In Chapter 5 of Democratic Rights, Corey Brettschneider continues his project of showing how the value theory of democracy generates support for individual rights that serve as substantive limits on the outcomes of democratic procedures. In this chapter, he asks how a democracy might justify the punishment of criminals, one of the most dramatic and severe forms of state coercion. In the course of answering this question, Brettschneider concludes that there are some forms of punishment that a democratic citizen reasonably could reject. A commitment to democracy, therefore, entails substantive limits on the state’s right to punish convicted criminals.
In this engaging chapter, Brettschneider elegantly and economically covers a lot of ground. The various arguments about the limits on the right to punishment are interesting in their own right, but they also usefully illuminate the wider theory of democratic contractualism. By following Brettschneider’s discussions of punishment, we get a better view of the kind of justificatory strategies his theory permits or denies. In the course of my discussion, I’ll try to raise questions not only about Brettschneider’s specific conclusions, but also about the wider theoretical issues that have already been nicely laid out by the discussants of the earlier chapters.
The basic question of the chapter is to what extent a democracy may legitimately punish convicted criminals. (We assume, for purposes of this chapter, that the procedures leading to conviction are just, and are as accurate as we can reasonably expect.) Following his general theory of democratic contractualism, Brettschneider approaches this question primarily by asking “which punishment a criminal qua citizen reasonably could accept” (97). The test for legitimate punishment is hypothetical, reasonable acceptance–not actual acceptance by criminals. Criminals are entitled to justification (on which more below), but not to an absolute veto. Consistently with Brettschneider’s broader emphasis on democratic citizens’ status as free and equal, he argues here that “this account of justification requires rights of criminals against punishments that undermine their status as citizens” (98).
Before approaching directly the question of which punishments might be justified, Brettschneider lays the groundwork for his approach by dismissing the view that criminals have no entitlement to justification because their crime has dissolved the contractual relationship. To extend the contract analogy, such a view holds that crime amounts to a “material breach” by the criminal which relieves the state from any contractual obligations, including justification, toward the criminal. Brettschneider, by contrast, insists that the state is never released from its obligations to justify coercion to its citizens. Even the most egregious breaches by citizens of their obligations do not strip those individuals of their citizen status, and so do not relieve the state of its obligation to justify punishment.
I think Brettschneider is right that the state always must justify its coercion, no matter how evil the object of coercion. I would think this obligation to justify extends even to non-citizen enemies (e.g., in a just war). The key question, as Brettschneider recognizes, is whether the unique justificatory obligations of democratic contractualism (as opposed to the general justificatory obligations we owe to humanity at large) continue to apply to a citizen who has committed a serious crime. Brettschneider says yes, and I’m inclined to agree: if there are special kinds of justifications we owe to fellow citizens in a democracy, then we owe those kinds of justification to criminals our democracy intends to punish. Presumably the content of the justifications for coercion we can give to criminals will be different from those we could give to law-abiding citizens-at least, that will be so if punishment is ever justified-but the basic requirement of justification persists.
I have a small quibble with Brettschneider’s description of Hobbes’s position in this section, a quibble which I hope is not just of antiquarian interest, because I think it raises a broader question about how Brettschneider conceives of contractualist justification. Hobbes is introduced as a proponent of the view that the state need not present a justification for punishment that is acceptable to the criminal (98-99). Brettschneider takes this to be the case because Hobbes famously thought that citizens always retain their right to self-preservation, and thus have a right to resist capital punishment. But does this right of resistance entail that there is no justification for such punishment? I’m not so sure. As Brettschneider recognizes, Hobbes does try to justify capital punishment, on the general deterrence grounds that punishment helps secure “the preservation of them all” by “disposing . . . men to obey the law” (Leviathan, chap. xxviii). Why is this not a sufficient justification for the criminal himself? Assuming Hobbes is right about the deterrent effect of the punishment, why should we consider the criminal’s rejection of the application of the punishment in his case to be a reasonable rejection? Even if we accept that the criminal has a right to resist the punishment, this does not commit us to saying that the state is wrong to inflict the punishment if it can. Hobbes certainly doesn’t think the state would be committing any wrong in such a case. While the state’s right to punish does not derive from the contract, because the criminal never waived his right to self-preservation, the right to punish is a natural right that the sovereign has not waived, and thus is consistent with the social contract, while the use of the right even promotes the ends of the contract (general security). One might reply that the general deterrence argument is unacceptable because it rests too much on consequentialist reckoning that insufficiently takes the condemned criminals’ interests into account, but leaving the argument at that may prove too much, as it would seem to rule out a large swath of public policies that coerce and harm some individuals in order to promote common ends. So apart from issues of Hobbes exegesis, I think there’s a question here about the reasonability of self-regarding rejections.
Having established that punishment requires justification, Brettschneider goes on to deny that retributivist theories of punishment adequately provide such justification. Such theories, he says, fail to recognize criminals’ status as citizens, a failure which largely stems from the theories’ apolitical focus on the desert of individual criminals rather than on the special problem of justifying state coercion. The retributivist, says Brettschneider, focuses on what should happen to the criminal, agent-neutrally, rather than on what punishment the state as agent may legitimately inflict.
Brettschneider uses three thought experiments to illustrate his point that claims about the moral desert of persons are insufficient to establish conclusions about legitimate punishment by states. If retributivism only provides support for the former, it does not do enough to support the latter, which is what a democratic theory of punishment requires. In the first experiment, we have a properly convicted, child-molesting murderer on death row who is killed by a cellmate outraged by the child molester’s crime. For purposes of argument, we assume that, by retributivist lights, death is an appropriate punishment for molestation-murder-death is what such a criminal “deserves.” Surely we would still object to killing by the cellmate. But with its focus on desert, says Brettschneider, retributivism does not have the resources to object to the killing, and so “cannot acknowledge a moral distinction between state punishment and private punishment carried out in the same manner” (102).
The second and third of Brettschneider’s thought experiments both try to drive a wedge between our intuitions about what criminals abstractly deserve, and our intuitions about what punishments the state can legitimately inflict. In the second, he invokes the case of Osama bin Laden. Brettschneider says, apparently with some sympathy, that a private person “well might believe that a mass murderer such as bin Laden deserves brutal, spontaneous, even bizarre retribution” (104). But, he says, it would clearly be illegitimate for a state to impose such a “ghastly death.” In the third experiment, Brettschneider discusses the principle of lex talionis, or “an eye for an eye.” Victims of crime, he says, might think it appropriate to retaliate in kind for their injuries. But such “deliberate infliction of pain” by the state would be illegitimate. In both the bin Laden case and the lex talionis case, then, we see that, contrary to certain intuitions about desert, the use of harsh punishments by the state is generally illegitimate. Desert therefore cannot be a sufficient condition for legitimate punishment.
I share the intuitions Brettschneider invokes in these experiments, so I support the propositions that (A) private retribution is generally wrong; and (B) it would be illegitimate for the state to inflict the harsh punishments described in the second and third thought experiments. But I’m not sure this is enough to dispense with retributivism, and not just because some people will have intuitions about the goodness of wrongdoers’ suffering that are stronger than their intuitions about legitimate state coercion. I think retributivists can consistently support both propositions (A) and (B). As for (A): I don’t think the retributivist is committed to the view that if a criminal deserves suffering S, then anyone, including private persons, may permissibly inflict S on the criminal. Some retributivists may believe that what wrongdoers deserve is not suffering per se but punishment, and may think that punishment conceptually involves some claim to right authority-a claim which, in ordinary circumstances, only the state may make. But a retributivist could presumably also reject private retribution for any number of reasons, including the basic pragmatic reason that practices of private vengeance often spiral out of control, causing disproportionate harm (sometimes to innocents). These reasons may not be themselves intrinsically retributivist, but they seem consistent with retributivism. I think Brettschneider is right that the retributivist owes us an explanation of why the state is entitled to make and execute judgments about what wrongdoers deserve, but I’m not sure that he has established that this is an impossible task.
What about (B), the claim that the state cannot legitimately impose certain harsh punishments? I think a sensible retributivist would accept (B). Instead of abandoning the idea that desert implies legitimate punishment, however, I think the retributivist would simply deny the original intuition that wrongdoers deserve such harsh punishment in the first place. It is simply not true, she might say, that Osama bin Laden deserves a brutal, torturous death, nor that criminals generally deserve an eye for an eye. (She remains a retributivist, so she thinks they deserve some suffering; just not that.) We may be able to understand why victims or their families desire that wrongdoers suffer terribly, and we may not blame them for such desires. But this is not to say that we would judge these desires to reflect an accurate account of what the wrongdoers deserve. So all but the harshest retributivists could support (B), that very harsh punishments are illegitimate, precisely because such punishments go beyond what criminals deserve. But they could hold on to their view that punishment is justified if and only if the criminal deserves the punishment.
Brettschneider makes what I take to be one other criticism of retributivist theories. He says that retributivism inappropriately depends on a “comprehensively moral code” rather than a “narrower conception of politically legitimate actions” (102-03). We have already seen that retributivism, as Brettschneider sees it, fails to explain why states should punish. This argument extends the complaint about retributivism’s “apolitical” nature, by claiming that it is not possible to provide a justification for state punishment that is recognizably retributivist. Any retributivist justification must refer to some theory about individual moral desert, which Brettschneider thinks would depend on a “comprehensive” morality. For familiar Rawlsian reasons, “democracy’s public reason” precludes recourse to such a view when engaging in public justification. Unfortunately, Brettschneider does not dwell on this point, but I’d like to know what it is about a theory of moral desert that makes it inevitably “comprehensive” in an inappropriate way. Is it really impossible to develop a “political conception of the person,” to use Rawls’s phrase, which includes basic ideas of responsibility sufficient to ground a workable theory of desert? If not, why not?
In a similar vein, I’d be interested to know what kinds of arguments in support of punishment are consistent with “democracy’s public reason.” Retributivism appears to be ruled out, and it seems that general deterrence is under severe strain, if not absolutely ruled out. (Perhaps deterrence concerns are acceptable under “democracy’s public reason,” though the consequentialist reasoning may violate the “inclusion principle.”) What reasons for punishment are acceptable in the first place? Is it only incapacitation? This may be a bullet easy to bite, but I’m curious to know the answer. If there aren’t many very acceptable reasons for punishment in the first place, the issue of what punishments are rejectable under the inclusion principle becomes a bit less pressing.
After his discussion of retributivism, Brettschneider goes on to examine specific democratic rights against punishment-that is, specific substantive limits on a democracy’s right to punish. Brettschneider accepts the legitimacy of restraint, so long as “it respects criminals’ continued status as citizens to the greatest extent possible” (105). Of course, deciding what this principle requires is a fairly context-sensitive matter, but Brettschneider argues that the principle rules out wanton infliction of pain and disproportionately harsh security measures. The focus on criminals’ continuing status as citizens also leads him to support fairly broad free speech rights for prisoners (consistent with basic security), to reject lifetime felon disenfranchisement, and to propose some limited voting rights even for incarcerated felons. I tend to find Brettschneider’s arguments in this section pretty compelling, so I won’t add much here.
After discussing the rights of prisoners, Brettschneider turns to the case of the death penalty, arguing that “capital punishment is inconsistent with respect for the status of criminals qua citizens” (103). Paralleling his comments on retributivism, Brettschneider eschews any abolitionist arguments about the human dignity of wrongdoers, as he sees such claims regarding “the moral desert of persons” as insufficiently political for the purposes of democratic contractualism. (Again, I’d like to hear more about why these moral arguments are inappropriately “comprehensive.”) Instead, Brettschneider proposes two arguments against capital punishment that he sees as more political: an argument from state fallibility and an argument from the interminability of the state’s obligation to justify coercion to its citizens.
First, the argument from fallibility: The state must not kill the innocent. But because the state (and its lawyers and judges and jurors) are fallible, there is always the chance that somebody convicted (even according to the most scrupulous procedures) might be innocent. A convicted criminal should always, therefore, have a right to appeal to prove his innocence. But “a procedural right presupposes an appellant who is alive” (109). This has the slightly odd consequence that the right to life is predicated on the apparently lesser procedural right to appeal and prove innocence, but Brettschneider is unembarrassed, arguing that this is perfectly appropriate for a theory that rests on political argument rather than on “divine or even moral justice in general.”
I think the fallibility argument has some plausibility, but notice that it depends on the principle that there are some circumstances in which the state may not engage in coercion even though there is a very high probability that the coercion is in fact justified. (For purposes of this argument, assume that capital punishment would sometimes be legitimate for the truly guilty.) This might be right, and the death penalty seems to be a reasonable candidate for application of this principle, but I wonder how far such a principle would extend. We may simply say “death is different” (perhaps because of the impossibility of compensation) in hopes of cabining the principle’s reach and preventing policy paralysis, but we’d still need an account of how the principle applies to other state policies involving risk of death.
Brettschneider’s second argument against capital punishment applies even in (counterfactual) cases of certain guilt. A citizen has a “fundamental” right to have coercion justified to him or her. “This right is so fundamental that it should be considered inalienable. If citizenship is inalienable, however, one cannot alienate one’s own right to life either, because citizenship presupposes personhood” (110). The need to justify coercion, says Brettschneider, presupposes a person to receive justification. The state may not kill a criminal, because by doing so it would impermissibly terminate the relationship between citizen and state (111).
I’m not sure I follow this argument. Let’s accept that justification presupposes a living recipient of justification. The convicted criminal is alive when convicted and sentenced, and the sentencing ideally involves a justification of the punishment handed down. The fact that the penalty is death does not prevent the state from justifying the killing itself to the convict. To be sure, once the convict is dead, there can be no more justification-but nor is there any more coercion by the state to justify. So if there is some non-rejectable justification for the death penalty, the fact that justification presupposes living personhood does not seem to undercut that justification. We are forced to directly address the question of whether there is in fact any suitable justification for the death penalty-I don’t think we can retreat to the more fundamental point that justification requires life. It does, but justification for a discrete, temporally delimited act, like killing, does not presuppose indefinitely continuing life.
Brettschneider uses Rousseau’s discussion of capital punishment in The Social Contract to illustrate his argument, but I’m not sure Rousseau makes any obvious errors. Brettschneider points out that Rousseau claims that the criminal is put to death “less as a citizen than as an enemy” (111, citing Social Contract, Book II, Chap v). Brettschneider suggests this comes close to an absurd form of “idealism, in which an individual can continue to exist as a citizen without being a person” (111). I agree that such a view would be silly, but I’m not sure Rousseau is committed to it. The “citizen/enemy” passage follows a fairly straightforward rationale for the death penalty: namely, that ex ante, all would endorse a regime which included capital punishment as a means to promote security. Rousseau believes that the death penalty is not rendered reasonably rejectable simply because, after committing the crime, the criminal realizes that he is going to bear the burden of this policy, and doesn’t want to. Whatever we think about the content of this argument, the justification does not depend on any future persistence of the citizen. (By the way, Rousseau’s support for capital punishment is conditioned by the proviso that we may put to death only those “who cannot be preserved without danger.”) Again, I think there is no option but to confront directly the proposed justifications for the death penalty, like deterrence; the future need to justify seems like too slim a reed, and the “value of never terminating the relationship between citizens and the democratic state” (111), without more, risks begging the question against proponents of capital punishment. It might be more fruitful for abolitionists to rely more heavily on arguments about the general wrongness of killing. One feature of this wrongness might be that it terminates a valuable political relationship, but that is unlikely to be the only feature, nor even the most prominent one.
As I hope these lengthy comments have shown, I really enjoyed reading and engaging with this chapter. Brettschneider has given us a lot to think about, and I look forward to future discussion.