I. Introduction

On August 16, 2015, Louisa Sewell, a 32‐year old woman living in Kidderminster, England, faced the magistrates’ criminal court and admitted to stealing a pack of chocolate bars. In mitigation of her guilty plea, Sewell’s defense lawyer argued that her client had been hungry for days and, with her welfare benefits suspended and unable to turn to friends or family for help, had decided to stay her hunger by shoplifting one of the cheapest food items in the store. The criminal court’s ruling dismissed the excuse and imposed a £328.75 fine: £73 as a criminal sanction, £150 in court administrative charges, £85 in prosecution costs, £20 as a victim surcharge, and 75 pence in compensation to the store. Upon sentencing, the chairman of the magistrates’ bench justified the decision by saying that “we do not readily accept you go into a shop to steal just for being hungry”.11 For a narrative of the case, see <https://theconversation.com/is‐it‐really‐criminal‐to‐steal‐food‐when‐youre‐destitute‐59190>.

Sewell’s case is not unique. Many contemporary liberal democracies that are expected to treat their citizens fairly allow for levels of inequality that push some people—most notably, those who find themselves subject to burdensome forms of social and economic disadvantage—to the brink of criminal acts. Some object that, when faced with Sewell‐like cases, magistrates are wrong not to admit hunger or other similar necessity defenses. Others worry that, whether hunger should count as an excuse or not, stealing chocolate once is not a blameworthy action or that responding to such petty theft represents a prudentially undue burden on the criminal justice system.22 For a review, see B. Ewing, “Recent work on punishment and criminogenic disadvantage,” Law and Philosophy, 37 (2018), 29–68.

A more radical objection is that punishing offenders like Sewell is wrong because people who commit crimes in a context where they are subject to serious injustice can sensibly argue that the state is no longer in a position to punish and condemn them. It is this latter objection that I will assess in this article. The objection holds that (at least some of) the crimes committed by socially disadvantaged offenders fall outside the scope of the state’s legitimate authority and that, consequently, they should not be punished. The thesis underlying this objection is that a state that fails to tackle serious injustice thereby forfeits its authority to punish those who are the victims of such injustice. Call this the authority thesis.

The authority thesis has many supporters, notably Antony Duff and Tommie Shelby.33 R. A. Duff, Punishment, Communication and Community (Oxford: Oxford University Press, 2001); R. A. Duff, “‘I might be guilty, but you can’t try me’: estoppel and other bars to trial,” Ohio State Journal of Criminal Law, 3 (2003), 245–59; R. A. Duff, Answering for Crime: Responsibility and Liability in the Criminal Law (Oxford: Hart, 2007), ch. 8; R. A. Duff, “Authority and responsibility in criminal law,” S. Besson and J. Tasioulas (eds), Philosophy of International Law (Oxford: Oxford University Press, 2010), pp. 589–604; T. Shelby, Dark Ghettos: Injustice, Dissent, and Reform (Cambridge, MA: Harvard University Press, 2016). Duff and Shelby disagree on its specific formulation, but they agree that a seriously unjust state lacks the authority to punish at least some of the offenses committed by disadvantaged offenders. The core contention of the authority thesis is that the state’s failure to ensure minimal social justice constitutes a flaw that the state needs to address and correct.44 My argument does not depend on a fully fledged conception of serious social injustice. All that matters here is that there is a level of social injustice that counts as serious under a plurality of plausible theories of social justice. It is precisely because it is the state’s job to prevent or eliminate serious social injustice that, when the state performs this job badly, it deserves criticism and has to suffer the consequences. One of the consequences is that the state loses its legitimate authority to punish the crimes committed by those who find themselves deprived as a result of the state’s poor justice record.55 I use the terms authority, legitimacy, and legitimate authority interchangeably, unless specified otherwise. I also use the expressions meeting minimal social justice and preventing/eliminating serious social injustice interchangeably.

The underlying presumption of the authority thesis is that the state, or a state‐like agency, is ultimately supposed to prevent serious social injustice and punish people for the crimes they commit. If that were not the case—that is, if it weren’t the state’s job to bring about social justice and if the state did not claim authority over penal matters—one could not explain why a seriously unjust state lacks penal authority. It would be odd to criticize an agent for asserting its authority if its authority depended on meeting a requirement the agent ought not to meet.

I shall argue, however, that for all of its apparent appeal, the authority thesis is ultimately untenable. My rejection of it relies on two moves. The first move argues that a state that has lost its authority to punish people who are socially deprived also lost its authority to punish those who are not socially deprived. If the state’s penal authority depends on its meeting certain social justice requirements, states that fail to satisfy these requirements lose their penal authority altogether. I will call this the non‐selectivity objection.

The second move is that, once one has accepted the non‐selectivity objection, the demand that the state should eliminate serious injustice can no longer be coherently sustained. For the social‐justice‐promoting demand to properly address the state, the latter has to be in a position to legitimately enforce justice‐oriented reforms. Since carrying out these reforms requires some authority for the state to punish those who would prevent them from doing so, a state without penal authority cannot enact justice‐oriented reforms. We cannot coherently demand that the state do something that it has no legitimate authority to do. I’ll call this the incoherence objection.

The article proceeds as follows. In Section II, I provide a characterization of the authority thesis, as formulated by Duff and Shelby. In Sections III and IV, I develop the non‐selectivity and the incoherence objections and explore some of the ways in which supporters of the authority thesis might reply. I conclude in Section V by briefly arguing that, although the authority thesis is incoherent in its own terms, formulating its underlying intuition in terms of a different account should inform the state’s treatment of socially disadvantaged offenders. I contend that punishing disadvantaged offenders is not wrong per se. Rather, the wrong consists in punishing some disadvantaged offenders through penal methods—for instance, excessive imprisonment or unreasonably high fines—whose imposition is foreseeably iatrogenic. This is to say that some forms of punishment will compound their disadvantage and thereby increase their risk of reoffending. The wrong, on this view, is that punishment contributes to increasing rather than reducing crime.

中文翻译：

---

社会不公，弱势罪犯和国家处罚机构*

一，引言

2015年8月16日，居住在英格兰基德明斯特的32岁妇女Louisa Sewell面对地方法院的刑事法庭，并承认偷了一包巧克力。为了减轻认罪，Sewell的辩护律师辩称，她的委托人已经饿了好几天，而且由于她的福利金被暂停并且无法求助于朋友或家人，她决定通过扒窃最便宜的食物之一来保持饥饿。商店中的物品。刑事法院的裁决驳回了这一借口，并处以328.75英镑的罚款：73英镑作为刑事制裁，150英镑法院管理费，85英镑诉讼费，20英镑受害者附加费以及75便士的商店补偿金。一经宣判，1有关案例的叙述，请参见<https://theconversation.com/is-它真的是犯罪分子，是您在目的地时偷窃食物-59190>。

塞维尔的情况并非唯一。期望公平对待其公民的许多当代自由民主国家允许一定程度的不平等，这些不平等将某些人（尤其是那些发现自己遭受沉重负担的社会和经济劣势的人）推向犯罪行为的边缘。有人反对在面对类似塞维尔的案件时，裁判官不承认饥饿或其他类似的必要辩护是错误的。其他人则担心，无论是否应将饥饿视为借口，一次偷巧克力都不是应受的谴责，或者对这种小偷小摸的反应是对刑事司法系统的审慎过度负担。22综述，请参阅B.尤因，“关于惩罚和致犯罪性的不利因素的最新工作”，《法律与哲学》，37（2018），29-68。

一个更激进的反对意见是，惩罚像Sewell这样的罪犯是错误的，因为在严重遭受不公正待遇的背景下犯罪的人可以明智地辩称，国家不再有能力惩罚和谴责他们。我将在本文中评估的是后一种反对意见。反对意见认为，社会处境不利者所犯的罪行（至少其中一些）不在国家合法权限的范围之内，因此不应予以惩处。该反对意见的依据是，一个国家无法解决严重的不公正现象，从而丧失了惩罚那些遭受不公正待遇的人的权力。称其为权威论文。

权威论文有很多支持者，特别是安东尼·达夫（Antony Duff）和汤米·谢尔比（Tommie Shelby）。33 RA Duff，《惩罚，沟通与社区》（牛津：牛津大学出版社，2001年）；RA Duff，“'我可能有罪，但你不能尝试我'：禁止反言和其他审判权，”俄亥俄州立刑法杂志， 2003年第3期，第245-59页；RA Duff，“对犯罪的回应：刑法中的责任和责任”（牛津：哈特，2007年），第1章。8; RA Duff，“刑法的权威和责任”，S。Besson和J. Tasioulas（编辑），国际法哲学（牛津：牛津大学出版社，2010年），第589-604页；T.谢尔比（T.Shelby），《黑暗的犹太人：黑暗，异议与改革》（马萨诸塞州剑桥市：哈佛大学出版社，2016年）。Duff和Shelby在其具体表述上存在分歧，但他们同意，一个严重不公正的国家没有权力惩罚至少一些处境不利的犯罪者所犯的罪行。权威论点的核心论点是，国家无法确保最低限度的社会公正构成了国家需要解决和纠正的缺陷。44我的论点不依赖于严重的社会不公的完整概念。在这里重要的是，在一系列合理的社会正义理论中，存在着一定程度的严重社会不公正现象。正是由于防止或消除严重的社会不公是国家的工作，当国家在这一工作中表现不佳时，应受到批评并必须承担后果。其后果之一是，国家失去了惩罚因国家司法记录不佳而被剥夺自己的人所犯下的罪行的合法权力。55除非另有说明，否则我会交替使用“权威”，“合法性”和“合法权威”一词。我还使用了满足最低限度的社会正义并交替预防/消除严重的社会不公的表达方式。

权威论断的基本假设是，国家或类似国家的机构最终应防止严重的社会不公正，并就所犯下的罪行对人们进行惩罚。如果不是这种情况，也就是说，如果不是由国家来实现社会公正的工作，或者如果国家不要求对刑事事务具有权威，那么谁也无法解释为什么一个严重不公正的国家缺乏刑事权力。如果代理人的权威取决于满足代理人不应该满足的要求，那么批评代理人断言其权威将是很奇怪的。

但是，我将争辩说，就其所有明显的吸引力而言，权威论点最终都是站不住脚的。我对此的拒绝取决于两个步骤。第一步是说，一个失去了惩罚被社会剥夺者的权力的国家也失去了惩罚未被社会剥夺者的权力。如果州的刑事权力取决于其是否满足某些社会正义要求，则不满足这些要求的州将完全失去其刑事权力。我将其称为非选择性异议。

第二步是，一旦人们接受了非选择性的反对意见，就不再连贯地维持国家消除严重不公正的要求。为了适当地解决国家对促进社会正义的需求，后者必须能够合法地执行以司法为导向的改革。由于进行这些改革需要国家有一定的权力来惩罚那些会阻止他们这样做的人，因此没有刑事权力的国家就无法实施以司法为导向的改革。我们不能一致地要求国家做它没有合法权限可做的事情。我称其为不连贯的异议。

文章进行如下。在第二部分中，我提供了Duff和Shelby提出的权威论断的特征。在第三节和第四节中，我提出了非选择性和不连贯性的反对意见，并探讨了权威论文支持者可以回答的一些方式。在第五节中，我简要地总结道，尽管权威论点本身并不一致，但以不同的说法来表达其基本直觉应该为国家对社会弱势犯的待遇提供依据。我认为，惩罚处境不利的罪犯本身并没有错。相反，错误在于惩罚一些通过刑事手段使处境不利的罪犯（例如，过多的监禁或不合理的高额罚款）可预见的是，这种施加是医源性的。也就是说，某些形式的惩罚将加重其不利条件，从而增加其再次犯罪的风险。这种观点的错误在于，惩罚有助于增加而不是减少犯罪。