More Thoughts on the Retributive Urge

My previous post summarized some fascinating psychological research on the urge to punish norm-violators.  I raised the question of whether and to what extent the criminal-justice system ought to try to satisfy this urge.  My thoughts on this question are still evolving, but I continue to feel, as I have for some time, both that just deserts ought to be regarded as a central aim of the criminal-justice system and that the system must beware of the risk of retributive rhetoric and emotions spiraling out of control into penal excess.

Let me unpack this a little.  The legal system exists primarily to keep the peace, to provide an alternative to the cycles of bloody private vengeance so vividly represented in Aeschylus’s Oresteia.  If the legal system is to preempt private vengeance and other forms of violent self-help, then it must be responsive to widely shared intuitions regarding who deserves what degree of punishment.

This does not mean, however, that the system should simply seek to gratify the punitive urge in full in every case.  My key premise is that there is no clear, stable quantum of punishment that is necessary to satisfy the intuitive desire to punish in most cases.  This premise presents both a risk and an opportunity.

The risk is that punishment will not really satiate the retributive urge, but will make it grow.  Jim Whitman writes very eloquently about the “intoxication” of punishment, and his work has always resonated with me.  I think we can see some evidence of the risk in the wave after wave of laws increasing sentencing severity over the past three decades.  Each new sentencing law raises the bar, so the next sentencing law targeting the next category of offense or offender that sparks public outrage must be that much harsher.  And when sentences measured in decades become routine, there is little hope that punitive urges will be satisfied by any sentence that does not involve a long term of imprisonment.

There is also opportunity.  If there is some malleability to what the retributive urge demands, then there is a possibility that careful framing of the issues may in some cases moderate the severity of punishment that must be delivered in order for the criminal-justice system to maintain its credibility as a substitute for private vengeance.  Indeed, one might imagine a progressive lowering of the severity bar over time — this is what Antony Duff calls the “decremental strategy.”

Why should the system seek to moderate public demands for punishment?  At one level, this might be seen as antidemocratic.  One might argue that the government should follow not modify public preferences.  What I suggest admittedly has a paternalistic flavor, although I’m not sure the populist-democratic alternative — “if what the people want is ever-greater punishment, then that is what the people shall have!” — could prove stable or coherent in practice, much less morally attractive in principle.

Here are some reasons the system should seek to moderate, rather than merely satisfy, public retributive urges:

  • Punishment may have an addictive character, as Whitman suggests; the more we punish, the more punishment we may demand — and if this is true, there is no such thing as an option for the government merely to follow public preferences; whatever the government does will affect public preferences, either for good or for ill.
  • The retributive urge comes from one part of our minds, but other parts of our minds are apt to produce quite different ways of thinking about punishment (as discussed in my previous post), and there seems no good reason to privilege our intuitive responses over our consciously reasoned responses — indeed, as Aristotle argued long ago, we might view our capacity to reason as what makes us most distinctively and fully human, and conclude that our reasoned responses ought to be encouraged and privileged as much as possible.
  • The practice of punishment is deeply at odds with our commitments to individual liberty and equality — although civilization seems inconceivable without a state-sponsored penal system of some sort, we may justifiably feel that punishment ought to be restrained as much as possible in our society in light of the threat it poses to the integrity of some our most cherished political values.
  • As Nietzsche observed, the urge to punish is deeply intertwined with a rather ugly and unhealthy set of emotions; the state should minimize the extent to which it validates and reinforces these emotions.
  • The state take care about responding to the violation of pro-social norms in ways that undermine or devalue other pro-social norms, like forgiveness, mercy, emotional restraint, and self-discipline — giving in fully to punitive urges may inculcate precisely the sort of unrestrained, thoughtless self-seeking that underlies much criminality.
  • Punishment is costly, both in terms of the financial burdens it imposes on taxpayers and in terms of the psychological and other harm it imposes on the person being punished and his or her family members and other associates.

I think Jonathan Simon’s metaphor of the reactor and the radiator is quite helpful.  A radiator dissipates heat, while a reactor intensifies it.  Simon argues that the law of homicide can function either as a reactor or a radiator of penal heat; an appropriately nuanced homicide law serves as a radiator by recognizing important distinctions among killers based on different degrees of culpability, provocation, and so forth.  Simon contends that the homicide law of California has for a generation failed to make such distinctions, which has served to build up an indiscriminate and escalating public fear and anger toward homicide defendants — a reactor in action.

I’m interested in broadening this line of thinking from homicide law to our criminal law and penal practices more generally — what do we need to do so that the system functions less as a reactor than a radiator, both giving meaningful play to retributive emotions and making those emotions more restrained and discriminating?

Here are a few preliminary ideas:

  • Reinforce traditional distinctions in criminal law based on harm, culpability, justification, and the like.
  • Bring more of what is important in criminal litigation out into the open — this may mean less or differently structured prosecutorial discretion in charging and plea-bargaining.
  • Ensure that there are fair opportunities to humanize the defendant at sentencing.
  • Involve victims in the process in constructive ways; restorative justice may be a highly effective way of doing this.
  • Use sentencing guidelines that are built around that same sorts of distinctions (culpability and the like) that have traditionally been so important to substantive criminal law.
  • At sentencing, attend both to the costs of different penal options and to sentences imposed on similarly situated offenders in other cases.
  • Beware of the risks of shaming and other sanctions that tend to stigmatize defendants in particularly acute ways.
  • Provide opportunities for defendants to redeem themselves and to earn early release or less severe treatment.
  • Eliminate or strictly cabin the availability of the most severe sentences, death and life in prison.
  • Run prisons as if we genuinely want and expect inmates to be successfully reintegrated into the community after release.
  • Enforce rigorous limits on sentence length based on proportionality — incapacitation and deterrence goals should rarely if ever be regarded as an adequate justification for a sentence that is disproportionate to the defendant’s culpability.