A. LEGALLYPROTECTEDINTERESTS
Our criminal code must enumerate those interests that it seeks to pro- tect. Th is requirement raises a number of questions. Th e fi rst ques- tion is whether, as a normative matter, we should “unpack” criminal wrongs into these building blocks. Although in Chapter 7 we defended such unpacking on conceptual grounds, in the fi rst part of this sec- tion we defend this approach on normative grounds. In the second part of this section, we turn to the question of which interests should be
protected by the criminal law. Although we cannot defend a theory of criminalization here, we raise some of the critical questions concerning punishing risks of harm to self and others, risks of off ense, and risks of harmless wrongs. We also briefl y discuss the role of consent.
1. A Normative Defense of Unpacking Crimes
In the previous chapter, we argued that it is conceptually preferable to understand a crime as a volition to move one’s body in a way that the agent believes will create a risk (which is unjustifi able) to a legally protected interest. Th is approach is conceptually superior because it resolves a host of thorny problems ranging from how to treat continu- ous courses of conduct to how to approach the double jeopardy clause.
Still, one might object that something is missing from a criminal code that does not list wrongs – those specifi c act types that harm spe- cifi c legally protected interests. We speak of rape, murder, and rob- bery, not of unjustifi able risks. Is there not something missing from an account of the criminal law that does not mirror our ordinary under- standing of wrongful conduct?
In our view, our approach brings precision, clarity, and deeper understanding to the criminal law. One signifi cant problem with an act-type system is that too much hangs on the ability to place any spe- cifi c act of risk creation within any given act-type category. As we noted in the previous chapter, rape is a clear example of this. Long ago, Susan Estrich bemoaned the disparate treatment between “date rapes” and
“real rapes.” 1 Her claim was that they are all rapes.
We disagree. We do not doubt that date rapes are serious wrongs, and we will assume that the criminal justice system still underpun- ishes these (and other) serious wrongs. What we question is whether the eff ort to place the myriad of diff erent unconsented-to acts of sex- ual intercourse within the rubric of “rape” is a worthwhile eff ort. Why, for example, should we understand rape, at its core, to be simply about the use of another human being?2 Should not the criminal law take into account whether the victim found the rape pleasurable (as might
1 See Susan Estrich, Real Rape 3–4 (1987).
2 See J. Gardner and S. Shute, “Th e Wrongness of Rape,” in Oxford Essays on Jurisprudence (J. Horder, ed., 2000).
happen in a case of deception); or whether the victim was sleeping;
or whether the victim did not communicate consent but also did not resist; or whether the victim was brutally beaten during the act? Let us be clear – society can accept that all of these acts are criminal. But they present diff erent risks to diff erent legally protected interests. Which interests of those aff ected are more important? Which acts risk more harm? Th ese are important issues, and they are obscured by the eff ort to place all of these acts into one criminal act type. We can more readily identify and rank these interests when we look at each one individu- ally. To summarize, one reason to abandon labels of particular types of wrongdoing is because such labels tend to obscure rather than to clarify the underlying normative justifi cations for punishing these diff erent types of culpable riskings.
A second signifi cant problem with the act-type approach is that the current system will oft en declare an act criminal and hence punishable without any thought to the interest that is being protected. Although, theoretically, a statute may be subject to a constitutional challenge if it lacks a rational basis, the number of statutes that are currently ratio- nally but remotely linked to actual harms is staggering. Drug off enses, particularly possession off enses, are a classic example of this phenom- enon. Although we may be able to articulate the crime of drug posses- sion, one will be hard-pressed to articulate its underlying rationale.
What precisely are the harms that are risked by an actor possessing one ounce of marijuana?3 By contrast, our approach equates the crime with its underlying justifi cation by equating crime with culpable acts, and culpable acts with risks imposed for inadequate reasons. Th us, if one cannot articulate an interest excessively put at risk, one cannot articu- late a crime.
A third reason to reject act-type categorization is that it cre- ates false distinctions between types of off enses, potentially allowing similarly culpable actors to receive signifi cantly diff erent penalties.
Consider murder and manslaughter. Currently, an actor may be guilty of murder on the basis of several diff erent culpability types. He may act
3 See Douglas Husak, Overcriminalization: Th e Limits of the Criminal Law 166–167 (2008).
Husak calls this unidentifi ed harm, “harm X.” See also Douglas Husak, Legalize Th is! Th e Case for Decriminalizing Drugs (2002).
purposefully, knowingly, or recklessly if such recklessness manifests a depraved heart. On the other hand, an actor is guilty of only man- slaughter if he kills recklessly or negligently. Th e distinction, then, between murder and manslaughter is a fi ne one – a jury decision along a continuum. At some point, a homicide becomes so reckless that the jury thinks it warrants more punishment (murder) than does another, slightly less reckless, homicide (manslaughter). Th e bottom line is that there is no defi nitive line between murder and manslaughter – they are diff erent in degree but not in kind. Th us, even though it appears that we have clear conceptual categories of homicide – murder and manslaugh- ter – the reality is somewhat diff erent. Both protect the same interest, and the culpability that supposedly distinguishes them is an arbitrary line (and a diff erent one in diff erent jurisdictions with diff erent juries) on a culpability continuum.
Finally, we should discuss the concern that our approach cannot capture a certain type of wrongdoing, and that is the group of inten- tion-drenched wrongs. One cannot lie without the intention to lie. One cannot torture without the intention to cause pain. Th ese crimes are unraveled by our approach not only because we focus on risks to inter- ests and not act types, but because we deny the signifi cance of intention as its own separate culpable mental state.
We believe that nothing is lost by our approach, however. Th ese off enses can and should be analyzed by their discrete elements. Indeed, one of the paradigmatic examples of an intention-drenched crime is the crime of attempt, a crime we have no problem unraveling.4
Moreover, although these crimes entail intention, there is no rea- son we need to rely on these particular crimes as currently understood.
First, as we have argued, conceptually, intention (or purpose) is but a particular species of recklessness so there is no conceptual reason why we cannot speak of recklessness instead of intentions. Even if con- sciously imposing a high risk that another will be misled is not “lying,”
it still may be – or may not be, depending upon the actor’s reasons – a culpable, reckless act with respect to the interest in not being misled.
4 Th is is a case where even those theorists who believe that crimes can be understood without mental elements, like Paul Robinson, create an exception. See Paul H. Robinson, Structure and Function in Criminal Law 133 (1997).
Indeed, recognizing the relationship between recklessness and intention may resolve current confusions. For example, New York courts have struggled with the question of whether a depraved indif- ference murder conviction is appropriate where an actor shoots his vic- tim at point-blank range but the jury acquits on an intentional murder charge. Current New York cases maintain that depraved indiff erence is inappropriate because intentional killings cannot be depraved heart killings.5 But this conclusion is certainly incorrect. A person who shoots at point-blank range with no good reason has surely manifested extreme indiff erence to human life. Intentional killings are just one (typical) spe- cies of such indiff erence.
Normatively, there is no reason to restrict punishment to instances of risking with the intent to bring about the harm risked. If the actor consciously disregards an unjustifi able risk of harm, her action is cul- pable even if she does not want that harm to occur. Indeed, given an actor’s reasons for acting, her “reckless attempt” may be more culpable than another actor’s “intentional attempt.”
Or, consider complicity. When Iago taunts Othello, he may not wish for Desdemona to die. Instead, Iago may believe it suffi cient to mentally torture Othello. But when Othello kills Desdemona, Iago is to blame for this action. His conduct – lying about Desdemona’s infi delity – created an unjustifi able risk that Othello would kill his wife. We see no reason why Iago should escape liability because Desdemona’s death was not within the scope of his intention.6
Indeed, although ordinary language and lay intuitions may be use- ful, they cannot be the last word on draft ing criminal codes. Even if, as a matter of semantics, an actor cannot “attempt” a reckless homi- cide because one cannot intentionally commit an unintentional act, we may alternatively label the reckless actor’s conduct “endangerment”;7 but, labels aside, the only distinction then between the “attempt” and the “endangerment” is the actor’s reason for imposing the risk. We see
5 See, e.g., People v. Payne, 819 N.E.2d 634 (N.Y. 2004); People v. Suarez, 844 N.E.2d 721 (N.Y.
2005).
6 See Larry Alexander and Kimberly D. Kessler, “Mens Rea and Inchoate Crimes,” 87 J.
Crim. L. & Criminology 1138 (1997); Sanford H. Kadish, “Reckless Complicity,” 87 J. Crim.
L & Criminology 369 (1997).
7 See generally R. A. Duff , “Criminalizing Endangerment,” 65 La. L. Rev. 941, 960–961 (2005).
no reason not to focus on the interest risked and the actor’s culpability regarding that risk rather than on the actor’s intention.
2. Which Interests?
Th e starting point for draft ing our criminal code is determining which interests the law should protect against culpable risk impositions.
Outside of our brief discussion in Chapter 1, we have had nothing to say about which risks the law should protect us from. Some risks are relatively straightforward. Th e law should protect us from inordinate risks of being killed, maimed, or even punched in the face. Beyond these obvious “harms to others,” however, lie a host of thorny ques- tions about what interests the criminal law ought to protect. Although we do not have time to resolve these questions here, we do fl ag some of these issues.
a. Harm to Others and Beyond: Following Mill and Feinberg, one might catalog the various justifi cations for criminalization as harm to others, harm to self (paternalism), off ense to others, and legal moralism.8 We discuss criminalization questions within each of these categories.
i. Harm to Others: Although preventing harm to others is the clearest justifi cation for state interference, there are diffi cult issues even here. First, there are some potential harms that may be diffi cult to specify. Th at is, it is easier to criminalize the conduct than to articulate the precise harm. Of course, when one sees a crime like this, one immediately suspects that the crime stands on normatively shaky ground. But this is not always true.
Take public corruption crimes. Under section 1346 of the United States Code, it is a violation of the federal mail and wire fraud statutes to deprive another of your “honest services.”9 Th is honest services statute certainly should apply to a public offi cial who takes a bribe. But does this apply to criminalize the conduct of the Internal Revenue Service
8 See generally John Stuart Mill, On Liberty (1859); Joel Feinberg, Harm to Others (1984);
Joel Feinberg, Off ense to Others (1985); Joel Feinberg, Harm to Self (1986); Joel Feinberg, Harmless Wrongdoing (1988).
9 18 U.S.C. § 1346 (2000).
employee who improperly reads others’ tax returns?10 Must a state employee’s action violate state law to fall within the statute?11 Although the latter question raises federalism concerns, the underlying issues are the type of duty of “honest services” owed and to whom that duty is owed – two extremely diffi cult questions to answer.12 For our criminal code, it will no longer be suffi cient to criminalize fi rst and analyze later.
Rather, an understanding of the precise interests threatened will be nec- essary in order to criminalize the conduct in the fi rst instance.
A second diffi cult category of harms involves fear and other emo- tional injuries. Although the criminal law prohibits some actions that cause fear, it typically does not prohibit them because they cause fear but because that fear is incidental to some other harm. For example, having a gun pointed at you may cause fear, but it causes fear because it risks harm, and it is the risk of harm that the criminal law is actu- ally targeting. Even tort law, which is oft en far more expansive than the criminal law, is unwilling to recognize all sorts of emotional distress as compensable harms. Although some theorists distinguish between public and private wrongs, we need more than this labeling in order to justify, say, not criminalizing cruelly breaking someone’s heart while simultaneously criminalizing the stealing of fi ve dollars. Although there is room to argue that we are entitled to more liberty within self-defi ning spheres, the leeway to which any individual is entitled is relevant to the actor’s justifying reasons, not to whether the victim’s interest should be protected in the fi rst instance.
Finally, we do not doubt that there will be particularly contested cases – such as whether fetuses should be protected by the law. Here, we only caution once more that the question of whether an interest is protected is not the question of whether risking harm to that interest may be justifi ed.
10 United States v. Czubinski, 106 F.3d 1069 (1st Cir. 1997).
11 Cf. United States v. Sawyer, 239 F.3d 31, 42 (1st Cir. 2001) (no need to prove lobbyist’s giving of gratuities and gift s to state offi cial violated state law) with United States v. Brumley, 116 F.3d 728, 734 (5th Cir. 1997) (“the offi cial must act or fail to act contrary to the requirements of his job under state law”).
12 For a recent attempt, see Joshua A. Kobrin, Note, “Betraying Honest Services: Th eories of Trust and Betrayal Applied to the Mail Fraud Statute and § 1346,” 61 N.Y.U. Ann. Surv. Am.
L. 779 (2006). For an analysis of white collar off enses generally, see Stuart Green, Lying, Cheating, and Stealing: A Moral Th eory of White-Collar Crime (2006).
ii. Paternalism: A second question is whether our code should protect individuals from themselves.13 Indeed, one might wonder how one shows insuffi cient concern for others when the only person the actor seeks to harm is himself. We fi nd this objection to be signifi cant.
Moreover, even if we could say that someone can be culpable for risking harm to herself, there is the question of whether the law should intervene to prevent such harms. Many theorists draw a distinction between hard and soft paternalism and would permit the law to step in when failures of information or rationality prevent an agent from choosing according to her own theory of the good. However, it may be doubted whether the distinction between hard and soft paternalism can be drawn. Let us explain.
One tenet of liberalism is that normal adults should not be pre- vented from acting as they choose merely because would-be preventers believe those choices will be harmful to the choosers. Yet liberals make an exception for cases labeled “soft ” paternalism. Th us, if Joy does not realize that the bridge of San Luis Rey is dangerously weak, and if she has no desire to end her life, it is supposedly permissible to restrain her from crossing the bridge. Although Joy believes she wishes to cross the bridge at the time her choice is interfered with, her false belief about the bridge’s condition and her desire to live – a desire that trumps any desire she might have to cross this bridge at this time – justify the inter- ference with her choice. Without undergoing any change in her stable values and commitments, Joy can be expected to welcome others’ inter- ference with her choice to cross the bridge once she is made aware of the bridge’s condition.
As we said, the kind of paternalism exemplifi ed in preventing Joy from crossing the bridge of San Luis Rey is termed “soft paternalism.”
Th e soft paternalist argues that choices such as Joy’s can be interfered with on grounds of her good only when she lacks information that, given her values, she would deem material to those choices, or when she is too young or too defective in rationality to process that informa- tion correctly. Th e soft paternalist contrasts her position with that of the hard paternalist, who believes in overriding choices whenever they
13 Th is section draws from Larry Alexander, “Scalar Properties, Binary Judgments,” 25 J.
Applied Phil. 85 (2008).
are contrary to the chooser’s good and irrespective of why those choices were made. Most liberals endorse the position of the soft paternalist but for obvious reasons reject that of the hard paternalist.
Th e relationship between hard and soft paternalism is, however, a matter of degree rather than a diff erence in kind. People are more or less rational and more or less informed. Th ey are on a smooth continuum in these respects, and there is no obvious threshold point marking a mor- ally relevant diff erence.14
Consider Joy2, who wants to cross the bridge. Th e defect that makes the bridge unsafe not only is not apparent to her but requires years of study of structural engineering to discern. Perhaps in twenty years Joy will fi nally come to understand that the bridge really was unsafe.
Perhaps she will never come to understand it and thus never thank – but always resent – those who in fact prevented her from falling to her death. Would the diffi culty in getting Joy2 to understand the bridge’s condition cast doubt on the propriety of stopping her from crossing the bridge?
Or consider Joy3, who understands that crossing the bridge is unsafe but is in a suicidal frame of mind. She believes, erroneously, that her life is no longer worth living. If she is prevented, she may in several years come to see that she was wrong and be grateful to those who prevented her death. Or she may never come to see that she was wrong – though she was – perhaps because her capacity for understanding her good is impaired. Would these facts about Joy3 impugn the interference with her choice?
Th ese examples are suffi cient, we believe, to illustrate the problem of identifying a threshold point that distinguishes hard from soft paternal- ism. Th e hard paternalist can always maintain that if an agent is choos- ing contrary to her own good, she must have either an informational or a rationality defi cit. And those defi cits can be diff erent only in degree and not in kind from the defi cits that the soft paternalist relies on.15
14 Th is is not the only area of law with such problems. See generally id. For the view that the law must adopt sharp distinctions, see Leo Katz, “Why the Law Is Either/Or” (working paper, on fi le with authors).
15 Th e line-drawing problems in justifying paternalistic interferences with acts have a direct parallel in the free speech area, where governments are frequently motivated to suppress speech because it will mislead, factually or evaluatively, some of the audience to its or to others’ detriment.