanford.edu/entries/legal-punishment/ Ellickson, Robert (1994). Order Without Law: How Neighbors Settle Disputes. Cambridge, MA: Harvard University Press. Enns, Peter. (2014). “The Public’s Increasing Punitiveness and Its Influence on Mass Incarceration in the United States.” American Journal of Political Science 58(4): 857–872. Farrand, Max (1901). The Declaration of the Rights of Man and of Citizens: A Contribution to Modern Constitutional History. New York City: Henry Holt and Company. Farrell, M.J. (1957). “The Measurement of Productive Efficiency.” Journal of the Royal Statistical Society, Series A 120: 253–90. Feinberg, Joel (1970). “The Expressive Function of Punishment.” The Monist 49: 397–423. Flanagan, Timothy and Longmire, Dennis (eds.) (1996). Americans View Crime and Justice, a National Public Opinion Survey. Los Angeles, CA: Sage. Foucault, Michel (1975). Discipline and Punish: The Birth of the Prison. New York City: Vintage Books. Friedman, David (1979). “Private Creation and Enforcement of Law—a Historical Case.” Journal of Legal Studies 8: 399–415. Friedman, David (2005). “From Imperial China to Cyberspace: Contracting Without the State.” Journal of Law, Economics & Policy 1: 349–69. Gambetta, Diego (1996). The Sicilian Mafia: The Business of Private Protection. Cambridge, MA: Harvard University Press. Gambetta, Diego (2011). Codes of the Underworld: How Criminals Communicate. Princeton, NJ: Princeton University Press. Garland, David (ed.) (2001). Mass Imprisonment: Social Causes and Consequences. Los Angeles, CA: Sage. Greenberg, David and West, Valerie (2008). “Siting the Death Penalty Internationally.” Law & Social Inquiry 33: 295–343. Greene, Joshua (2007). “The Secret Joke of Kant’s Soul.” In Moral Psychology, Vol. 3: The Neuroscience of Morality: Emotion, Disease, and Development, ed. Sinnott-Armstrong. Boston, MA: MIT Press. Greiff, Pablo (2002). “Deliberative Democracy and Punishment.” Buffalo Criminal Law Review 5: 373–403.

324  Daniel J. D’Amico Guala, Francesco (2012). “Reciprocity: Weak or Strong? What Punishment Experiments Do and Do Not Demonstrate.” Behavioral and Brain Sciences 35: 1–15. Hampton, Jean (1984). “The Moral Education Theory of Punishment.” Philosophy & Public Affairs 13: 208–38. Hayek, Friedrick. (1960). The Constitution of Liberty. Chicago: University of Chicago Press. Hoebel, E. Adamson (1954). The Law of Primitive Man: A Study in Comparative Legal Dynamics. Cambridge, MA: Harvard University Press. Houser, Daniel, Xiao, Erte, McCabe, Kevin and Smith, Vernon (2008). “When Punishment Fails: Research on Sanctions, Intentions and Non-Cooperation.” Games and Economic Behavior 62: 509–32. Kaminski, Marek (2004). Games Prisoners Play: The Tragicomic Worlds of Polish Prison. Princeton, NJ: Princeton University Press. Kant, Immanuel. (1781 [2008]). Critique of Pure Reason. London: Penguin Classics. Kleinig, John (1973). Punishment and Desert. Boston, MA: Martinus Nijhoff. Kornai, Janos (1992). The Socialist System: The Political Economy of Communism. Princeton, NJ: Princeton University Press. La Porta, Rafael, Lopez-de-Silanes, Florencio, Shleifer, Andrei and Vishny, Robert (1997). “Legal Determinants of External Finance.” Journal of Finance 52: 1131–50. La Porta, Rafael, Lopez-de-Silanes, Florencio and Shleifer, Andrei. (2008). “The Economic Consequences of Legal Origins.” Journal of Economic Literature 46: 285–332. La Porta, Rafael, Lopez-de-Silanes, Florenico, Shleifer, Andrei and Vishny, Robert. (1998). “Law and Finance.” Journal of Political Economy 106: 1113–55. Lacey, Nicola. (2008). The Prisoners’ Dilemma: Political Economy and Punishment in Contemporary Democracies. Cambridge: Cambridge University Press. Lacey, Nicola (2012). “Political Systems and Criminal Justice: The Prisoners’ Dilemma After the Coalition.” Current Legal Problems 65: 1–37. Leeson, Peter (2007). “An-arrgh-chy: The Law and Economics of Pirate Organization.” Journal of Political Economy 115: 1049–94. Leeson, Peter (2009). The Invisible Hook: The Hidden Economics of Pirates. Princeton, NJ: Princeton University Press. Leeson, Peter (2011). “Trial by Battle.” Journal of Legal Analysis 3: 341–75. Leeson, Peter (2012). “Ordeals.” Journal of Law and Economics 55: 691–714. Leeson, Peter (2014a). “Pirates, Prisoners, and Preliterates: Anarchic Context and the Private Enforcement of Law.” European Journal of Law and Economics 37: 365–79. Leeson, Peter (2014b). “Oracles.” Rationality and Society 26: 141–69. Leeson, Peter (2014c). Anarchy Unbound: Why Self-Governance Works Better than You Think. Cambridge: Cambridge University Press. Leeson, Peter and Coyne, Christopher (2012). “Sassywood.” Journal of Comparative Economics 40: 608–20. Maine, Henry (1861 [1906]). Ancient Law: Its Connection with the Early History and Its Relation to Modern Ideas. New York City: Henry Holt and Company. Markel, Dan (2011). “What Might Retributive Justice Be?” In Retributivism: Essays on Theory and Policy, ed. Mark White. Oxford: Oxford University Press, 49–72. Mauer, Marc (1995). “Americans Behind Bars: The International Use of Incarceration.” The Prison Journal 75: 113–23.

Potentials, Limitations of De-Incarceration  325 Mauer, Marc (2003). “Comparative International Rates of Incarceration: An Examination of Causes and Trends.” Presentation to U.S. Commission on Civil Rights. Merryman, John (1969). The Civil Law Tradition: An Introduction to the Legal Systems of Europe and Latin America, 3rd edition. Redwood City, CA: Stanford University Press. Mocan, Naci (2008). “Vengeance.” NBER Working Paper No. 14131. Mocan, Naci (2013). “Vengeance.” The Review of Economics and Statistics 95: 969–82. Moore, Michael (1997). Placing Blame: A Theory of Criminal Law. Oxford: Oxford University Press. Mundle, Clement (1969). “Punishment and Desert.” In The Philosophy of Punishment, ed. H. Acton. New York City: St. Martin’s Press. Murakawa, Naomi (2014). The First Civil Right: How Liberals Built Prison America. Oxford: Oxford University Press. Murphy, Jeffrie (1985). “Retributivism, Moral Education and the Liberal State.” Criminal Justice Ethics 4: 3–11. Murphy, Jeffrie (2007). “Legal Moralism and Retribution Revisited.” Criminal Law and Philosophy 1: 5–20. Neapolitan, Jerome (2001). “An Examination of Cross-National Variation in Punitiveness.” International Journal of Offender Therapy and Comparative Criminology 45: 691–710. Newman, Graeme (ed.) (1999). Global Report on Crime and Justice. Oxford: Oxford University Press. Newman, Graeme, Bouloukos, Adam and Cohen, Debra (2002). “Introduction” World Factbook of Criminal Justice Systems. Washington, DC: Bureau of Justice Statistics. Olson, Mancur (1993). “Dictatorship, Democracy, and Development.” The American Political Science Review 87: 567–76. Oppenheimer, Heinrich (1913 [1975]). The Rationale of Punishment. Hudson: Patterson Smith. Ostrom, Elinor (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press. Parisi, Francesco (2001). “The Genesis of Liability in Ancient Law.” American Law and Economics Review 3: 82–124. Pinker, Steven (2012). The Better Angels of Our Nature: Why Violence Has Declined. New York City: Viking Books. Plucknett, Theodore (1929 [2010]). A Concise History of the Common Law. Indianapolis, IN: Liberty Fund. Pollock, Frederick and Maitland, Frederic (1898 [1968]). The History of English Law Before the Time of Edward I. Volumes I and II. Indianapolis, IN: Liberty Fund. Posner, Richard (1980). “A Theory of Primitive Society, with Special Reference to Law.” The Journal of Law and Economics 23: 1–53. Rawls, John (1955). “Two Concepts of Rules.” Philosophical Review 65: 3–32. Rawls, John (1993). Political Liberalism. New York City: Columbia University Press. Rejali, Darius (2000). Torture and Democracy. Princeton, NJ: Princeton University Press. Roche, Sebastian (2007). “Criminal Justice Policy in France: Illusions of Severity.” Crime and Justice 36: 471–550.

326  Daniel J. D’Amico Ruddell, Rick (2005). “Social Disruption, State Priorities, and Minority Threat.” Punishment and Society 7: 7–28. Rusche, Georf and Kirschheimer, Otto (1939 [2003]). Punishment and Social Structure. New Brunswick, NJ: Transaction Publishers. Ryberg, Jesper (2004). The Ethics of Proportionate Punishment, a Critical Investigation. Dordrecht: Kluwer Publishers. Savelsberg, Joachim (1994). “Knowledge, Domination, and Criminal Punishment.” American Journal of Sociology 99: 911–43. Savelsberg, Joachim (1999). “Knowledge, Domination and Criminal Punishment Revisited: Incorporating State Socialism.” Punishment & Society 1: 45–70. Savelsberg, Joachim (2008). “Punitiveness in Cross-National Comparison: Toward a Historically and Institutionally Founded Multi-Factorial Approach.” In International Perspectives on Punitivity, ed. H. Kury and T. Ferdinand. Bochum: Universitätsverlag Brockmeyer, 13–31. Schafer, Stephen (1965). “Restitution to Victims of Crime—an Old Correctional Aim Modernized.” Minnesota Law Review 243–54. Schafer, Stephen (1970). Compensation and Restitution to Victims of Crime. Hudson: Patterson Smith. Seabright, Paul (2004). The Company of Strangers: A Natural History of Economic Life. Princeton, NJ: Princeton University Press. Sellin, Johan and Wolfgang, Marvin (1964). The Measurement of Delinquency. San Francisco, CA: John Wiley & Sons. Simon, Herbert. (1991). “Bounded Rationality and Organizational Learning.” Organization Science 2: 125–34. Smart, John (1973). “An Outline of a System of Utilitarian Ethics.” In Utilitarianism for and Against, ed. John Smart and Bernard Williams. Cambridge: Cambridge University Press. Smith, Vernon (2009). Rationality in Economics: Constructivist and Ecological Forms. Cambridge: Cambridge University Press. Soares, Rodrigo (2004a). “Crime Reporting as a Measure of Institutional Development.” Economic Development and Cultural Change 52(4): 851–71. Soares, Rodrigo (2004b). “Development, Crime and Punishment: Accounting for the International Differences in Crime Rates.” Journal of Development Economics 73: 155–84. Sobel, Russell and Osoba, Brian (2009). “Youth Gangs as Pseudo-Governments: Implications for Violent Crime.” Southern Economic Journal 75: 996–1018. Spamann, Holger (2015). “The U.S. Crime Puzzle: A Comparative Perspective on U.S. Crime and Punishment.” American Law and Economics Review 18: 33–87. Stringham, Edward and Caplan, Bryan. (2008). “Privatizing the Adjudication of Disputes.” Theoretical Inquiries in Law 9: 503–28. Stringham, Edward and Levendis, John (2010). “Chapter 6: The Relationship Between Economic Freedom and Homicide.” In Economic Freedom of the World: Annual Report, ed. James Gwartney, Josh Hall and Robert Lawson. Vancouver: Fraser Institute, 203–18. Stuntz, William (2011). The Collapse of American Criminal Justice. Cambridge, MA: Harvard University Press. Sutton, John (2000). “Imprisonment and Social Classification in Five Common-Law Democracies, 1955–1985.” American Journal of Sociology 106: 350–86.

Potentials, Limitations of De-Incarceration  327 Sutton, John (2004). “The Political Economy of Imprisonment in Affluent Western Democracies, 1960–1990.” American Sociological Review 69: 170–89. Tasioulas, John (2006). “Punishment and Repentance.” Philosophy 81: 279–322. Tiebout, Charles.(1956). “A Pure Theory of Local Expenditures.” Journal of Political Economy 64: 416–24. Von Hirsch, Andrew (1993). Censure and Sanctions. Oxford: Clarendon Press. Von Hirsch, Andrew and Ashworth, Andrew (eds.) (2005). Proportionate Sentencing: Exploring the Principles. Oxford: Oxford University Press. Wacquant, Loic (1999). “ ‘Suitable Enemies’ Foreigners and Immigrants in the Prisons of Europe.” Punishment and Society 1: 215–22. Wacquant, Loic (2001). “Deadly Symbiosis: When Ghetto and Prison Meet and Mesh.” Punishment and Society 3: 95–134. Walmsley, Roy (2003). “Global Incarceration and Prison Trends.” Forum on Crime and Society 3: 65–78. Walmsley, Roy (2011). World Prison Population List, 9th Edition. London: International Centre for Prison Studies. Wilkins, William, Newton, Phyllis and Steer, John (1991). “The Sentencing Reform Act of 1984: A Bold Approach to the Unwarranted Sentencing Disparity Problem.” Criminal Law Forum 2: 359–62. Zane, John (1927 [1998]). The Story of Law, 2nd ed. Indianapolis, IN: Liberty Fund.

Contributors

Eyal Aharoi is an assistant professor of psychology, philosophy, and neuroscience at Georgia State University. His current research draws on interdisciplinary approaches to understanding the ways in which extra-rational factors, such as emotion, influence decision-making in legal, criminal, interpersonal, and policy domains. Chris Barker is an assistant professor of political science at Southwestern College. He has published articles on nineteenth-century thought in History of European Ideas; American Political Thought; and Law, Culture, and Humanities. He recently completed his first book manuscript on J. S. Mill’s political writings. He has previously held positions at Ohio University, Boston College, and Harvard University. Kristen Bell is a Senior Liman Fellow in Residence at Yale Law School. Her dissertation, “Mercy and Criminal Justice,” explores mercy and the sense in which conditions of social injustice vitiate the moral justification for criminal punishment. As a former Soros Justice Fellow, Bell has advocated on behalf of prisoners serving life sentences for juvenile convictions. David Boonin is a professor of philosophy at the University of Colorado at Boulder. He is the author of Thomas Hobbes and the Science of Moral Virtue (Cambridge University Press 1994), A Defense of Abortion (Cambridge University Press 2003), The Problem of Punishment (Cambridge University Press 2008), Should Race Matter? (Cambridge University Press 2011), and The Non-Identity Problem and the Ethics of Future People (Oxford University Press 2014). Jason Brennan is the Flanagan Family Chair of Strategy, Economics, Ethics, and Public Policy at Georgetown University. He is the author of seven books, including Against Democracy (Princeton University Press, 2016) and Markets without Limits, with Peter Jaworski (Routledge, 2015). Andrew Cohen is an associate professor of philosophy at Georgia State University. He is the author of Toleration (Polity, 2014) as well as numerous papers on individualism, exchange, toleration, and waste.

330 Contributors Daniel J. D’Amico is a lecturer in economics at Brown University, where he is the associate director of the Political Theory Project. His research is broadly focused upon the causes and consequences of incarceration around the world and over time. Josh Dohmen is a lecturer in philosophy at the University of West Georgia. His research interests include feminist philosophy and twentieth century continental philosophy, especially in relation to ethical and political concerns about disability, prisons, education, and labor. Cara H. Drinan is a professor of law at the Catholic University of America Columbus School of Law. She is the author of numerous law review articles and the forthcoming book The War on Kids: How American Juvenile Justice Lost Its Way (Oxford University Press). William Glod is a program officer of philosophy at the Institute for Humane Studies at George Mason University. He has published articles in Southern Journal of Philosophy, Social Theory and Practice, and Ethical Theory and Moral Practice. Nathan Goodman is a graduate research fellow at the Mercatus Center at George Mason University. He has previously worked as the Lysander Spooner Research Scholar in Abolitionist Studies at the Center for a Stateless Society and as a Fellow at the Fully Informed Jury Association. Abigail R. Hall is an assistant professor in economics at the University of Tampa in Tampa, Florida, a Research Fellow with the Independent Institute, and an Affiliated Scholar with the Mercatus Center at George Mason University and the Foundation for Economic Education. John Hasnas is a professor of business at Georgetown’s McDonough School of Business, a professor of law (by courtesy) at Georgetown University Law Center, in Washington, DC, and the executive director of the Georgetown Institute for the Study of Markets and Ethics. His scholarship concerns ethics and white-collar crime, jurisprudence, and legal history. Michael Huemer is a professor of philosophy at the University of Colorado at Boulder. He is the author of more than 50 academic articles in ethics, epistemology, political philosophy, and metaphysics, as well as four books: Skepticism and the Veil of Perception (Rowman and Littlefield 2001), Ethical Intuitionism (Palgrave Macmillan 2005), The Problem of Political Authority (Palgrave Macmillan 2013), and Approaching Infinity (Palgrave Macmillan 2016). Robert C. Hughes is an assistant professor in the legal studies and business ethics at the Wharton School of the University of Pennsylvania. His work in legal philosophy focuses on the relationship between law and coercion, the moral obligations to obey unenforced and under-enforced laws, and issues of justice in medical care and research.

Contributors  331 Johanna C. Luttrell is a visiting scholar in political philosophy at the Hobby School of Public Affairs of the University of Houston. Her work in political philosophy centers around issues in human rights, transnational feminism, and philosophy of race. Veronica J. Mercier is a student at the University of Tampa majoring in economics and math. Eddy Nahmias is a professor of philosophy and neuroscience at Georgia State University. His research focuses on free will and moral responsibility, as well as how these issues are informed by modern science. He is the co-editor of Moral Psychology: Historical and Contemporary Readings (Wiley-Blackwell 2010). John F. Pfaff is a professor of law at Fordham Law School in New York City. He is the author of Locked In: The Real Causes of Mass Incarceration—And How to Achieve Real Reform, published in 2017 by Hachette Book Group. Chris W. Surprenant is an associate professor of philosophy at the University of New Orleans, where he directs the Alexis de Tocqueville Project. He is the author of Kant and the Cultivation of Virtue (Routledge 2014), and co-editor of Kant and Education: Interpretations and Commentary (Routledge 2011) and Kant and the Scottish Enlightenment (Routledge 2017). Bill Wringe is an assistant professor of philosophy at Bilkent University in Ankara, Turkey. He has published widely on a range of topics in ethics, political philosophy, and the philosophy of mind, including collective intentionality and obligation, punishment, emotion, and perceptual states. His book, An Expressive Theory of Punishment, was published by Palgrave Macmillan in 2016.

Index

Annie E. Casey Foundation 202 – 5 apology 147 – 57, 168, 188, 310, 316 asset forfeiture 210, 216 – 21, 234 Audre Lorde Project 58, 63 authoritarianism 60 – 1, 285, 290 Barnett, Randy 128, 131, 133 – 4 Bastiat, 243 Bastoy Island 98 – 9 Beccaria, Cesare 69 – 75, 77 – 9, 81 – 3 behavior modification 147, 236 Benatar, David 304 Black Lives Matter 279 Bloomington School 58, 62 – 3 Bradley, Kelly 35 – 6 broken windows 225 Brooks, Thom 5 – 6 Bundy, Ted 51 – 2 Busta Rhymes see Smith, Trevor Byrd, Sharon 10, 103 clemency 9, 78 communitarianism 277 – 80, 284 – 6 community policing 52, 55 – 7 Crime Victims’ Rights Act of 2004 63 criminal intent 23, 25 – 6, 37, 100, 107, 125, 128, 130, 146, 149 – 56, 188 Department of Justice (DOJ) 209, 216 – 19 doctrine of double effect 37 Duff, Antony 106, 109, 116, 145, 147, 157, 166, 169, 173 – 5, 300 – 1, 311, 315

Federal Office of the Pardon Attorney 78 Federalism 56 – 7 Feinberg, Joel 15 – 7, 105 – 7, 109 – 10, 146, 226, 239, 321 forgiveness 27, 150, 156, 187, 189 Foss, Adam 245, 253 – 4 Foucault, Michel 82, 265 – 7, 275, 314, 321 gangs 54, 58, 173, 182, 189, 228, 234, 282 – 4, 290, 296, 321 Glasgow, Joshua 108 Graham v. Florida 195, 200 Grimes, Jacobia 1 Guantánamo Bay 79 Hamilton, Alexander 50 Hammurabi’s Code 3 – 4 Harrell, Samuel 35 – 6 Hart, H.L.A. 4 – 5 Herman, Barbara 93, 102 Hobbes, Thomas 11, 19, 23 Hynes, Joseph 247 incarceration: alternatives to 41 – 2, 278 – 87, 290 – 1, 294 – 307; incarceration, benefits of 34 – 5, 43 – 5; incarceration, costs of 34 – 9, 53, 61, 69, 105, 137, 166, 201, 205, 213, 215 – 6, 238, 245, 256 Iowa v. Lyle 199 – 200

Eighth Amendment 195, 201, 206 entrepreneurship 57 – 9, 102

Judeo-Christian values 17, 29 juvenile punishment 187 – 9, 194 – 207, 294

Fay, Michael 294 Federal Bureau of Investigation (FBI) 247

Kant, Immanuel 7 – 13, 17, 69 – 73, 90 – 1, 93 – 8, 102, 304 – 5, 309 Kennedy, Anthony 195

334 Index leniency 55, 69 – 72, 74 – 5, 221, 255 – 6, 265 Lippke, Richard 116, 230, 239 Locke, John 19 Madison, James 50 Maldives 278 – 80, 284, 286 – 8 mandatory minimums 55 – 6 Maricopa County, Arizona 296 Marx, Karl 188 McDonald, Laquan 246 Mill, James 79 Mill, John Stuart 16, 18, 70 Miller v. Alabama 196 Missouri Model 202 – 5 Model Penal Code of the United States 5, 146 moral education 90, 98 – 9, 102, 321 Moskos, Peter 297 – 9, 301 National Lawyers Guild 44 New Jersey Supreme Court 249 New Orleans 1, 79, 82 Nixon, Richard 210, 212 – 3 Norman Conquest 21, 60 – 1 Nozick, Robert 89, 114 Ostrom, Elinor and Vincent 49 – 52, 58, 62 – 4 pardons 77 – 8 parole 1, 27, 76, 141, 157, 162 – 77, 186 – 91, 194 – 6, 199, 201, 206, 213, 221, 227, 239, 249, 253, 266, 313 paternalism 225 – 7, 229 – 34, 236 – 8, 298, 305 Paul, Rand 220 plea bargaining 56, 70, 76 – 7, 157, 242 – 57 Potter, Nelson 72 prison labor 270, 273 – 4 private prisons 3, 26, 180, 209 – 16, 218, 220 – 1 punishment: communicative theory of 9, 99, 106, 109 – 10, 112 – 8, 144 – 51, 153, 155 – 7, 164, 166 – 73, 300 – 1, 313; deterrence theory of 4 – 10, 16, 20, 23 – 5, 27 – 9, 34 – 5, 41 – 3, 50, 60, 70 – 2, 74, 77, 80, 82, 98 – 9, 105, 107, 124, 132, 134, 144 – 6, 151, 155, 164, 166 – 72, 174, 184 – 6, 190, 217, 229, 235 – 6,

254, 298, 299 – 300, 302, 312, 317; expressivist theory of 105 – 19, 144, 146, 151, 282; retributivist theory of 3 – 12, 15, 18, 20, 35, 38, 42, 46, 70 – 1, 75, 82 – 3, 90, 98, 107, 125 – 30, 145 – 7, 149, 151, 164 – 73, 254, 298 – 9, 302, 311 – 3 punishment drift 230, 232, 239, 241 Ramirez, Israel 53 Rawls, John 4 – 5, 102, 311 – 2 Reagan, Ronald 213 recidivism 144, 151, 155, 168, 171, 173, 186 – 7, 203 – 4, 229, 300, 302 rehabilitation 3 – 6, 15, 27, 34, 39, 43, 74, 80, 82 – 3, 89, 144 – 7, 151, 157, 162, 164, 167 – 9, 171 – 3, 176 – 7, 184, 195 – 6, 200 – 3, 213 – 4, 230, 237, 240, 254, 265, 284, 294, 298 – 300, 302, 306 – 7 restitution 17 – 8, 22 – 4, 27 – 8, 35, 41 – 3, 45, 59 – 61, 83, 122, 126 – 41, 166, 182, 298, 301 – 2, 310, 316 restorative justice 5 – 6, 27, 146, 149, 156 – 7, 164, 174 – 7, 180 – 91, 282, 301 Ripstein, Arthur 10 – 1 Roper v. Simmons 195 school-to-prison pipeline 180 – 1, 185, 187 Seychelles 1, 242 Singapore 278, 284 – 6, 290 – 1, 302 Singer, Peter 305 Skarbek, David 54, 58, 63, 239, 296 Smith, Trevor 53 Sotomayor, Sonia 234 State v. Brimage 51, 253, 258 State v. Vasquez 249 Stevenson, Bryan 187 – 8 Strawson, Peter 148 Supreme Court of the United States 195 – 201, 205, 249, 254, 281, 286 talking circles 182 – 3 Thompson, Kenneth 246 Tocqueville, Alexis de 50, 78 torture 36, 123, 265 – 6, 296, 302 – 4, 316, 321

Index  335 transgender 58 – 9 Trump, Donald 250 Turner, Brock 1 Utah v. Strieff 234 utilitarianism 4, 69 – 70, 77, 79, 81, 254

Wadhwa, Anita 182 war on drugs 64, 210 – 3, 218, 226, 232, 239 – 40 Whitman, James 73 – 5, 78, 80, 82 – 3

Victims of Crime Act of 1984 55

Yates v. US 77 Yew, Lee Kuan 285 – 6 Young, Iris Marion 263

Wacquant, Loïc 180, 314

Zibilich, Franz 1