Monday, October 12, 2009
Although Aristotle first distinguished the two ideals more than 2300 years ago, the relationship between corrective and distributive justice has been the topic of a good deal of discussion in recent years. Indeed, I just returned from a roundtable on tort theory where the relationship was debated at length. Larry Alexander, Jules Coleman, Stephen Perry, Ronen Avraham and Issa Kohler-Hausmann, Kevin Kordana and David Tabachnick, and Steven Walt, among many others, all have written about the relationship between corrective and distributive justice. Why has that relationship been thought to be so problematic?
In “Causation and Corrective Justice: Does Tort Law Make Sense” (pdf), for example, Alexander maintains that “[c]orrective justice as an independent, non-derivative principle of justice is…impossible to square with principles of distributive justice because the two types of justice have completely coextensive domains.” (p. 6). Faced with the choice between corrective and distributive justice, he believes that corrective justice must be rejected because distributive justice is “logically prior to corrective justice” insofar as “there must be a distribution relative to which loss and compensation are measured” (p. 7). What I wish to offer here are some thoughts on how corrective justice might vindicate its independence as an ideal with its own normative force.
As an initial matter, even if Alexander is right both that corrective and distributive justice have co-extensive domains of application and that the two sets of principles can therefore conflict, the conjunction of those facts, accepted arguendo, poses no problem for either principle of justice. Reasons of justice, whether corrective or distributive, are certainly important normative reasons, but they are nevertheless just reasons. They can therefore stand in all of the myriad relationships that other normative reasons can stand in with respect to other normative reasons: they can outweigh as well as be outweighed and silence as well as be silenced, for example, all depending on the circumstances. Reasons of corrective and distributive justice are not all-things-considered final judgments of what must be done, but (admittedly important) elements in an all-things-considered final judgment. As general conflicts of reasons are ubiquitous, it should not be surprising that reasons of justice themselves can conflict – it is widely recognized, after all, that justice can conflict with other values, like beneficence and mercy. Such conflicts, though, are never evidence that one of the conflicting reasons must really be an imposter and not really a reason at all. And so, it seems to me, there is no embarrassment in admitting that reasons of corrective and distributive justice can conflict.
What is likely really doing the work in Alexander’s argument, then, is the priority of distributive justice. If distributive justice is in some sense prior to corrective justice, how can corrective justice generate any non-derivative moral reasons of its own? The problem on this reading isn’t so much the conflict itself between reasons of corrective and distributive justice, but the fact that reasons of corrective justice are only bona fide reasons if they do not conflict with those of distributive justice, on which they depend. Coleman captures the thought nicely in "Risks and Wrongs": “[I]f corrective justice provides moral reasons for repairing a loss, then the underlying claims sustained by corrective justice must themselves express requirements of distributive justice….[T]his relationship appears to rob corrective justice of its moral independence.” (p. 348).
Towards both characterizing and solving the puzzle of corrective justice, consider an analogy with the duty to obey the law. Where a state’s laws are just and the persons under the state’s dominion are responsive to moral reasons, the duty to obey the law is either epiphenomenal or non-existent, depending on how one understands the duty to obey. It is epiphenomenal, like a shadow, if one understands the duty loosely, because where the law is just one would already be morally required to do as the law commands. It is non-existent if one understands the duty to obey strictly, as on Joseph Raz’s influential view, for it would be wrong to obey just law, following it merely because it is the law – one ought instead to conform to just law, following it because it is just. The puzzle with corrective justice is analogous: where distributive justice obtains, reasons of corrective justice are either entirely epiphenomenal, adding nothing to antecedent reasons of distributive justice, or non-existent, because the reasons anyone would have would simply be those of distributive justice. The puzzle of corrective justice therefore appears to be analogous to that of the duty to obey the law: the independent power of corrective justice seems to be circumscribed by prior norms of distributive justice, much as the duty to obey the law is circumscribed by prior duties of morality.
Might the solution to the puzzle of corrective justice also be analogous? Even though Raz denies that there is a general obligation to obey the law, he accepts that there can be a piecemeal one where a particular law is either modestly unjust or simply not just; the duty to obey can carry citizens over rough patches of low-grade injustice and also direct them to act in cases where there is no antecedent moral reason, as with coordination or regulatory problems. Perhaps corrective justice serves a similar role. First, corrective justice and the value of personal accountability that it embodies may be important enough in its own right to carry citizens over rough patches of low-grade distributive injustice. This would help answer Coleman’s doubt that corrective justice’s dependence upon distributive justice “greatly reduces [corrective justice’s] potential as a principle underwriting our current legal practices, practices that do not require that underlying entitlements protected by law meet the strictest test of justice” (p. 348). Second, corrective justice also may be able to direct people to act where distributive justice itself is indeterminate. This would afford corrective justice a robust independence, for as Perry points out in “On the Relationship Between Corrective and Distributive Justice,” plausible theories of distributive justice are simply too dynamic to be able to issue guidance about specific duties at any given moment. Corrective justice therefore seems to have enough independence from distributive justice to allay any deep worries about its ability to be a non-derivative ideal, and of course to keep us from embracing eliminativism about corrective justice and thus tort law.
Associate Professor of Law
Co-Director, Rutgers Institute for Law and Philosophy
Rutgers School of Law - Camden