Punishment, Consent and Value

Abstract

In this paper I take another look at the view, defended by C. Nino, that we may punish criminals because, by knowingly breaking a law, they have consented to becoming liable to the prescribed punishment. I will first rebut the criticisms usually aimed at this view in the literature, aiming to show that they are inconclusive. They are all efforts to show that criminal offenders in fact do not consent to becoming liable to punishment simply by committing crimes. I then turn to a different line of criticism, which I find more promising. I argue that the moral power of effecting normative changes by consenting reflects the power holder’s value as a person, and show how this idea makes sense of how refusal to recognize that power wrongs a person. I then argue that the “power” of consenting to punishability does not fit that model, and is better explained as reflecting the value of other people, whom the offender has wronged. Hence the power of consenting is not involved in typical cases of wrongdoing.

One view of punishment is that we may punish criminals because, by knowingly breaking a law, they have consented to becoming liable to the prescribed punishment. Punishing them therefore does not infringe any right of theirs. While this consent view of punishment has certainly had its share of criticism, much of that criticism seems to me to miss the point. The goal of this paper is to show why that is, and then present a better reason for rejecting the consent view, a reason based on a certain conception of moral rights.

1 The Consent View and its Critics

As should be apparent to anyone familiar with the recent philosophical debate over punishment, the consent view is due to Nino (1983, 1991). However, in this discussion I wish to abstract away from one important aspect of his view.

Nino believes that a consequentialist justification of punishment would work, were it not for the fact that it permits an unfair distribution of benefits and burdens (1991, p. 266). Consent enters the picture because, he claims, a distribution is not unfair if those it disadvantages have consented to it; and in particular, the distribution generated by punishment is not unfair if offenders consent to their liability (1991, p. 268). However, such consent can obviate the charge of distributive injustice in the punitive context only if certain moral requirements are also met. Nino offers the following list: “the punishment is attached to [the violation of] a justifiable obligation,…the [legislative and adjudicative] authorities involved are legitimate,…the punishment deprives the individual of goods he can alienate, and…it is a necessary and effective means of protecting the community against greater harms” (1983, p. 299). If and only if these conditions are satisfied, he holds, consent is sufficient to render fair an otherwise unfair distribution.

In this paper I wish to set consequentialism aside. The question that concerns me is simply that of whether the consent view as defined above is correct; that is, whether we could explain why punishment does not wrong offenders by appeal to consent. (Note that I will simply take for granted here that punishment does not wrong offenders.) Similarly, I will not be concerned primarily with questions of distributive justice or fairness — nor will I contest Nino’s claim that the moral conditions he mentions (and for all I know others) are also necessary for justifying punishment.

I will retain the following component of Nino’s view: what the offender consents to is to lose his legal immunity from punishment. There are several noteworthy features of this claim. In the first place, the object of the consent is precisely a (legal) normative consequence of an action (the loss of a legal immunity), not a factual consequence. Hence I follow Nino in denying that the offender, strictly speaking, consents to the punishment itself (1983, pp. 295–6). He may believe that punishment unlikely, or even intend to evade it, yet he has consented to the loss of his immunity all the same. Second, consent to lose one’s legal immunity is a conventional matter, determined by what the law happens to say (1991, p. 281) — though not any old law could make the offender’s consent legitimize punishment, as Nino’s additional conditions imply. Hence, talk of the offender’s consent presupposes the existence of a state and a legal system. It does not apply to the state of nature. In the absence of state and law, after all, it is unclear at best both to whom and to what a malefactor consents simply by his wrongdoing. Third, and relatedly, we need not assume that the offender consents to losing his moral immunity from punishment, if that is supposed to be something distinct from losing his legal immunity. It is indeed unclear what such “moral” consent amounts to (see the discussion of the “knowledge condition” in Section 2 below). Insofar as the offender’s consent plays an essential role in explaining why he lacks a right against punishment, that is simply his consent to lose his legal immunity, as defined by convention.

Criticisms of the consent view are typically aimed at showing that there are necessary conditions on consenting not met by typical offenders, and therefore that consent could not explain why punishment violates no right of the offender’s.¹ In particular, we could distinguish three conditions on (valid) consent that offenders typically would not meet: consent has to be voluntary, it has to be intentional, and the agent must know that he is consenting.² These conditions are presumably not independent of one another, but it is still worthwhile to address them separately. They are also, as I hope is obvious, not supposed to be jointly sufficient for consent generally. Rather, I take them up precisely because they are necessary conditions that offenders fail to meet, if critics are to be believed. In the following section I aim to show that it is at least not obvious that offenders do not meet these conditions, once they are properly understood.

Even if we grant that offenders do consent to losing their immunity to punishment, it could still be argued that this consent is invalid, and so lacks normative consequences. As we saw, Nino holds that validity in this sense is subject to certain moral conditions.³ Granting that these are met, though, and (again) granting that criminal offenders are indeed liable to punishment, there still seems reason to doubt that this liability is due to their having consented. Something still seems to be missing. In the two remaining sections I ask what more might be needed. First, in Section 3, I inquire quite briefly, and rather skeptically, into the possibility of dismissing the offender’s consent on the grounds that what one consents to must in some sense be beneficial for one. In the final section I suggest instead that a person’s moral power to effect normative changes by consenting is a reflection of this person’s own value, and that the idea of offender’s consent fits this fact badly.

2 Do Offenders Consent to Lose their Immunity to Punishment?

I turn now to a discussion of the three conditions on consent noted above.

• (1) Intentionality

There is some disagreement in the literature on the precise content of the consent-constituting intention. For instance, Alexander holds that “[t]o consent is to form the intention to forgo one’s moral complaint against another’s act” (1996, p. 166). Owens seems to take a similar position, when he asserts that you consent to X just in case “you (intentionally) communicate the intention of hereby making it the case that someone would not wrong you by X-ing” (2011, p. 408). By contrast, Hurd (1996) advocates a stronger interpretation, according to which consenting to another’s action involves intending that action. This last proposal raises a problem I would rather avoid, that of whether it is possible to intend another’s action. I also prefer to skirt the issue that seemingly divides the first two suggestions, that of whether consent has to be communicated. Even if non-communicated consent is possible, it is presumably rare. I will therefore assume that consent is communicated. (Later I will mention a reason for being skeptical about non-communicated consent.) I focus instead on the idea uniting Alexander and Owens, namely that consent essentially involves an intention to change the moral status of another’s action.

Normally we say that the object of an intention is an action, if only a purely mental one, such as to think about some problem. Hence, if consent involves an intention to change the moral status of another’s action, that “changing” (or “forgoing” or “making”) must be an action. But what kind of action is it? Suppose I say “You may have my apple.” There is disagreement over whether I have done one thing or many —the saying and the consenting could be different actions or distinct descriptions of the same action — but surely there are not two distinct intentions here. Indeed, no doubt we would most likely describe my intention in the way Owens suggests: to consent to the apple’s becoming yours for the taking by uttering the words “You may have my apple.” But how does this differ from a case in which I intend to utter these words without consenting? To sharpen the problem, suppose that I am well aware that my words are typically understood to express consent, and that you and everyone else will almost certainly so understand them (in other words, the knowledge condition is met — see below)? If that action does not amount to consent, what exactly is missing? It is not clear to me what this could be.⁴

Now, I have granted that consent involves communicating a message (that one waives one’s right). But could an action “communicate a message” unless it is performed with the intention of communicating just that message? I answer yes. An action could communicate a message when the agent foresees that his action will be interpreted as communicating the message, but does nothing to prevent that (though he could have). Much the same holds of other performative speech acts, such as marrying people and naming ships. For instance, to count as having named a ship it is not necessary that one smashed the champagne bottle against it while intoning the words “I hereby name you the Queen Elizabeth” with the intention of naming it. It is enough that one realized that in one’s circumstances that action would be universally so taken.

The root of the intentionality condition seems to me to be the consenting agent’s obvious need for control: exercising a power is something one can choose to do. That is why coercion invalidates such exercises, by removing, or at least diminishing, control. And it explains why mere desire is at least not sufficient for consent, as one has little control over one’s desires. Further, even an offender has control at least in the sense that he performs his wrongful action willingly, and intentionally. That gives him all the control he could reasonably ask for. I cannot see that an intention to make himself punishable adds anything of significance, even if such an intention makes sense. What we can say, then, is that insofar as some action is to count as expressing consent, it must be intentional. That is the grain of truth in the intention condition. However, the content of the intention need not be the exercise of a power.

There is a weaker version of the intentionality condition, according to which (valid) consent at least requires that one not intend not to consent. This weaker condition may seem damaging enough to the consent view, on the grounds that punishment is justified even if an offender has such a contrary intention. A related idea is that punishment would not violate an offender’s rights even if he explicitly refuses to consent.⁵ This worry rests on the presumption that consent is always cancelable (e.g. by saying “I do not consent” as one is performing the supposedly consent-generating action). That presumption seems false, however: while consent often is cancelable, that is not invariably so. A good example (of a kind used by Nino (1983, p. 296)) are cases in which a person voluntarily assumes a risk, thereby consenting to bear any cost that may accrue thereby. In ignoring an “Enter at own risk” sign, one could not avoid incurring any liability by simply (verbally) refusing to do so. For another example, I am also inclined to say that a person who signs a contract while saying (verbally) that he does not consent to its terms still counts as having incurred the relevant obligations. It is his signature that matters. More generally, and as I have already made clear, what counts as consenting is largely a conventional matter — though I concede that that term is a bit artificial when applied to the case of offenders. As a consequence, by voluntarily, and in full knowledge of the consequences, committing certain actions a person counts as having brought about certain changes in his normative status, regardless of his intentions.

• (2) Knowledge

I have in effect already endorsed this condition.⁶ That is, I have said that (valid) consent requires knowledge of the conventional, or “natural,” meaning of one’s action. Nino explicitly accepts the condition as well, and uses it to explain why retroactive laws are unjust (1983, p. 299). He faces a problem in the fact that all legal systems subscribe to the dictum that “ignorance of the law is no excuse.” That is, they punish offenders who do not meet the knowledge condition, and thus do not consent to punishability.⁷ We could respond on his behalf by retreating to an attempt to justify only an ideal system of punishment, in which all offenders know the law. Alternatively we could try weakening the knowledge condition to allow also for information the consenting party in fact lacks but ought to have. We could also side with those philosophers who have expressed skepticism about the traditional dictum.⁸

This is not the place to assess these alternative responses in any detail, or the extent to which they might be combined. However, I will at least say a little about the second, which is potentially the least revisionist of the three (which is not to say that I reject the other two). Boonin has objected that it amounts in effect to abandoning the consent view on the grounds that, presumably simply by definition, we cannot consent to consequences of which we are unaware (2008, p. 163). Strictly speaking this seems true enough. The question is whether it is of more than merely verbal significance. Nino’s analogy with the voluntary assumption of risk is relevant here (see above). We would presumably allow that someone who does not know the risk he is running, but ought to know, has to bear the cost he may incur as a consequence of his action, no less than someone who is aware of the risk. We might not want to say that the first of these two persons “consented” to run the risk, given his (culpable) ignorance, but the moral consequence seems to follow regardless. Much the same could be true of the culpably ignorant offender. The similarity between him and the fully informed offender seems more significant than the difference. To be sure, if we take this path we might wish to abandon the label ‘the consent view’, but we would still be retaining the basic idea.

I turn now to a closer consideration of the knowledge condition, asking: which knowledge must an offender have (or, possibly, be in some sense required to have) for him to count as consenting to his punishability? We must first remember a point I granted for the sake of argument: consent has to be communicated. This fact tells us something about what the consenter needs to know, namely that his action counts as communicating his consent. In many instances this communication uses language. A person can say “I consent…” or “You may…” or suchlike. Plainly these utterances count as consent only if the person making them is aware that they are conventionally used to express consent. As a consequence, consent through language requires the existence of conventions. However, it should also be clear that not all consent proceeds by way of conventional, and hence more or less arbitrary, signs. If it did it would be hard to explain, for instance, how two strangers who do not understand one another’s languages could have consensual sex. It is certainly not obvious how to sort out what is convention and what is nature, when the former is not limited to positive law, but this line-drawing issue need not detain us. More important is that I have in effect framed the knowledge condition in circular terms: to count as consenting one has to know that one consents, or that others take one to be consenting. We can easily avoid reference to consent by talking instead directly of its moral effect, the offender’s liability to punishment. However, it seems that one could only know of things that hold independently of one’s knowing them. And if the knowledge is an essential part of consent and consent creates the change in moral status, that principle is violated. The likeliest way out is that what a consenting offender must know is that there are certain legal, or other conventional, facts that hold independently of his consenting. For example, an offender needs to know that by acting in a certain way he breaks the law, and has thereby surrendered his legal immunity to punishment (whether or not he will in fact ever be punished for his deed).

Three comments about the knowledge condition as just formulated are in place. The first two concern the case of punishment specifically, while the last is general. First, a right violator who finds himself in an environment in which there is no law against his action, or even a non-legal convention of retaliating against such violators, does not meet this condition, and thus has not consented. If punishment is justified even in such a case, the consent approach could not account for that fact. Second, it might seem that the knowledge condition requires also that one know something about precisely which punishment one becomes liable to as a result of a given violation. That is, in order for the state to be justified in giving an offender a particular punishment on grounds of consent it is not enough that he knew that he was breaking the law and is therefore liable to some punishment or other; he must also have known that his infraction was subject to just that punishment.⁹ I would say, however, that those cases in which this strong knowledge condition might initially seem to make a difference — namely, cases in which offenders face punishments much harsher than could reasonably have been foreseen — can be dealt with in other ways. For one thing, we have already seen (in Section 1) that justified punishment requires more than mere consent. Third, my formulation of the knowledge condition seems in effect to presuppose that consent is communicated. As I have not ruled out the possibility of non-communicated consent, I face the question of how to understand the knowledge condition in cases of such consent. I have no answer to that question, but again do not want to rule out the possibility that one could be found. (This is the reason I mentioned earlier for being skeptical of non-communicated consent.)

• (3) Voluntariness

For this condition to be plausible, we have to be careful about what exactly is supposed to be voluntary. As Owens has noted explicitly, we must distinguish consenting to something, such as having sex, and wanting it (2011, p. 418). What I want in consenting to have sex with you might not be to have sex with you, but rather simply that you not count as a rapist if we do have sex. (This observation matches Nino’s point, noted earlier, that what one consents to is, strictly speaking, a normative consequence of the consent, rather than some physical treatment.) Yet this point leaves open that a consenter must desire the change in rights his consent brings.

As a second step we should recognize that even if that possibility is actual, it is not plausible to interpret the voluntariness condition as demanding that the consenting person desire the moral change for its own sake. By contrast, quite often when we consent to a loss of a claim we desire it only as a necessary means to getting something else. To adapt one of Nino’s examples (1983, p. 294), if I step into a taxi and say “Take me to the railroad station,” and the driver cheerfully replies “This one’s on us!”, I am not disappointed (though probably surprised). Other things equal, I will probably prefer to get the service for free. This observation leads to a “deflationary” take on the voluntariness condition, similar (perhaps identical) to the one suggested earlier for the intentionality condition: the only kind of “voluntariness” needed for consent is the kind that any voluntary action possesses. In particular, an offender’s consent does not require any “voluntariness” beyond this.¹⁰ We may dispute whether he also desired the resulting moral change, but that question is secondary at best.

Summing up, I conclude that criminal offenders could meet all three conditions on consent by offending, or at any rate that it is not clear that they could not.

3 Consent and Benefit

A natural enough thought is that the ability to consent is in general, and in many of its contexts of use, itself valuable for the person who has it. One would be willing to pay to acquire it if one did not already have it. (However, presumably one could not acquire it by paying. After all, doing that requires having the power to begin with.) But the ability to consent to punishability is not desirable. If anything one would be better off not having it. Is this alleged fact about offenders’ consent a reason for dismissing it as spurious, or at least not as valid?

There are reasons for doubting both that the supposed condition on validity is genuine and that offenders do not meet it, even if it is genuine. In the first place, there can clearly be no such “benefit condition” on individual acts of consent. We are all perfectly capable of consenting to treatment that is in fact, and indeed expectably, harmful to us. To deny this would amount to endorsing an objectionable form of paternalism. To be sure, it may be that the ability to consent in general must be beneficial for the person who has it, if consent is to count as valid. That may be why we treat children as unable to consent validly. But this very general benefit condition does not obviously “distribute” to various contexts of consent, such as the offender’s consenting to liability to punishment. It does not follow that I can consent validly to forgo my claim against some type of treatment only if such treatment is beneficial as a rule, nor would such a demand be plausible in any case.

It could also be argued that offenders do, at least in certain circumstances, benefit from losing their immunity to punishment. Even granting that they typically do not benefit from actually being punished, and hence not from losing their immunity to punishment upon committing a particular crime, they could still benefit from the system of punishment as a whole — because of the degree of protection if offers, or at least ideally should offer, even to themselves. Talk of such “benefits” is admittedly unclear, as the baseline of comparison is not specified. Are we to compare with a Hobbesian (or a Lockean?) state of nature? Or with some alternative form of civil society that does not use punishment at all? And what would such a society in any case be like? I could not answer these questions here (if they even have answers), but at least they leave open the possibility of a penal system from which offenders do benefit as a whole, at least ex ante, even if that is true of no actual system. Further, we might insist that this condition has to be met for such a system to be justified in the first place. It resembles those advocated by Nino and stated in Section 1, and could perhaps be added to these (though I will remain neutral on that score).

All in all, then, appealing to the consenter’s benefit is unlikely to provide any strong objection to the consent view of punishment.

4 Consent and Value

I wish now to consider a different perspective on consent that I believe does cast serious doubt on the idea that offenders validly consent to punishability by offending. To see it, we need first to take a step backwards. The ability to effect normative changes by consenting is a moral power. Understood in Hohfeldian terms, it is a power to extinguish a claim of one’s own, temporarily or permanently, and along with that claim its corresponding duty or obligation, belonging to someone else.¹¹ As is sometimes said, a power is a second-order normative relation, consisting in an ability to manipulate first-order relations (and also relations of higher orders).¹² Now, a crucial question about all such normative relations is that of why they obtain. It is too large to be settled here, yet surely some answer to it is necessary if we wish to identify the normative relations in which persons actually stand. My strategy here, therefore, will be to assume a certain view of such relations — incompletely specified, to be sure — and then draw some conclusions from it relevant to the consent view. The view I assume is inevitably controversial, but by no means idiosyncratic. Indeed, I hope it will strike many as a fairly natural deontological view of such relations.¹³

On the view I assume, a claim is an expression of the claim holder’s value, which I take to mean that it exists in virtue of whatever properties make him valuable in his own right. Or at any rate this is true of a central class of claims, excluding at least claims that hold simply in virtue of an exercise of some power (such as, most obviously, promissory claims, and possibly also property claims). To respect another person’s claims, then, is to respect that about him which makes him valuable in his own right. But then what about powers? What I now wish to argue is the less familiar point that the moral power of consent is also expressive of its holder’s value. (I would say the same of all moral powers, but will not argue the point.)

Just as there are ways for others to respect, and fail to respect, a person’s claim there are ways to respect, and fail to respect, a person’s power of consent.¹⁴ In particular, as that power amounts to an ability to alter the shape of the person’s own claims, to respect the ability in turn amounts to acknowledging these alterations. To acknowledge the alterations is to behave as if claims and their correlative obligations are now shaped the way the consenting agent intended (or at least foresaw). If I consent to give my car to you, the property right in the car has transferred from me to you. For third parties to respect my power of consent (and yours) in this instance is for them to acknowledge this change by henceforth treating you as the car’s owner, just as they previously treated me. It is important to see that this decision whether to acknowledge a consent-generated alteration of rights and obligations matters to whether one treats the consenting agent with the respect due a person. An example can make this point vivid.

Imagine a father who refuses to recognize his adult daughter’s power of consent to have sex. He comes in two varieties. In the first, more benign, version he is overprotective, still thinking of his daughter as “daddy’s little girl.” In his eyes she is still a child, and children of course lack the maturity necessary to make decisions about sex, even if they are old enough to have developed an interest in it. In the second, more sinister, version, the father is extremely patriarchal and regards his daughter as his personal property. Hence he claims for himself the exclusive power of deciding when and with whom she has sex. While the second father no doubt comes across as morally worse overall than the first, in one respect they both wrong their daughter in the same way, namely precisely in that they refuse to recognize her power of consent. We have to be careful here, though, not to confuse this instance of wronging with the related one of violating the daughter’s claim right to have sex with a willing partner (taking for granted that such a right exists). The father would violate the latter right by somehow preventing his daughter from exercising her right (say, by locking her up). To fail to respect her power is rather to treat her consensual sex as in effect non-consensual. This in turn involves treating her sex partners as rapists (in the case of the overprotective father) or as some sort of property violators (in the case of the patriarchal father). Hence the overprotective father mistakenly, and indeed presumably wrongfully, regards his daughter’s sex partners as having wronged her, whereas the patriarchal father even more mistakenly regards them as having wronged himself (and perhaps he does not acknowledge that his daughter can be wronged at all in her own right, any more than other items of property can).¹⁵ But most importantly for present purposes, in either case the father’s failure to respect his daughter’s power of consent involves wronging her by treating her as incapable of deciding over a matter of intimate concern to her, one over which she alone should decide, as an autonomous agent.

Note that the father does wrong his daughter in not respecting her power. Normally we think of a “wronging” action as violating, or at least infringing, a claim. But the key feature of a claim infringement, that makes it a wronging — we are now assuming — is that it fails to respect another person’s value. And this value could be expressed by a power, no less than by a claim. Further, a wronging must plausibly involve giving insufficient heed to some reason. If all that is going on in the case above is that the father behaves as if he had a certain reason to act or relate towards his daughter that does not in fact exist — such as a reason for seeking to punish her sex partners as rapists — it would be hard to explain the sense that he wrongs his daughter. The natural explanation of the wronging is that the father has a special reason to treat his daughter as able to extinguish certain reasons — a sort of second-order reason, if you will — and that he fails to heed this reason. Also, this reason corresponds to his daughter’s value as a person capable of making decisions for herself — or simply her value as an agent.

We may take the father-daughter case just described as a paradigm instance of wronging others by failing to respect their power of consent. The paradigm case is helpful here precisely because it tells us something about what it means to have a power of consent in virtue of one’s agency, and hence what it means for other persons to fail to respect that power, and therefore one’s value as an agent. Hence we have something to work from when we ask whether offenders consent to make themselves punishable simply by offending. To be sure, the offender case differs from the paradigm in one obvious way: should the state refuse to acknowledge the offender’s supposed consent to punishability, that would not (necessarily) amount to a complete refusal to acknowledge his power to consent to such a normative consequence. It might simply amount to refusing to recognize offending as such as a form of consenting. The state could still recognize that a person (offender or not) could consent to punishment-like treatment in other ways — say, by signing a special waiver form. Yet that difference is not obviously decisive here. For as long as all the relevant conditions on valid consent are met in a given case — and I argued in Section 2 that it is at least not obvious that offending typically does not meet the standardly recognized conditions — then refusing to recognize that consent (treating it as invalid) should still at least to some extent wrong the consenting party, even if one is willing to recognize other forms of consent. If that is so, though, we have an argument against the consent view. I say so because I would also assert that the offender in such a case is not wronged. From these two premises it follows that there is no valid consent — no moral power exercised — and the consent view is therefore false. If the offender’s immunity is indeed lost, that is not due to his having consented by offending. Further, there must then be some additional condition on valid consent that he has failed to meet.

This very simple argument is sound, I hold, yet it is hardly satisfactory on its own. It needs supplementation in two ways. In the first place we must say something to those who are inclined to object that the offender is wronged if he is not treated as having lost his immunity (and who therefore deny the argument’s second premise). Second, we must say something about why the offender does not consent validly, which the simple argument itself does not tell us, and this explanation should be in tune with my analysis above of the father-daughter case. (To do this is to identify the just-mentioned additional condition on valid consent not met by offending.) As we will see, these two tasks are related.

Turning to the first task — that of responding to the charge that an offender is wronged after all if he is treated as not having lost his immunity to punishment — I begin by noting that we must be careful in identifying exactly why such a person is wronged, if indeed he is. For on one view of punishment, associated above all with Hegel, offenders have a right to be punished. The most appealing version of that view, defended by Morris (1968), stresses that punishment acknowledges the criminal’s responsibility for his deed, and thereby also his agency, in a way that at least some other reactions to crime do not. If Morris is right — which I will not deny, though I have my doubts — there is reason to believe that a policy of refusing punishment could in one way wrong (responsible) offenders by treating them as not responsible for their actions. However, and this point is crucial (if subtle), even if non-punishment does wrong the offender in this way, it does not follow that it wrongs him by failing to recognize his power of consent to punishability, as opposed to his being responsible for certain actions deserving of punishment. I am inclined to believe (but of course could not prove) that any sense we might have that an offender is wronged when he is not treated as punishable is due to the sense that his right to punishment is violated — or at least, more modestly, that we object to his not being treated as responsible for his actions — and not that his power of consent is ignored. Hence such an intuition does not undermine the above argument against the consent view.

To clarify: if Morris is right, an offender presumably is liable to punishment, as satisfying his right to punishment could not wrong him, but (again) it does not follow that this liability is due to his consent. On the contrary, it would seem that on such a view his being liable to punishment is a consequence of, or is at least in some sense secondary to, his having a right to be punished (meaning a right to be treated as responsible for his actions).

Our second task was to explain why the change in the offender’s status is not due to consent (given that the standard conditions on consent are met). Our recent analysis of the right to be treated as responsible for one’s actions now stands us in good stead. For, whether or not such a right exists, surely the wrongful nature of the offender’s deed (for which he is responsible) plays an essential role in accounting for his changed moral status. (Perhaps his deserving punishment for his deed also plays such a role). In this respect the present case differs from those in which a change in moral status is (uncontentiously) due to consent. Consider, for instance, the case, described above, in which a person consents to “punishability” — here meaning simply that he consents to losing his immunity against being treated like a criminal — by signing a waiver form. In that case, wrongdoing on the consenter’s part is irrelevant. The same is true in general of exercises of the power of consent. The moral status of the consent-generating action itself is not important to whether one has consented validly.¹⁶ Further, the fact that the nature of this action does not matter to the normative effect — indeed, as we have seen, the action is largely identified by convention — shows that its value is not relevant to that effect either. Yet in light of the gravity of the effect, surely some value is at stake; and that, I have suggested, is the consenting agent’s own value. In the offender’s case, by contrast, his value qua agent could at most play an indirect role in accounting for his change in moral status, for it is mediated by the value of the offending action(s) he has performed. It is precisely the lack of a direct role for the agent’s value qua agent that makes the offender’s case differ from one involving a change in moral status due to consent. Hence the second task is met.

Yet a third task may also be performed. For there is an alternative, and more plausible, explanation of why an offense changes the offender’s moral status. It is not (primarily) because the offender is an agent (and so valuable), as the consent view implies if I am right, but rather because other people are, and in particular his victims. Put in the broadest terms, the idea is that what a person can claim from us at any given time depends, to some extent, on how he treats others, and that this fact in turn is a reflection of these others’ value. In other words, the idea is that if we do not let our views about how we are required to treat a person reflect how he has in turn treated others, we fail to properly recognize the value of these others. Note also that the wrongful nature of a crime reflects its victim’s value. It is wrongful precisely because it fails to respect that value.

Summing up: even granting that an offender has consented to his punishment, we have seen reason for thinking that the state’s refusing to treat that consent as valid, meaning its refusal to treat him as having lost his immunity to punishment, does not wrong him, because his value as an autonomous agent is not directly at stake. If we do insist that he has been wronged by being treated as retaining his immunity, that seems to be due not to his having consented, but rather to his not being treated as he deserves, or as being responsible for his own actions. His value as an autonomous agent may still be at stake in such a situation, but only indirectly, mediated by his particular value-bearing actions.

Let me end this discussion by saying something about its implications for the distinction between consent and forfeiture. In the literature these two are typically contrasted. Indeed some go as far as to define forfeiture as a “non-consensual loss of a right.”¹⁷ And others at least argue that forfeiture, unlike consent, need not meet the three conditions discussed above (intentionality, knowledge and voluntariness).¹⁸ Be all of that as it may, though, we have now seen a distinguishing mark. In short, I would say, the crucial difference is that consent is a power, in the sense I have outlined, but the “ability” to forfeit is not.¹⁹