It is common knowledge that the United States incarcerates more people per capita than any other country – indeed, five times the rate of most countries. My state of Maryland, which has a below-average rate compared with other states, incarcerates 531 people per 100,000 of the population. The United Kingdom – the country next highest on the list – incarcerates a quarter as many: 133 per 100,000. Almost as well-known is that prison sentences in the US are longer and harsher than those in any country with which we like to compare ourselves. The US dispenses life sentences at a rate of 50 per 100,000, about the same as the entire incarceration rate for countries like Finland, Sweden and Denmark. One in seven prisoners in the US is serving a life sentence: more than 200,000 people, a greater number than were incarcerated for all crimes in 1970.
The number of people in the US serving life without parole (LWOP) sentences has increased 66 per cent since 2003. Germany outlawed LWOP in 1977, and in 2013 the European Court of Human Rights decided that Article 3 of the European Convention of Human Rights prohibited LWOP as a form of ‘inhuman or degrading treatment or punishment’. Because of the barriers to getting parole, many life sentences are in fact LWOP sentences. Moreover, ‘virtual life sentences’ – those greater than 50 years – effectively condemn almost all those serving them to a life in prison. In 2016, more than 44,000 people in the US were serving virtual life sentences.
Can anything justify these harsh sentences, in particular life or other very long sentences?
The answer is no. It’s easy to show that the public safety arguments for these sentences are fallacious, which leaves the conclusion that only retribution could possibly warrant them. But when understood in any reasonable way, retribution also fails to justify them.
Justifications for punishment are typically divided into the ‘forward-looking’ (consequentialist or utilitarian) justifications on the one hand, and the ‘backward-looking’ (retributive or deontological) justifications on the other. Forward-looking justifications mainly concern public safety – general and specific deterrence, incapacitation of dangerous offenders – and rehabilitation. Consider first general deterrence. The idea is that we punish people in order to send a message to other would-be law-breakers that bad things will happen to them if they commit crimes. (Oddly, the aims of general deterrence could be satisfied if no one were actually punished, as long as the public believed they were punished.) Probably no one denies that general deterrence has a legitimate role to play in the criminal justice system. If nobody were ever punished for their crimes, there would be more of them. Nevertheless, the question is what deterrents actually reduce the likelihood of crime, why, and by how much.
Much recent work shows that it’s not the severity of the punishment that deters but its likelihood. There are several reasons why. First, deterrability varies. People suffering from severe mental illness or those acting impulsively or in the heat of passion may not be at all deterrable. Different people have different attitudes toward risk. Few perform cost-benefit analyses when contemplating a crime; those who do may do so poorly. Second, people are often ignorant of the penalties attaching to different crimes, and tend to underestimate their severity. Perhaps most important are the many steps between crime and punishment – being caught, accused, tried, convicted, and sentenced – which greatly reduce the likelihood of punishment. All these factors give us reason to doubt the deterrent effect of life sentences.
What about incapacitation and specific deterrence? These terms can easily be conflated. The most literal meaning of specific deterrence is that the unpleasant experience of incarceration will somehow prevent offenders from re-engaging in criminal activity, and in the best case will lead them to give it up. That might be because the person is morally rehabilitated – no longer inclined to antisocial behaviour – or simply that they become sufficiently averse to further punishment. (So it’s not always easy to distinguish specific deterrence from rehabilitation.) Incapacitation means that, while locked up, people are less likely to commit crimes, or at least less likely to harm members of the public.
It’s not the severity of the punishment that deters but its likelihood
But none of these purposes is served by life sentences. The central reason is that most offenders, including those who engage in violence, age out of crime before middle age. The journalist Dana Goldstein writes: ‘Homicide and drug-arrest rates peak at age 19, according to the Bureau of Justice Statistics, while arrest rates for forcible rape peak at 18 … For most of the crimes the FBI tracks, more than half of all offenders will be arrested by the time they are 30.’ And, she continues, criminal careers are short-lived: ‘for the eight serious crimes tracked by the FBI … five to 10 years is the typical duration that adults commit these crimes, as measured by arrests.’ As the neuroscientist Robert Sapolsky put it in an interview: ‘The greatest crime-fighting event on Earth is the 30th birthday.’
There are at least two reasons why criminal behaviour skews young. One is its physical requirements, which are more demanding as people age. At least as important is that people change. Someone in their teens or 20s is not the same in his 40s. The case is especially strong for young offenders, as has been extensively argued in recent years. That includes not only literal adolescents but those as old as 25. Those aged 18 to 25 – although around 10 per cent of the US population – comprise 25 per cent of arrests and 19 per cent of admissions to adult prisons. The young brain is not fully developed, with less impulse control and greater dependence on peer approval than adults have. According to a 2018 report by the Sentencing Project more than 11,000 prisoners serving life sentences – more than 5 per cent of all those serving life – committed their crimes before they were 18. And even those who commit crimes after 25 are capable of radical transformation, as anyone who has spent time, as I have, with those currently or formerly serving long sentences can attest.
Of course, some offenders may have to stay locked up for life because they remain dangerous. How to assess and manage their sentences is an important question I will ignore here. But such cases account for a very small fraction of the nearly quarter of a million people incarcerated for life in the US today.
Rehabilitation is the other forward-looking justification for punishment I must mention – but only to dismiss it. That’s because the idea that one would permanently incarcerate someone in order to rehabilitate them is impossible to get one’s mind around. Perhaps a life sentence is necessary to keep others safe; perhaps it’s justified by a person’s desert. But to lock up someone for 50 years to make them fit for human society? The idea is paradoxical, to say the least.
That leaves retribution as the only plausible justification for life sentences. This is no surprise. It’s what rises quickly to the surface when supporters express their views, including those who realise that the deterrence arguments fall short. All but the most saintly among us are likely to want to lash out when we or someone we love is the victim of attack. Forget about deterrence; punishment is what wrongdoers deserve, and the world is out of whack until they get it! Evolutionary psychologists have sought to give the urge for revenge a biological basis.
But retribution needs some philosophical credibility if it is to justify, rather than simply explain, our punitive practices. Over the centuries, many philosophers have obliged, most notably Immanuel Kant. Retributivism is the name of the moral theory that attempts to justify the practice of retribution. It comes in several varieties, but the basic idea is that culpable wrongdoers ought, morally, to be punished proportionately to their offences and their culpability. (A person who is only partially culpable, ie has some kind of mitigating excuse, should be punished less harshly than one who is fully culpable.) But this leaves unresolved at least two issues that must be addressed for retributivism to be a clear and coherent theory.
These issues are illustrated by Kant’s unforgettable assertion in The Metaphysics of Morals (1780) that if a person
has committed murder he must die. Here there is no substitute that will satisfy justice. There is no similarity between life, however wretched it may be, and death. Even if a civil society were to be dissolved by the consent of all its members … the last murderer remaining in prison would first have to be executed, so that each has done to him what his deeds deserve.
This passage refers obliquely to two unresolved questions.
The first is what constitutes proportional punishment. Kant seems to express the traditional understanding of retributivism as lex talionis, the law of retaliation – an eye for an eye. The punishment should equal the crime. Thus, the murderer must die: ‘There is no likeness or proportion between life, however painful, and death.’
Most offenders, including those who engage in violence, age out of crime before middle age
But the implications of lex talionis are repellent. Should we torture the torturer, and rape the rapist? If the answer is no, the question is why. Some would say no one deserves to be tortured or raped. But a retributivist may be more inclined to the view that that the rapist may deserve to be raped; yet we should refrain because it would debase, demean or degrade the person who would carry out this act or the society that permits it. Torturing the torturer and raping the rapist are uncivilised, beneath us. And state-sponsored rape and torture would water down a message that should never be diluted: rape and torture are always impermissible.
Either response leads to an alternative view of retributivism: not lex talionis but proportional retributivism. Punishments should be in proportion to crimes, which means at the very least they should be not too harsh and not too lenient. A life sentence for shoplifting is too harsh; a week in jail for rape is too lenient. On this view, we need to construct a scale of crimes and punishments in which the worst crimes get the worst punishments, the next worst get the next worst, etc. Of course, this is an ideal, perhaps even a fiction, that can be achieved only very roughly. Nevertheless, if we reject lex talionis we must implicitly accept some kind of matching of crimes and punishments.
That leaves open just what punishments are appropriate to what crimes, and which cross the line into forbidden territory. Raping the rapist may cross the line, but what about the death penalty? Life, with or without parole? Are some or all of these also beyond the pale, either because they exceed an offender’s desert, or because they degrade those who literally or politically must carry them out?
Retributivists may well deny that either of these conditions is met in the case of life imprisonment (or the death penalty). They may believe that killers deserve to be killed, and also that killing them lacks the icky features that make punishment by torture or rape unacceptable. Life imprisonment as a punishment for first-degree murder likely seems reasonable to many or even most Americans. By contrast, Norway caps the maximum sentence for any crime at 21 years (though in rare circumstances this can be extended by five-year increments). There is probably no single objective standard for the appropriate sentence. One might argue directly for the conclusion that life sentences are too long, full stop, but it is easy to see how this strategy might appear to beg the question.
Nevertheless, when we reject lex talionis we see at the very least that a commitment to retributivism does not entail endorsing the death penalty or life imprisonment for serious crimes. It entails a belief that the wrongdoer ought to suffer punishment, but not how much. And that’s not the end of the story. Another central question must be resolved to determine retributivism’s reach.
That question, also revealed in the passage from Kant, is how to understand the ‘ought’ in the claim that culpable wrongdoers ought to be punished because they deserve it. Kant’s view is unambiguous: the wrongdoer must be punished no matter what. ‘Even if a civil society resolved to dissolve itself … the last murderer lying in prison ought to be executed before the resolution was carried out.’ There are no mitigating circumstances or countervailing factors.
This is detached from reality. Retributivists should instead say that culpable wrongdoing creates a presumption in favour of punishment, perhaps even a strong presumption, but one that can be overridden. So stated, that still leaves a lot open. How strong is the presumption? What other factors are relevant, and how powerful a counterweight do they provide to the moral force of retributivism? In what follows, I’ll consider four factors that help determine the force of retributivism’s ought – and thus, how harsh a sentence should be.
One reason life sentences are unjust is that they disproportionately affect people of colour, who comprise more than two-thirds of those serving such sentences in the US. In particular, Black people make up almost half of lifers, despite comprising only 13 per cent of the US population. Of course, this in itself demonstrates bias only if these sentences are out of proportion to the involvement of people of colour in criminal activities. They are. As most people know by now, Black people are treated differently than whites ‘at every stage of the criminal justice system’, as the American Civil Liberties Union put it in 2014. That includes the death penalty and life sentences. One reason is racial bias, implicit and explicit. Another is that people of colour tend to be poorer, and poorer people are more likely to commit crimes than rich ones; they are also less likely to receive adequate legal representation.
The question is how to respond to these systemic inequities. Suppose a certain crime carries a possible life sentence, but that Black people receive it disproportionately, with similarly situated white people routinely receiving milder sentences. One might think the appropriate response is to ‘level up’ – give white people life sentences too – rather than to ‘level down’ by reducing the sentences of Black people. Which tack to take might depend partly on whether one thinks life sentences are ever a just and appropriate punishment. Proponents of levelling up can’t simply assume the answer is yes, when that is the very question at issue.
Here is one reason to think levelling down is appropriate. While the US criminal justice system treats Black and brown people worse than it treats whites, it seems plausible that the treatment of white people adheres more closely to our society’s views of justice, since white people are thought to represent the ‘normal’ or the default. People of colour, then, are subject to a surplus penalty rather than white people being let off easy. If that is so, reducing the penalties Black people pay rather than increasing those of whites would bring us closer to justice.
You cannot reasonably believe that someone should spend their whole life in prison unless you believe they will always present a significant threat to public safety, or that they deserve such a sentence. I have already argued that the overwhelming majority of the more than 200,000 people currently serving life sentences in the US do not present a lifelong threat to public safety. Whether they ought to spend their whole lives in prison depends, then, solely on whether they deserve it. They can deserve it only if they are fully responsible for their criminal acts. In the law, responsibility is best understood by what it excludes. One who is not responsible (liable) for their criminal act has a partial or complete excuse. A complete excuse, making one wholly nonculpable, has required, since at least the 19th century, that one not know what one is doing, not know it is wrong, or not be able to control one’s actions (be under the influence of an irresistible impulse). Under these definitions, most people serving life sentences are probably responsible for their acts.
Growing up in environments of deprivation increases the likelihood that a person will go on to commit crimes
But a broader understanding of responsibility is essential. For one thing, the insanity defence is a complete excuse that renders a person legally innocent (which doesn’t mean they walk free; they are likely to be institutionalised in a psychiatric facility). But there are many more whose responsibility is diminished, even though not entirely absent. The Death Penalty Information Center in its 2021 report found that:
[A]ll but one of the 11 people executed in 2021 had one or more significant impairments, including: evidence of mental illness; brain injury, developmental brain damage, or an IQ in the intellectually disabled range; or chronic serious childhood trauma, neglect, and/or abuse.
There is every reason to believe that those sentenced to life also have extremely high rates of impairments. Many have experienced abuse, neglect, and trauma. By some reports, almost 40 per cent of prisoners in state and federal facilities suffer from some form of diagnosed mental illness.
But actual mental illness is not the only factor relevant to determining culpability. It’s clear that growing up in environments with certain kinds of deprivation – high poverty, neglect or abuse, poor schools, prevalence of guns and drugs, non-intact families, inadequate access to decent employment – greatly increases the likelihood that a person will go on to commit crimes. For example, if you happen to grow up in the city of Baltimore, the probability that you will commit a violent crime is more than five times greater than for residents of the US as a whole. This comparison significantly understates the effect, since the figures for the US include many places with high crime rates, and this obscures the contrast with safer places. Baltimore’s rate of violent crime is more than 30 times that of Frederick, Maryland, a small city about an hour west of Baltimore. We cannot ignore such disparities in judging lawbreakers’ responsibility.
Such talk makes people uneasy. It seems to suggest a hard determinism that is incompatible with holding people responsible. How can we avoid sliding down the slippery slope to a place where no one acts freely, and no one is culpable for what they do?
We avoid it by compromising between two powerful, intuitive perspectives, neither of which we can abandon. One is that, practically and humanly, we must hold people responsible for their actions most of the time. We cannot think of ourselves or, usually, others as beings whose behaviour is the inevitable outcome of everything that happened to them before. This is the lesson of (or perhaps the reason for) compatibilism, probably the dominant view of the free will problem among contemporary moral philosophers and criminal law theorists.
A commitment to retributivism does not entail endorsing the death penalty or life imprisonment for serious crimes
Compatibilism says that determinism (universal causation) and free will or moral responsibility are compatible – that if one’s actions are caused in the right way or by the right things (and different theories will offer different accounts of what the right way or the right things are), then we are morally responsible for them; our wills are, for the most part, as free as they need to be. After all, the compatibilist rightly points out, if our actions were not caused, they would be uncaused, ie random, and that would hardly make them free. The view is satisfying as long as you don’t push on it, which we can often avoid doing. But the principle and its upshots clash with another, equally indispensable principle: that many of the factors that contribute to a person’s committing crimes severely limit their freedom.
The way to square this circle is to punish, but to punish less harshly than if a more robust conception of free will were in play. From a purely logical point of view, this solution may appear inadequate. It wants to have it both ways, and distorts both the principles under consideration. But the compromise, I believe, does as much justice as we can hope to achieve in this world.
Retributivism presupposes moral responsibility. To the extent that a person’s responsibility is compromised, their punishment should also be reduced.
Another reason to abolish life sentences rests on the strangeness of continuing to punish a person who committed a crime years earlier but has changed radically since that time. This is the situation of many people serving very long sentences. They may have murdered someone (or done something much less serious: there are people serving life sentences for nonviolent crimes) as teenagers, and remain incarcerated 30 or 40 or 50 years later. Leaving aside the morality of continued punishment, we may question its rationality. What’s the point of continuing to punish a person who recognises the wrongness of what they have done, who no longer identifies with it, and who bears little resemblance to the person he was years earlier? It is tempting to say that he is no longer the same person.
That assertion might seem to land us in dense philosophical thickets. But is it really so complicated? Certainly, in describing people, we often talk this way. As Jennifer Lackey argues, because we take mental states as relevant to punishment, we should also take two stages of the same person ‘with radically different attitudes toward his crime, as deserving of different punishments’.
The three factors I’ve talked about so far – bias, responsibility, identity – affect retributivism internally: to the extent that each of them holds, they weaken the ‘retributive force’ of retributivism by reducing a person’s desert. The following considerations are, by contrast, forward-looking. They focus on the benefits of less-harsh sentences.
There are, first, the incarcerated people whose sentences would be shortened and who would sooner or later join the outside world. That they benefit is not quite as obvious as it might seem, since formerly incarcerated people can find it very difficult to secure decent lives on the outside in the absence of education, training and money, which they so often lack. Nonetheless, even without such advantages, they will almost always be better off out of prison than in it. That is why most of them fervently want to get out.
Of course, we cannot ignore the interests of victims of crime. Some victims want offenders to receive the maximum penalties possible, and may otherwise feel unhappy or insecure. Leaving aside how to weigh the feelings of victims in sentencing and re-sentencing decisions, we also know that not all of them believe that, because one life has been lost, others must be as well.
Another group are the families and communities of those who have committed crimes, who are also victims. The harms of having their members – especially men and, disproportionately, young Black men – disappear from the community for years at a time are incalculable, even taking account of the benefits of having violent people removed.
How would society more generally be affected by the abolition of life sentences? The waste of human lives condemned to prison for life, or even for decades, is tragic as well as irrational, and can be justified only by some powerful offsetting benefits. As we have seen, there is scant evidence that long sentences have either general or specific deterrent value. Incarceration is very expensive, and becomes more so as prisoners age.
Also relevant are the benefits of adopting a less punitive approach. If less-harsh policies result from democratic decision-making, that includes ‘us’ as a society, in addition to individual agents of reintegration. To embrace this approach is to express a certain optimism about the possibilities of good and redemption in human beings.
The philosopher Ryan Preston-Roedder has explored this terrain, arguing that faith in humanity makes the world better both for those who have faith and for those in whom they have faith. Take first the idea that it is good for those in whom one has faith. Goethe proclaimed that ‘if we treat people as if they were what they ought to be, we help them become what they are capable of becoming.’ It sounds nice, of course, but is it too nice to be true? No. There is good social scientific evidence showing, for example, that people tend to internalize others’ view of them, and that when people have certain expectations of others’ behavior they may send subtle signals to which those others then conform. For these and other reasons, Preston-Roedder writes that ‘having faith in people’s decency tends to encourage them to act rightly.’ It’s not foolproof; we can make mistakes, and we can sometimes be taken in by people. Blind trust is not advisable. But an attitude that does not reduce people to their worst acts, as the civil rights lawyer Bryan Stevenson puts it, and that does not permanently label them as criminals is more likely to succeed.
Preston-Roedder argues that faith in humanity is also good for one’s own well-being. That alone is not sufficient to recommend it. But we can count this trait as a virtue if we agree that on balance it benefits those who possess it as well as others. A world in which we do not give up on people who have done terrible things, and aim to facilitate their journey to a different place, is a better world than the alternative.