Although this article appears to be about the permissibility of aggressive anticrime measures, it really concerns particularism in ethical theory.
Three strikes laws, zero tolerance policies, and the like have been criticized for violating nonconsequentialist restrictions on punishment. According to this criticism, when the state sentences a former felon to 25 years in prison for stealing three golf clubs, see Ewing v. California, it acts improperly because it imposes a sentence more severe than the defendant deserves for the crimes he has committed. That such a restriction on punishment exists, it is claimed, can be shown by considering telishment (short for “teleological punishment”), in which the government intentionally punishes someone it knows to be innocent because crime will thereby be prevented. Although telishment would, ex hypothesi, have good consequences, it is nevertheless wrong. And so–-proceeding according to what Brand-Ballard calls a generalization argument–-punishment under three strikes laws must also be wrong, for it shares all morally relevant intrinsic features with telishment. Both, after all, impose a quantum of undeserved harm.
Brand-Ballard challenges not merely this criticism of three strikes laws, but also all generalization arguments, in which a policy is held impermissible because it shares all its morally relevant intrinsic features with another policy that is clearly impermissible. He asks us to consider an act of punishment that is clearly permissible–say, a 5-year prison sentence for bank robbery. Since this sentence inevitably imposes harm upon the punished’s spouse (the “collateral damage” of the article’s title), he argues it also shares all morally relevant intrinsic features with telishment. The spouse, like the telished, does not deserve harm, although imposing it might have good consequences.
Of course, Brand-Ballard’s argument can work only if the harm to spouses that arises from imprisonment shares all morally relevant intrinsic features with telishment. The most important challenge he must face is that the doctrine of double effect is a morally relevant intrinsic difference between the two. The state, one might argue, does not intend to harm the spouse, but merely foresees the harm as a side-effect of the imprisonment. In contrast, the telishing state intends the harm suffered by the scapegoat. Although Brand-Ballard does not deal with this objection with quite the care one might wish, I think he is right that the doctrine of double effect cannot be used to distinguish the two.
His best argument is that the coercive separation of the spouse from the prisoner is often an intended quality of the punishment. The state wants the prisoner to be denied the companionship of the spouse. To be sure, one might argue that it does not really want the spouse to be denied the companionship of the prisoner. If it could separate the prisoner from the spouse without separating the spouse from the prisoner, it would. But here I think Brand-Ballard could respond that we might just as well say that the telishing state doesn’t really want the harm to the scapegoat–if it could bring about the crime reduction without the harm, it would.
Given that telishment, three strikes laws, and imprisonment without unlimited conjugal visits share all morally relevant intrinsic features, what does Brand-Ballard think follows? Not that prisons must allow unlimited conjugal visits, and not that telishment should be permitted. He does not even take a stand on the permissibility of three strikes laws. His point is that our moral assessment of all of these policies cannot be made without a consideration of their extrinsic features.
Notice that this is not necessarily consequentialism. He leaves open the possibility that considerations other than consequences matter. Rather, he borrows a theme from the particularists: the way that intrinsic considerations matter is never invariant across extrinsic contexts.
The common view is that telishment is wrong because it violates nonconsequentialist restrictions protecting the scapegoat. Likewise, imprisonment without unlimited conjugal visits is fine, because the prisoner’s spouse has no comparable nonconsequentialist protection of her interests. Brand-Ballard argues that this is too hard on telishment and too easy on prisons’ refusal to allow unlimited conjugal visits. Both share the pro tanto wrongmaking feature of imposing undeserved harm. Therefore, what makes telishment all-things-considered impermissible and imprisonment without unlimited conjugal visits all-things-considered permissible are extrinsic features.
He offers only brief suggestions about what these extrinsic features might be. In the case of telishment, he suggests that, in any normal society, allowing officials to telish would open up an avenue for abuse–they would have a convenient argument for punishing those with whom they disagree. Concerning unlimited conjugal visits, he points to the cost such a policy would impose upon prisons. He admits that these reasons may not reflect the phenomenology of our moral reactions to these policies. But, he argues, to the extent that we think nonconsequentalist considerations are solely driving these reactions, we are misdescribing the sources of our own moral intuitions.
What then about three strikes laws? He argues that the common criticism of such laws–that they violate retributive restrictions on punishment–will not work. Any argument against them is going to have to be contextual and empirical, in particular that they inflict more undeserved harm than they prevent.
The most significant aspect of this highly original article, however, is less its assessment of three strikes laws than its exemplification of the modern trend toward moral particularism. Notice, however, that Brand-Ballard’s particularist sympathies are limited. The way that intrinsic features morally matter, he argues, is never invariant across extrinsic contexts. That leaves open the anti-particularist possibility that there are extrinsic features, or nontrivial sets of intrinsic and extrinsic features, with moral effects that are invariant across contexts. From these features, moral principles might be derived.