Is strict criminal liability in the grading of offences consistent with retributive desert?

Oxford Journal of Legal Studies

Volumn 32, Issue 3, 2012, Pages 445-466

  Simons, Kenneth W ^a  

^a BOSTON UNIVERSITY   (United States)

Author keywords

Culpability; Mens rea; Proportionality; Retribution; Strict liability

Indexed keywords

EID: 84865849717     PISSN: 01436503     EISSN: 14643820     Source Type: Journal    
DOI: 10.1093/ojls/gqs012     Document Type: Article

References (69)

1. I use the terms 'culpability' and 'mens rea' broadly to refer to any fault requirement, including negligence.
2. The example is a famous illustration from R v Serne 16 Cox Crim Cas 311, 313 (Central Criminal Court) (James Fitzjames Stephen J).1887
3. The article does presuppose a retributivist rationale for punishment. How the analysis would change under a mixture of retributivist and consequentialist rationales, I do not pursue here.
4. This analysis is drawn from my own earlier account, Kenneth W Simons, 'When is Strict Criminal Liability Unjust?' (1997) 87 J Crim L and Criminol 1075, 1121; and from the account of AP Simester, 'Is Strict Liability Always Wrong?' in AP Simester (ed), Appraising Strict Liability (OUP 2005) 44. The term 'strict liability in grading' is mine. The term 'constructive' liability is often used to capture a similar concept. See RA Duff, Answering for Crime: Responsibility and Liability in the Criminal Law (Hart; AP Simester and GR Sullivan, Criminal Law: Theory and Doctrine (3rd edn, Hart 2007) 166, 186-91, 372-79 (discussing 'constructive liability' generally and 'constructive manslaughter' committed by an unlawful act).2007 235, 253-55
5. In Simester's words, 'an offence imposes substantive strict liability when it contemplates the conviction of persons who are blameless for committing that particular offence'. See Simester, 'Is Strict Liability Always Wrong?' (n 4) 23. See also Simons, 'When is Strict Criminal Liability Unjust?' (n 4) 1085-93. As Duff points out (Duff, Answering for Crime (n 4)), there are four possible permutations of formal and substantive strict liability: (i) formally non-strict and substantively non-strict; (ii) formally strict and substantively strict; (iii) formally non-strict and substantively strict or (iv) formally strict and substantively non-strict. I agree with Duff that (i) is unproblematic, and that (ii) and (iii) are clearly unacceptable to a retributivist; thus, the proper focus of analysis is whether and when (iv) is acceptable.233-35
6. Douglas Husak, 'Strict Liability, Justice, and Proportionality' in Simester (ed), Appraising Strict Liability (n 4) 94-98 (distinguishing two problems of proportionality, proportionality in censuring and in hard treatment); Andrew Ashworth, 'A Change of Normative Position: Determining the Contours of Culpability in Criminal Law' (2008) 11 New Crim L Rev 232, 252 (noting that proportionality principles must address both fair labelling and the length of a sentence). Simester perhaps overstates when he suggests that the main proportionality objection is the undeserved censuring and labelling of the offender, not the undeservingly harsh treatment (Simester, 'Is Strict Liability Always Wrong?' (n 4) 33-35). Rather, which objection is more serious depends on the particular criminal label and the particular mode and degree of punishment. To be unjustifiably labelled a murderer or rapist is indeed highly stigmatic, and much more stigmatic than to be unjustifiably branded a manslaughterer or a perpetrator of sexual assault. However, if the severity of the sentence is seriously disproportionate, that type of disproportionality will be more important to most offenders. Few offenders would prefer a 15-year term in prison for manslaughter to a 10-year term for murder. Similarly, reclassifying a strict liability in grading element (for example, a homicide offence of 'intentional drug distribution causing death') as a sentencing element ('intentional drug distribution, with a harsh sentence authorized when a death results') avoids mislabelling but hardly solves the proportionality problem. See Husak, 'Strict Liability.. ' (n 6) 101-02.
7. The most prominent advocates of the argument are John Gardner and Jeremy Horder. John Gardner, 'Rationality and the Rule of Law in Offences against the Person' (1994) 53 CLJ 502
8. Jeremy Horder, 'A Critique of the Correspondence Principle' [1995] Crim L Rev 759.
9. Ashworth, 'A Change of Normative Position' (n 6). I have invoked the argument myself, without sufficient explanation. Simons, 'When is Strict Criminal Liability Unjust?' (n 4) 1124.
10. To my knowledge, English law is relevantly similar: although the amount of property stolen or drugs possessed is a sentencing factor, not a factor differentiating the degree of the crime, no mens rea is required as to that sentencing factor
11. George P Fletcher, Rethinking Criminal Law (Little, Brown and Company 1978 723-30
12. Husak, 'Strict Liability, Justice, and Proportionality' (6) 98-99
13. Duff, Answering for Crime (4) 256
14. Stephen P Garvey, 'When Should a Mistake of Fact Excuse?' (2009) 42 Tex Tech L Rev 359, 370-71.
15. Simester, 'Is Strict Liability Always Wrong?' (4) 45.
16. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1104
17. Ashworth, 'A Change of Normative Position' (6) 245-47.
18. Defendants are entitled, for example, not to appeal a sentence, even a sentence of death.
19. Ashworth briefly concedes that some 'change of normative position' cases properly fall within the categories of gross negligence manslaughter or reckless manslaughter, but he declines to discuss the point further (Ashworth, 'A Change of Normative Position' (n 6) 234, n 12). In my view, whether and when these cases can be so categorized is a critical part of understanding when 'change of normative position' is a defensible approach.
20. The appropriate offence category also depends on other factors, including fair notice and fair labelling. This article focuses on the appropriate severity of the sentence, not the appropriate offence category.
21. See Model Penal Code s2.02(2)©. English law is in accord: 'Not every case of foresight amounts to recklessness. In order for a defendant to be reckless, the risk that she chooses to run must also be an unreasonable one.' Simester and Sullivan (n 4) 134.
22. To be sure, the threshold itself could be higher or lower: it could demand a gross or instead merely an ordinary deviation from the standard of reasonable conduct; and it could be understood more objectively or more subjectively, taking into account more of the defendant's personal attributes and circumstances. A similarly impoverished understanding of unjustifiability is typical of contemporary accounts of criminal negligence: the trier of fact looks only at whether the actor was unjustifiable in lacking awareness of the risk, but not at how unjustifiable or culpable he was for being unaware. 18 A counterexample is the category of extreme indifference murder, many instances of which are also instances of reckless homicide in which the actor's reason for running the risk is especially unjustifiable. I also concede that when an offence category permits significant sentencing discretion, degree of unjustifiability sometimes affects the sentence.
23. Guyora Binder, 'The Culpability of Felony Murder' (2008) 83 Notre Dame L Rev 96
24. Providing this type of rationale for a qualified felony-murder rule); Larry Alexander and Kimberley Kessler Ferzan, Crime and Culpability: A Theory of Criminal Law (CUP) ch 2 (endorsing a unified formula for culpability in which the reason for taking the risk is a critical factor).2009
25. For defences of the principle, see Barry Mitchell, 'In Defence of a Principle of Correspondence' [1999] Crim LR 195; Andrew Ashworth, Principles of Criminal Law (6th edn, OUP 2009) 76-78
26. For criticism, see Horder, 'A Critique of the Correspondence Principle' (n 7); Jeremy Horder, 'Questioning the Corresponding Principle-A Reply' [1999] Crim LR 206.
27. I am employing the Model Penal Code definition of negligence, s2.02(2)(d), but many other definitions of negligence are consistent with this analysis.
28. Is the holistic culpability principle just another name for, or just a subcategory of, the highly unjustifiable risk principle? Not quite. The principles often overlap in practice. But sometimes culpability is holistic yet the reason is not because the risk is highly unjustifiable. For example, some would support a recidivist premium even when the separate crimes that the actor committed are only modestly culpable. More generally, X might only barely exceed the threshold of unjustifiability, but still, causing Y in the course of doing X displays holistic culpability.
29. Suppose an impatient driver runs over a drunk lying asleep in the middle of a narrow mountain road; or a panicky swimmer who unreasonably fears he is about to drown grabs the only available life preserver from a child.
30. For example, which is worse or more culpable, treason or murder? Raping one person or wounding 50? Stealing $1 million from a bank or forcibly robbing someone of his wallet? More generally, which is worse, causing harm or evil H with culpability C, or causing greater harm or evil H+ with lesser culpability C The correspondence principle is not a complete theory of culpability because it cannot resolve this comparability problem: We cannot simply rank all harms and wrongs in one scale of significance, then rank all mens rea or culpability criteria in a second scale, then multiply the values to arrive at a composite measure of the actor's just deserts. Kenneth W Simons, 'Understanding the Topography of Moral and Criminal Law Norms' in RA Duff and Stuart P Green (eds), Philosophical Foundations of Criminal Law (OUP 2011) 228, 241.
31. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1091-92
32. Simons, 'Topography' (n 24) 231. An extensive philosophical literature discusses comparability, commensurability, rough comparisons, and values that are 'on a par'. For an introduction, see N Hsieh, 'Incommensurable Values' (The Stanford Encyclopedia of Philosophy (Fall 2008 Edition)) accessed 26 May 2012.http://plato.stanford.edu/archives/fall2008/entries/value-incommensurable/
33. These three principles also apply outside the strict liability in grading context, as we have seen; they explain why committing crime X with a lesser mens rea as to Y often deserves to be punished as harshly as if the actor had brought about Y with a higher mens rea (but not as a byproduct of committing X). One other principle helps explain why some instances of formal strict liability amount to genuine fault: the principle that some strict liability criteria are really negligence criteria in rule-like form. An example is a strict liability rule that one may not sell alcohol to a person unless she shows two forms of identification indicating that she is of legal age. See Simons, 'When is Strict Criminal Liability Unjust?' (n 4). This principle is unlikely to apply, however, to the types of cases discussed in this article.1125-31
34. Simester, 'Is Strict Liability Always Wrong?' (n 4) 44-45. See also Simester and Sullivan, Criminal Law (n 4). I suggested a very similar approach in an earlier article (Simons, 'When is Strict Criminal Liability Unjust?' (n 4) 1125). Simester characterizes intrinsic risks as risks that form part of the reason why, even ex ante, the actor's conduct was wrong (Simester, 'Is Strict Liability Always Wrong?' (n 4) 45). He does not make clear, however, whether intrinsic risks include risks that explain how wrong the conduct is, even if they are not relevant to making the conduct wrong. For purposes of analysing strict liability in grading, I believe that we must identify a conception of relevant risks that also explains the extent of the wrong or of the actor's culpability, not simply whether the act was wrong or culpable vel non. See also Ashworth, ('A Change of Normative Position' (n 6) 252-53), who advocates a restriction of constructive liability to cases in which Y is within the scope of the foreseeable risks posed by X. It is not clear whether, by 'foreseeable risks', Ashworth means to include a broader range of risks than Simester's 'intrinsic' risks. See discussion of 'minimally foreseeable' risks in the text below.197-98
35. Simester, 'Is Strict Liability Always Wrong?' (4) 44.
36. On this account, the state should still have to prove that the risk of Y was an intrinsic rather than extrinsic risk of X, or else the court must be prepared to take judicial notice of this. An analogous issue arises in the American felony-murder doctrine: when a court requires that an eligible felony be 'inherently dangerous', does this require proof that in the particular case, the felon's acts posed a sufficient risk of death, or instead proof that the felony 'in the abstract' tends to pose a sufficient risk of death?
37. In my prior analysis of this issue, I made a similar claim, but less precisely. See Simons, 'When is Strict Criminal Liability Unjust?' (n 4) 1123-25 (arguing that if the foreseeability of the risk is just below the threshold that would be required for negligence liability in a case where the underlying activity was innocent, this level or risk should nevertheless suffice for negligence liability if the underlying activity was a serious felony).
38. The determination that the risk is more than trivial and is made more probable by the defendant's act should be from the perspective of a reasonable person. (Simester's 'Jane' would flunk part (a) of this test because she poses such a tiny risk of property damage.)
39. Relevant here is Judge Henry Friendly's argument, albeit in a tort case, against a requirement that every category of damages be foreseeable. So long as the defendant created a foreseeable risk of certain types of harm, he argued, 'the existence of a less likely [and unforeseeable] additional risk.. should inculpate him further rather than limit his liability'. Petition of Kinsman Transit Co, 338 F 2d 708, 725 (2d Cir 1964
40. Ashworth, 'A Change of Normative Position' (6) 250-53
41. How does this analysis differ from proximate cause analysis? There is some overlap, insofar as foreseeability is an element of both. However, other aspects of proximate cause analysis, especially requirements of physical or temporal proximity, are quite distinct. For example, suppose that D holds up V1 and V2 in an armed robbery. V1 immediately dies of a heart attack. V2 dies 20 years later of a heart attack, caused by V2's memory of the robbery together with other contemporary stressful events. D is properly considered culpable for the death of both, given that their deaths are at least minimally foreseeable. But proximate cause principles might preclude criminal liability for the death of V2.
42. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1110.
43. RA Duff, Criminal Attempts (Clarendon 1996); Michael S Moore, Placing Blame: A General Theory of the Criminal Law 1997
44. Peter Westen defends a non-desert justification for punishing more harshly when harm occurs (Peter Westen, 'Why Criminal Harms Matter: Plato's Abiding Insight in the Laws' 1 Crim L and Phil 307(2007)
45. In Westen's view, although the occurrence of the harm does not increase the actor's desert, its non-occurrence provides the state with a justifiable reason to be merciful. Of course, many retributivists deny the relevance of moral luck. See eg Andrew Ashworth, 'Taking the Consequences' in Stephen Shute, John Gardner and Jeremy Horder (eds), Action and Value in Criminal Law (OUP 1993); Sanford H Kadish, 'Foreward: The Criminal Law and the Luck of the Draw' (1994) 84 J Crim L and Criminol 679
46. Alexander and Ferzan (n 19) ch 5. Moral luck principles seem equally plausible (or implausible) whether Y is a circumstance element or a result element. See Simons, 'When is Strict Criminal Liability Unjust?' (n 4). However, what counts as the 'merely fortuitous occurrence of Y', triggering the moral luck principle, is a matter of some controversy, which I do not explore here. cf n 36.1115-16
47. One additional reason, not discussed in the text, is that not all strict liability in grading cases involve 'last act' attempts or endangerments; and it is doubtful that moral luck principles should apply outside of such 'last acts'. However, quite a few 'strict liability in grading' cases do involve 'last acts', so I do not place primary reliance on this argument. Let me explain. The issue of whether we should add a moral luck supplement arises in pure form only in such 'last act' cases, not in cases where the actor has not yet taken the last step towards his intended goal or has not yet irrevocably unleashed a risk of harm (Simons, 'When is Strict Criminal Liability Unjust?' (n 4) 1112-13, 1115-16). It arises in pure form, for example, as a reason to increase the punishment of M2 relative to M1 (see text above), but not if the actor intends to fire the gun but has not yet done so. In the latter case, it is still up to the actor whether to fire the gun, and the criminal law might have good reason to impose a lesser punishment on an actor who has not yet finally irrevocably unleashed the risk than on an actor who has. Although many 'strict liability in grading' cases involve a last act or the irrevocable unleashing of risk, not all do. Suppose D believes that V is 13, when she is actually 9 years old, and he is arrested just prior to having intercourse with her; and suppose negligence is required as to her being under the age of 16 years, but no mens rea is required as to whether she is under the age of 10 years. See Model Penal Code ss213.4(4), 213.6(1). Or consider Andrew, from the introduction. Suppose he walks up to his victim and pulls back his fist menacingly, but is restrained before he can strike the victim.
48. This might roughly translate into a 100% increase in the length of imprisonment, though I concede that the equivalence is at best rough. For discussion of the difficulties of translating just desert principles into actual conditions of punishment, see Adam J Kolber, 'The Subjective Experience of Punishment'; Kenneth W Simons, 'Retributivists Need Not and Should Not Endorse the Subjectivist Account of Punishment' (2009) 109 Colum L Rev Sidebar 1. The moral luck supplement could also be sensitive to the nature of the result; perhaps a 150% supplement is warranted for death, but only a 10% supplement for bodily injury. And perhaps the supplement should take the form of an absolute rather than a relative (eg percentage) increase in the severity of the punishment for the original attempt or endangerment. These variations do not affect the basic argument in the text.(2009) 109 Colum L Rev 182
49. Similarly, the 'highly unjustifiable risk' and 'roughly comparable culpability' arguments are independent, in their weight and significance, from the moral luck supplement.
50. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1119.
51. Yet another relevant factor is whether, and to what degree, the authorized punishment for X already takes into account (and should take into account) the extent to which X creates a risk of Y. Armed robbery is punished as harshly as it is in part because it creates a risk of death. For reasons of space, I have only briefly discussed this factor.
52. In this stylized analysis, the numbers refer to abstract quantities of harsh treatment (or of censure). It is of course controversial how these should be translated into real-world punishment conditions, such as months or years in prison, fines, probation or parole, or other forms of loss of liberty, wealth, or other opportunities.
53. Notice that, to explain these punishment levels, we no longer can rely on the corresponding punishment levels of Dawn, since these are 0. So it might seem that the moral luck supplement does, in this example, fully explain the punishment levels. But that is not the case. In order to evaluate the culpability of each Don, and to explain why each is culpable while each Dawn is not, we still need to apply the holistic culpability or highly unjustifiable risk principle. However, applying those principles is more complex than before, for they can no longer be cashed out as a simple supplemental formula (eg as 50% more than analogous Dawn would deserve).
54. In the real world, where formal strict liability very often does not instantiate substantive fault, the coherence and nomenclature issues remain. The question then is whether a (non-moral) luck principle can justify increasing the penalty when the act fortuitously causes harm or fortuitously satisfies a circumstance element. For example, suppose a jurisdiction recognizes a strict liability crime of speeding (eg it is no defence that the speedometer malfunctioned). And suppose it also recognizes a more serious crime of speeding causing death. Does the 'luck' principle justify a higher punishment for the more serious crime? If a persuasive argument for the moral luck principle relies only on the actor's causal responsibility for the effects he brings about in the world, without regard to his culpability, the answer could be yes.
55. Ashworth, 'A Change of Normative Position' (6) 255
56. However, Ashworth seems to use the phrase to refer to a proximate cause limit.
57. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1111-12:
58. The 'highly unjustifiable risk' and 'roughly comparable culpability' arguments have less purchase in this example, at least if the actor believes the victim is just under the legal age and if statutory rape of a person just under the legal age is a minor crime.
59. To be sure, if one both endorses moral luck and believes that the proper moral luck punishment supplement is very large, then it could be justifiable to impose a very harsh punishment on one who is not even negligent about the victim being under the age of 12 years. But those beliefs would also justify extremely draconian punishments across the board whenever tiny risks posed by criminal conduct eventuate in harm. I assume that this result is not consistent with the best retributive theory.
60. Ashworth, 'A Change of Normative Position' (6) 254
61. My own prior analysis of this point too readily accepted strict liability in this scenario. Simons, 'When is Strict Criminal Liability Unjust?' (4) 1092.
62. Oregon Rev Stat ss161.605(3), 161.615(1), 161.615(3), 164.043(1)(b), 164.045(1)(b), 164.055(1)(a) (2009).
63. For a disturbing litany of cases exhibiting the serious dangers of over-punishment under the felony-murder doctrine
64. Guyora Binder, 'Making the Best of Felony Murder' (2011) 91 BU L Rev 405-07
65. According to Binder, however, most jurisdictions would not permit felony-murder liability in these cases; and he suggests that a more qualified felony-murder doctrine can be consistent with retributive and consequentialist principles.
66. Kenneth W Simons, 'Rethinking Mental States' (1992) 72 BU L Rev 463
67. Kenneth W Simons, 'Does Punishment for 'Culpable Indifference' Simply Punish for 'Bad Character'?' (2002) 6 Buff Crim L Rev 219.
68. Consider, for example, this question. What counts as 'awareness' of a risk, sufficient to satisfy a criminal recklessness requirement? The answer is complex. Perhaps readily available, latent knowledge about risks that the actor could easily access should satisfy an 'awareness' requirement; yet from one perspective, this seems to collapse the distinction between recklessness and negligence. Similarly, if at one time an actor knows a legally relevant fact but later forgets it, should we treat him as 'knowing' the fact at that later time? Perhaps we should, if he is clearly negligent in failing to remember what he once knew. For discussion, see Kenneth W Simons, 'Book Review: Retributivism Refined-Or Run Amok?' (2010) 77 U Chi L Rev 551, 568-79
69. Douglas Husak, 'Negligence, Belief, Blame and Criminal Liability: The Special Case of Forgetting' (2011) 5 Crim L Phil 199