Release, retribution and risk
Here are some thoughts on release, retribution and risk, stimulated by my participation over the last few days in a symposium on ‘Prisoner Resettlement in Europe’ at the University of Greifswald by our very own Ineke Pruin…
Early release schemes cause controversy in many jurisdictions. Critics often suggest that they are an affront to justice (and to victims) and that the bring the criminal justice system into disrepute because they mean that the sentences passed by the court — typically expressed in terms of years of imprisonment — are very rarely served as stated. For example, in Scotland, someone serving a 12 year sentence of imprisonment might be released on parole at 6 years and will be released on a non-parole licence no later than 8 years. In such circumstances, early release violates the need for ‘truth in sentencing’ or so it is argued.
How then might early release be justified? The problem is perhaps most acute for retributivists, since they justify punishment retrospectively — with regard to the severity of the crime — and that retrospective judgment of deserts can’t be changed by anything that happens during imprisonment, for example to reduce risk. If 12 years is the proportionate, parsimonious and just penalty — all things considered — then 12 years it should be.
But what things should be considered in determining proportionality? Typically, in deliberations about sentencing, we allow a number of factors to diminish the proportionate severity of the sanction. Some of these factors relate to the degree of culpability for the criminal act (for example, maturity or provocation or mental distress), some serve somehow to mitigate the gravity of the offence (for example, evidence of remorse). But whatever the proper ambit of such considerations, they are applied at the time of the sentence, so we might argue again that imprisonment cannot change them. Nothing here should alter the sentence once it is passed.
Among the considerations that might alter the effect of the sentence post-hoc, we might consider the types of circumstances that invoke pleas for mercy, clemency or compassion (for example, terminal illness on the part of the prisoner). But again, these are exceptional grounds of early release — detached from a retributive logic — and they are not the same as the sorts of more ‘routine’ early release systems that are our concern here.
Nonetheless, I argue that early release (for non-exceptional reasons) can be justified through retributive theory. My reasoning is this: the calculation of the proportionate, parsimonious and just penalty must take account of the foreseeable but unintended collateral consequences of imprisonment. Although retributivism looks back at the offence, it must (and it does) also look forward at the severity of the penalty. Without looking forward in this sense, I suggest, retibutivism can’t make an accurate judgment of proportionality.
If we accept — as is now commonplace — that the punishment of imprisonment should rest solely in the deprivation of liberty (and not in the harsh conditions of confinement), then surely some account must be taken of other de facto sufferings that arise from imprisonment. To neglect these sufferings, it seems to me, is to violate the principle of nulla poena sine lege (no punishment beyond the law). The precise nature of these sufferings will vary in each and every case but they will commonly include (but are by no means limited to) loss of employment and earnings, loss of (public) housing, damage to personal relationships and the stigma of conviction and of having been imprisoned. Criminological and sociological research suggests that these sufferings are real and enduring, even if none of them is pronounced as a part of the court imposed sanction.
From a retributivist perspective therefore, early release might be justified as a sentence discount that compensates the prisoner for these unintended consequences of imprisonment; it is a mechanism not for violating or traducing proportionality but rather for securing it.
The argument might be extended further to suggest a resettlement duty falls on the state (and on civil society) to take all possible steps to reintegrate the released prisoner to a situation as close as possible to his or her social position and standing before imprisonment, so as to minimize these unintended and illegitimate sufferings. On this logic, the less that a state is prepared to sanction early release, the more it must to do to support and guarantee reintegration.
Notably this duty creates an imperative to support the released prisoner but not to supervise (or seek to constrain or control) him or her. If post-release supervision is considered necessary, this would perhaps need to be justified with respect to public safety. Post-release supervision would then be seen as a preventive measure and not a criminal sanction. In other words, it would fall outside the ambit of retributivism.
However, there is an alternative available to retributivists. This is to pass proportionate and parsimonious sentences that explicitly combine imprisonment and supervision. The determination of how much of that sentence should be served in prison, and of what conditions should apply to the period of supervision in the community, might fall to the sentencing court (of first instance), to an implementation court, or to a parole board. But whatever the mechanism, the decision-makers would, in deciding the timing and conditions of release, have to balance both the duty to resettle and the duty to protect the public… but, crucially, it would have to do so within (and not beyond) the envelope of proportionality fixed in a calculus of duration of liberty deprivation and restriction. And duration would need to be balanced with a parallel calculus of the precise conditions of liberty deprivation and restriction (through imprisonment and supervision). To borrow Ben Crewe’s terms, we would need to think about proportionality not just in relation to the length of the sentence, but also in relation to its ‘depth, weight and tightness’.