Four rationales for punishment (2024)

As any law student who’s studied criminal law will tell you, there are four traditional rationales for punishment: retribution (giving someone their just deserts), deterrence (preventing harm in the future), rehabilitation (transforming someone into a better person through punishment), and incapacitation (keeping a person away from society so they don’t hurt others). In discussions of criminal justice policy, these rationales are very often mixed up and conflated, which produces incoherent discussions.

Rehabilitation has long fallen out of favor as a rationale - I don’t think you’d find many system actors who would say the reason to send someone to jail or prison is that it will make them better than they were. (A lot of the blame should go to this very flawed article in the 1970s, whose claim that rehabilitation doesn’t work gave permission to gut programs for millions of people; the author later retracted his claims). With the destruction of a huge number of learning, education, and support opportunities that were previously available in prison, it’s become pretty clear that prisons and jails are places where people are going to suffer and that’s about it.

The second thing to know is that ‘incapacitation’ is a misnomer, as people suffer a substantial amount of crime in prison at the hands of guards and other prisoners, so sending someone to prison doesn’t reduce crime in the way folks imagine, unless you zero-out the moral value of a person who is in prison and therefore don’t care if they become crime victims inside. Moreover, 95% of people who go into prison come out at some point, usually within a few years, and face immense challenges in rebuilding relationships and a life after the isolation and brutality of prison. Even if we wanted to ‘disregard’ crime in prison as a social concern, incapacitation in prison as we know it in America is not a solution for a safer society.

Getting to the two that people focus on the most – retribution and deterrence – the first thing to know is that these two aims tend to work at cross purposes. This is not what most people assume. The common sense understanding, if people consider it at all, is that the punishment suited for retribution – making the person suffer for causing harm – is also well suited to deterring or reducing future harm. But a mountain of evidence tells us that the opposite is true. If we want less harm in the future, we have to invest in other solutions. Indeed, those solutions are happening already: investing in health, poverty alleviation, and housing is what works to reduce crime. A true commitment to deterrence requires focusing a much higher percentage of social resources on these interventions (see Measure J for an example).

I have long been curious what people would say in a poll if asked to choose between retribution and deterrence. Could anyone rationally say that they will tolerate future harms, being certain they will increase if retribution is enacted, as the acceptable cost of ensuring retribution? Or is their support for retribution really based on the belief that punishment is a good deterrence? In other words, is it about a moral commitment to retribution no matter the harm it causes, or is this an empirical question about what works to deter? And if it’s empirical, what evidence would persuade people? Or what narratives do they need to embrace before they can accept the fact that a prison sentence is going to harm more than it helps?

To be clear, all of this is starting from a place of wanting to increase safety and reduce harm in the future. It’s not about justice or being nicer to people who are in prison, or trading off freedom and safety. If we care only about safety, we have to think about what increases safety, and the evidence is clear. But for many people, that’s as hard to accept as a rabbit being pulled out of a hat.