¹ One outstanding exception is Hart, H. L. A., Punishment and Responsibility (Oxford, 1968), pp. 17–21Google Scholar. This criticizes the ‘rationale of excuses’ in Chapter 13 of Bentham, Jeremy, An Introduction to the Principles of Morals and Legislation, ed. Burns, J. H. and Hart, H. L. A. (Oxford, 1996), pp. 160–2Google Scholar. Numbers in the text refer to pages in this edition.

² The passages in Bentham's Introduction that I focus on have parallels in Bentham's The Rationale of Punishment. Much of Rationale was written in the mid-1770s, although parts of it were clearly written decades later. It was not published in English until 1830. See Bentham, Jeremy, The Rationale of Punishment, ed. McHugh, James (Amherst, NY, 2009), p. 13Google Scholar. Rationale seems to contain earlier versions of some chapters in Introduction. Book I, Chapter 6 of Rationale closely parallels Chapter 14 of Introduction; Book I, Chapter 4 of Rationale parallels Chapter 13 of Introduction. See n. 24 below, where I argue that one passage in Introduction seems to be assuming an argument only made explicitly in Rationale. I cite Rationale by book and chapter, as well as by pages in The Works of Jeremy Bentham, ed. Bowring, John, vol. 1 (Edinburgh, 1843)Google Scholar.

³ Ewing, A. C., The Morality of Punishment (London, 1929), pp. 52–4Google Scholar; H. L. A. Hart in Bentham, Introduction, p. cv. See also Stephen, James Fitzjames, Liberty, Equality, Fraternity 2nd edn., ed. White, R. J. (Cambridge, 1967), pp. 152–4Google Scholar; Halevy, Elie, The Growth of Philosophic Radicalism, trans. Morris, Mary (New York, 1928), pp. 69–70Google Scholar. Maitland, F. W., ‘The Relation of Punishment to Temptation’, Mind 5 (1880), pp. 259–64CrossRefGoogle Scholar, largely agrees with Bentham. Draper, Anthony, ‘Punishment, Proportionality, and the Economic Analysis of Crime’, Journal of Bentham Studies 11 (2009), pp. 1–32Google Scholar, is the only previous study that tries to situate Bentham's arguments in Chapter 14 of Introduction in his larger theory.

⁴ Moore, Michael, Placing Blame (Oxford, 1997), p. 29Google Scholar.

⁵ There Bentham speaks of ‘propinquity’ and ‘remoteness’. ‘Proximity’ occurs at Introduction, pp. 169–70.

⁶ This important shift in meaning is clear in his explanation of the meaning of ‘profit’, but the same point obviously applies to ‘value’. See Introduction, pp. 166, note c (‘expectation of profit’), 167 (‘apparent profit’).

⁷ Bentham seems to recognize this point at Introduction, p. 162, paragraph 11.

⁸ The fundamental claims of Bentham's psychological hedonism are presented at Introduction, pp. 96–100. In this material, Bentham does not mention the factor of ‘proximity’. See Introduction, pp. 169–70. Bentham's later ‘A Table of the Springs of Action’, Deontology, ed. Goldworth, Amnon (Oxford, 1983), pp. 79–115Google Scholar, largely follows Introduction, although it seems to abandon psychological hedonism at p. 100, and tends to neglect the roles of certainty and proximity.

⁹ Introduction, pp. 166, note c, 167. Bentham does not identify these authors. It is clear that William Eden (1745–1814) is his main target. See Eden, William, Principles of Penal Law (London, 1771), pp. 7–9Google Scholar. The reference to ‘authors’ is thus puzzling. Eden's book went through four editions by 1775, but all were published anonymously. However, Bentham knew that Eden was the author: The Correspondence of Jeremy Bentham, ed. Sprigge, Timothy, vol. 2 (London, 1968), pp. 100, 114Google Scholar. Eden was a prominent politician in the late 1770s, when Bentham was writing Introduction. Eden's eminence may have made Bentham reluctant to criticize him by name. Two other authors are mentioned in Radzinowicz's authoritative work in relation to Introduction, pp. 166–7: Blackstone and Paley. But they basically agree with Bentham. Radzinowicz, Leon, A History of English Criminal Law and its Administration from 1750 (London, 1948), p. 384Google Scholar. Cf. Draper, ‘Punishment’, pp. 21–31.

¹⁰ Bentham seems still to have accepted this argument later in his life. Jeremy Bentham, ‘Specimen of a Penal Code’, in Works, vol. 1, pp. 164–8, at 166–7.

¹¹ Bentham does not explicitly mention there subjective proximity, but it is included in the Best Version of Rule 1, and is explicitly mentioned in Rule 8 (Introduction, p. 170).

¹² The example of theft of food by a starving person, and the claim that leniency is appropriate, are common. Thomas Aquinas, Summa Theologica (1274), II-II, 66, 7; Hobbes, Thomas, Leviathan, ed. Macpherson, C. B. (Harmondsworth, 1968), ch. 27Google Scholar, marginal heading, ‘Total Excuses’, p. 346; Blackstone, William, Commentaries on the Laws of England (Chicago, 1979), vol. 4, p. 15Google Scholar; Maitland, ‘Relation’, p. 262; Ewing, Morality, p. 52. Bentham himself discusses such an example. Introduction, p. 140. Cf. the slightly different example in Rationale, I 6, p. 400.

¹³ Hart uses this term, and it is now in general use: Hart, Punishment, pp. 13–14. Hart confined the domain of legal excuses to various kinds of psychological limitations of offenders, and this, too, is now generally accepted. Punishment, p. 14.

¹⁴ He also mentions four types of case where a type of act should be criminalized, but ‘occasional circumstances’ render it unprofitable to punish one or more offenders. One example he gives is where there are so many offenders that punishment of all of them will not produce the most happiness. Introduction, pp. 163–4. The issue discussed in sect. VI is structural, not occasional.

¹⁵ Cf. Hart, Punishment, p. 16; Fletcher, George, ‘The Individualization of Excusing Conditions’, Southern California Law Review 47 (1974), pp. 1269–88, at 1274Google Scholar (with regard to necessity). Bentham may have been thinking along the same lines. Introduction, p. 162, paragraph 10, seems to assert that some cases of mistake about the existence of circumstances that would provide legal justification for acting should be granted excuses.

¹⁶ Bentham notes that the profit of an offence is not always ‘proportioned to the mischief’ (Introduction, pp. 168–9, note k). But he does not see that this point, which he discusses in relation to Rule 2, is relevant to the argument he makes about deterring people subject to strong temptation. In some cases, the profit of stealing bread is so great that instances of it should not be an offence at all.

¹⁷ Given Bentham's conceptions of when an act should be treated as legally justified and when it should be treated as legally excused, it would be possible for one act to be such that it should be both legally justified and excused. This is because someone who is undeterrable could perform an act which produces the most happiness. When these conditions hold it would be better to say that it should only be treated as legally justified. In the rest of this paragraph I am discussing acts which should not be treated as justified, but should be excused.

¹⁸ Bentham takes Rules 7, 8 and 9 to be closely related to Rule 1. Rules 7 and 8 are entailed by the Best Version of Rule 1. Rule 9 brings in a distinct consideration, namely, the likelihood that someone who commits one offence had committed others, and was not punished for them. Introduction, p. 170.

¹⁹ Many of the other rules can be seen as implications of these two rules. Rules 3 and 4 can each be seen as an implication of Rule 2. Rule 12 can be seen as an implication of Rule 5. Rules 6, 10 and 11 can be seen as implications of the conjunction of Rules 2 and 5. (Rule 13 is second-order, requiring simplicity in the set of legislated rules.) Introduction, pp. 168–71.

²⁰ When we understand the normative force of Rules 2 and 5, Rule 1 comes to seem superfluous. Rule 1 focuses on a single causal channel of crime reduction, deterrence via a threat antecedent to an offence. However, punishment can reduce crime in other ways, as Bentham knew. For example, punishment can prevent crimes by disabling a criminal after an offence (Introduction, pp. 181–2). All of the channels of crime reduction should be governed by the utilitarian cost-benefit analysis, which is roughly articulated by Rules 2 and 5. This normative analysis can be stated at a level of abstraction in which deterrence antecedent to an offence has no special role.

²¹ Bentham recurrently contrasts ‘partial’ or ‘confined’ and ‘extensive’ or ‘enlarged’ benevolence, the latter being the more likely to be objectively correct. Introduction, pp. 117–18, 128, 135.

²² This discussion in Chapter 14 lays out the utilitarian argument against inefficacious punishments more explicitly than the well-known passage in Chapter 13.

²³ When punishment actually imposed diminishes an offender's tendency to break the law Bentham speaks of its ‘reforming’ effect (Introduction, p. 180–1). When it diminishes the tendency of other agents to break the law he speaks of its serving as an example, or its exemplarity (Introduction, pp. 178–9). Bentham also recognizes that some punishments simply ‘disable’ an offender from committing crimes (Introduction, pp. 181–2). Cf. n. 20 above.

²⁴ Bentham seems to deny this when he says that such punishment is cruelty ‘to the offender himself, by punishing him to no purpose’. He does not spell out his reasoning, but it is presumably given at Rationale, I 6, p. 399. There he says, ‘If … a man, having reaped the profit of the crime, and undergone the punishment, finds the former more than equivalent to the latter, he will go on offending’. Bentham is presumably picturing an eligible offender with a standing desire to commit a given offence, whose strength does not change after punishment. Such an offender is also being pictured as having subjective probability and proximity values that do not change after punishment. Bentham is arguing that if such a person believes, after undergoing the punishment for it, that he is better off than he was before committing it, then he will go on committing it. In that case, punishing him (at the same severity level) is wasted pain, at least with regard to preventing his offending. However, there are other types of offender: punishing them after failing to deter them can deter them in future. For example, after punishment an offender may change his belief about the probability of being punished. If so, the expected net value for him after t2 of offending again may be less than the expected net value for him of obeying the law, even if the expected net value for him at t1 of committing the offence was greater than the expected net value for him of obeying the law.

²⁵ Speaking of an error made by Anglo-Saxon laws he writes that it is also made by any legal system ‘wheresoever the punishment is fixed while the profit of delinquency is indefinite: or, to speak more precisely, where the punishment is limited to such a mark, that the profit may reach beyond it’ (Introduction, p. 167). This passage follows his statement of Rule 1.

²⁶ For a clear discussion, see National Research Council, The Growth of Incarceration in the United States, ed. Travis, J., Western, B. and Redburn, S. (Washington, 2014), pp. 138–40Google Scholar.

²⁷ Hart Punishment, pp. 14–17, speaks of ‘mitigation’, which is now the more common term. Bentham speaks of ‘extenuation’ at Introduction, p. 167.

²⁸ Introduction, pp. 69–70, 169, paragraph 15. Rationale contains a brief chapter: VI 1, pp. 516–17. Later in his career Bentham developed his thinking on how a utilitarian legal system should be codified, and how judicial officials would be legally authorized to apply these codes to cases, including, presumably, to sentencing. For two opposing interpretations of Bentham's position on the general question of judicial decision-making and discretion, see Postema, Gerald, ‘Bentham and Dworkin on Positivism and Adjudication’, Social Theory and Practice 5 (1980), pp. 347–76, at 350–8CrossRefGoogle Scholar, and Ferraro, Francesco, ‘Adjudication and Expectations: Bentham on the Role of Judges’, Utilitas 25 (2013), pp. 140–60CrossRefGoogle Scholar. But Bentham's views in 1780 on these matters were undeveloped. As Hart noted, Introduction contains no discussion of constitutional law (Hart, Introduction, p. cx; cf. p. 281, n. a). Bentham, ‘Specimen’, is much more detailed on sentencing than Introduction, but still sketchy on the limits of judicial discretion. Furthermore, English sentencing law c. 1780 did not provide Bentham with a usable model. It was extremely rigid in theory, at least with regard to felonies. Judges often had no discretion, and were required to sentence convicted criminals to death, even for property crimes. Various mitigating techniques existed, some of dubious legality. Langbein, John, The Origins of Adversary Criminal Trial (Oxford, 2003), pp. 57–61, 324–5, 334–6Google Scholar.

²⁹ Draper, ‘Punishment’, pp. 27–31, discusses this chapter, but largely rests his interpretation on an unpublished manuscript.

³⁰ For tutelary motives, see Introduction, pp. 134–5, 145–6.

³¹ This disregards the goal of general deterrence, but it is not clear that it would favour punishing the Nearly Starving Man more severely than the rich man.

³² Cf. Introduction, p. 142, which occurs after Rule 3 and its illustration.

³³ I thank three anonymous referees, Dale Miller, Robert Howell, Matt Lockard, Luke Robinson, Charles Curran, Alastair Norcross and, especially, Justin Fisher for helpful comments.