a Rawlsian approach to punishment

John Rawls, though setting out a liberal theory of justice, stopped short of justifying a particular view of punishment. Still, Sharon Dolovich is convinced that we can use the Rawlsian veil of ignorance to legitimize punishment in a liberal society. LDers contemplating the current hate crime resolution might take note of her article, "Legitimate Punishment in a Liberal Democracy," found in the Buffalo Criminal Law Review, available online [pdf].

Dolovich first warrants the use of deliberation in the "original position," in contrast to most punitive schemes, which operate in an ex post framework.* She writes,

[I]t is a basic assumption of the argument I develop--and indeed, of liberal democracy itself--that all members of society are moral equals, entitled to due consideration and respect as fellow human beings and fellow citizens. From this assumption, it does not follow that all citizens are entitled to equal treatment. To the contrary, by their actions, individuals may forfeit certain goods that other citizens enjoy. But on the theory of liberalism I adopt here, forfeiture in this sense does not negate an individual’s moral status: he or she is still a subject of justice, entitled to consideration as such.

This means that any legitimate theory of punishment necessitates the prior deliberation ("behind the veil") of those who, someday, will flout the law. Why should we allow the possibility that a reasonable deliberator might err when in the "real world?" Three reasons, Dolovich argues. First, imperfect legal sanctions mean that sometimes innocents will be punished. Second, people are people, and people make mistakes. Third, the unequal distribution of resources means that certain worse-off members will be more prone to commit crimes simply due to circumstances. (It is important to recognize that even a just society in a Rawlsian sense is not perfectly equal.)

Dolovich spends pages analyzing some of the potential results of the behind-the-veil deliberations, and the principles of just punishment that would result.

1. There shall be no incarceration for non-serious offenses, unless doing so would appreciably deter the commission of serious offenses.
2. Punishments of incarceration, when imposed for serious offenses, may be only as severe as necessary to appreciably deter offenses causing harm of equal or greater severity (the parsimony principle).
3. Before any punishment may be imposed, its deterrent effect must be shown to be reasonably certain or imminent, on the basis of standards and modes of reasoning acceptable to all.
4. Consistent with these principles, the state must do all it can to reform the criminal justice system in order to reduce as much as possible the danger of convicting the innocent or retaining them in custody.

The parsimony principle is central to concerns of proportionality, and relies on both retributivism for its moral grounding and utilitarian considerations for its deliberative outcome; in other words, the original architects of a legitimate punitive system would know that crime deserves punishment, but would look to deterrence considerations for determining the proper proportion, because of the three qualifications on human conduct listed above. Ultimately, the needs of the law-abiding set the ceiling for punishment, as Dolovich argues:

If any punishment is to be legitimately imposed, those parties advocating its imposition must demonstrate convincingly, in terms that all could be expected to accept, that this imposition on the security and integrity of targets of punishment is immediately necessary or at least reasonably certain to result in greater protection for the law-abiding.

Though Dolovich doesn't list them specifically, hate crime enhancements, which often carry mandatory minimums, may run afoul of the parsimony principle.

If legislatures are to honor this principle, they must maintain the flexibility to rethink sentences in particular cases, or indeed to rethink the whole legislative approach to punishing certain offenses when the circumstances demand such reevaluation. Of particular concern in this regard are omnibus mandatory sentencing schemes which lump together a range of offenses and prescribe the same minimum sentence for each. Such schemes preclude the possibility for the focused consideration of the characteristics of each offense and the likely deterrent effect of the prescribed punishment which application of the principles demands.

Last, I should emphasize a few things. First, this summary of some of Dolovich's core arguments hardly begins to describe the complexities of her arguments, which is why I've linked to the original. Second, these arguments are Dolovich's, and not Rawls'. Finally, the advantage of a Rawlsian framework is its grounding in democratic principles. Theories such as utilitarianism and retributivism do not always clearly define the role of the state in providing punishment, especially in a state where laws are democratically created and changed. Whether hate crime enhancements meet Dolovitz's ideal is a matter for the affirmative--and negative--to consider.



*Sharp-eyed readers will note that, over at the other post, I originally thought that Rawls wouldn't "come near this resolution." I was partly right: not Rawls, but a Rawlsian, changed my mind.