Adam Kolber (Brooklyn Law School), "Punishment for the Greater Good
Oxford University Press, 2024
By Adam Kolber
Part I: Introduction
More than ten million people are incarcerated around the world, even though punishment theorists have struggled for centuries to morally justify incarceration and other punishment practices. Efforts so far are incomplete, referencing parts of theories that have yet to be fleshed out. What can philosophers say about the justification of incarceration today?
Retributivists claim that people deserve punishment because of their wrongdoing. Punishment can be morally justified, they say, provided it is proportional to wrongdoing. Consequentialists, by contrast, claim that if punishments like incarceration are justified, they are justified because they lead to good consequences, such as crime prevention and offender rehabilitation, that more than make up for the suffering and other bad consequences they inevitably cause. In my new book, Punishment for the Greater Good (Oxford University Press, June 2024), I argue that a pure consequentialist approach is better than the standard retributivist approach if we seek to address incarceration in the here and now.
As is typical in the punishment theory literature, I discuss consequentialism and retributivism at some level of generality without specifying every feature of either theory. The consequentialism I consider is “pure,” however, in the sense that it disavows the key feature of its main competitor: it denies the intrinsic value of making people suffer (or otherwise treating them harshly) for their misdeeds. To pure consequentialists, suffering in prison is intrinsically bad, though it may be permissible when it leads to greater good overall by reducing harm through deterrence, incapacitation, and rehabilitation. When I use the term “consequentialism,” I will refer to pure consequentialism, unless otherwise clear from context.
I define “standard retributivists” as those who believe: (1) it is intrinsically good when wrongdoers receive the punishment they deserve; (2) offenders deserve such treatment in proportion to the seriousness of their wrongdoing; and (3) we should never purposely or knowingly punish in excess of what people deserve. Standard retributivists also: (4) defend firm deontological constraints; (5) recognize certain fundamental rights owed to criminal defendants; and (6) apply some form of consequentialism in at least some situations where both retributivism and deontology are silent. When I use the term “retributivism,” I refer to standard retributivism, unless otherwise clear from context.
Part II: Problems with Retributivism
Here are three problems with the retributivist approach:
(1) Retributive Proportionality is a Mystery
The idea that wrongdoers deserve proportional punishment lies at the heart of retributivism. Yet retributivists haven’t successfully explained how to determine when a punishment is proportional. Some murderers are punished with life imprisonment, while others, under similar circumstances, are punished with several years of imprisonment. Which duration is appropriate? Some crimes are clearly worse than others, but how much worse? Do we just use our intuition? A growing movement declares that incarceration should be abolished. Intuiting that an offender should spend even a day in jail will fail to satisfy carceral abolitionists. You can’t simply intuit that their views are wrong. Incarceration can’t be justified by gut feelings alone.
Some scholars claim that we just haven’t figured out proportionality yet. While this is never a satisfying response, it is particularly inadequate when we look at punishment from a here-and-now perspective. We can’t incarcerate someone today based on a promise to deliver an adequate theory of punishment sometime in the future.
Consequentialists, by contrast, do not give proportionality a primary role. They seek to punish when the good of doing so exceeds the harm. Since both crime victims and punished perpetrators suffer, consequentialists generally seek to reduce total crime-related suffering. While many people overestimate the ability of incarceration to prevent crime, there is little doubt that it deters an enormous amount of crime and prevents the most dangerous offenders from regularly harming others. When incarceration is not a good tool for public safety, consequentialists oppose its use and seek a better tool. They don’t make people suffer just for the sake of past bad conduct.
(2) Retributive Proportionality is Unappealing
Many think that consequentialism is unappealing because it could lead, under some imaginable circumstances, to the punishment of the innocent or the overpunishment of the guilty. I argue that, even if we could perfectly assess proportionality, retributivist proportionality is also unappealingly counterintuitive, making retributivism not clearly advantageous in an arena where it is often assumed to shine.
Consider two equally blameworthy offenders named Sensitive and Insensitive. They are alike in all pertinent respects except for the amount they suffer in prison. Sensitive suffers tremendously. He is usually distressed and anxious. Insensitive suffers too but manages to cope and make good friends. If these equally blameworthy offenders spend the same three years in prison, I claim that they have not been punished equally in any sense that matters from a moral perspective. Moreover, if Insensitive’s sentence was proportional, then retributivists need to explain what justifies the additional suffering we knowingly impose on Sensitive.
We could try to punish in ways that take sensitivities into account. But doing so leads to counterintuitive results as well. Suppose Sensitive suffers so much because, prior to prison, he lived a life of luxury (with wealth obtained non-culpably). Few would welcome punishing Sensitive for a shorter duration (or in better conditions) to accommodate his wealth-induced sensitivity. Nevertheless, it’s hard to see how retributivists can knowingly make Sensitive suffer more than Insensitive when they are equally blameworthy.
We reach similar results even if we speak in terms of deprivations of liberty rather than suffering. If Jeff Bezos committed a crime and was as blameworthy as some ordinary offender, putting them both in prison for the same duration would deprive Bezos of far more liberty. For the period of his incarceration, he would lose access to his many homes and his mega-yacht. To deprive him of the same liberty as an ordinary offender, we would have to give Bezos a shorter sentence or much better prison conditions to make his punishment equal in magnitude to that of the ordinary prisoner. This is a conclusion few retributivists will find appealing.
Consequentialists have to take suffering into account too. But they’re not specifically committed to proportional punishment and so needn’t reach the particular counterintuitive results that retributivists face here. At the same time, many people do have the strong intuition that punishment should be proportional. Consequentialists may find that imposing punishments viewed as proportional often has the good effect of meeting people’s expectations about just punishment. Satisfying those expectations may itself lead to good consequences, including increased compliance with laws that are viewed as fair and just. But if consequentialists do punish a wealthy person like Bezos with the same carceral conditions for the same period of time as an ordinary person, they recognize that the two are not receiving equal impositions of suffering or liberty deprivation.
Many think consequentialism is highly counterintuitive because, in principle, it might allow punishment of an innocent person contrary to a firm deontological principle that we may never punish in excess of desert. In chapter 4 of the book, I argue that most reasonable retributivists will also punish the innocent under sufficiently catastrophic conditions. The dispute largely comes down to setting the threshold at which punishment of the innocent is permissible, and it’s hardly obvious that consequentialists have the worse end of this perennial dispute.
(3) Retributive Punishment is Too Risky
Third, most retributivists believe that in order to punish, we must believe someone committed all the elements of a crime beyond a reasonable doubt. Some say this means we shouldn’t punish unless we are 95% or even 99% confident that an accused person is guilty. The problem is that, in order to actually punish someone, retributivists must believe a long list of propositions that they are unlikely to collectively hold with high confidence. For example, they must believe people have the kind of free will that can make them seriously morally responsible. They must believe it is morally appropriate to make people suffer (or be deprived) because of their past wrongdoing, even though some think this view is barbaric.
I discuss many additional propositions retributivists must believe. Even if they can hold some of them with high confidence, they cannot hold the conjunction of all of them with sufficient confidence to actually punish, at least not in a manner consistent with the values that seem to underlie their commitment to the beyond-a-reasonable doubt standard. Why be so demanding in your confidence in factual guilt when you’re considerably less sure that the person has free will and that it is morally appropriate to respond to wrongdoing with suffering or liberty deprivation?
Consequentialists needn’t have the same worries. They seek to punish bad behavior regardless of whether the offender has metaphysical free will, and they don’t view deserved suffering as intrinsically valuable. Granted, they should be morally uncertain about a variety of matters, but they are not paralyzed in the same way that retributivists are. Consequentialists can balance two kinds of risks: uncertainty about factual or moral matters might lead us to overpunish, but it might alternatively lead us to underpunish. Because consequentialists are not committed to the beyond-a-reasonable-doubt principle in the way that most retributivists are, consequentialists can make reasonable punishment decisions even in the face of substantial uncertainty.
Part III: Consequentialism More Generally
There is, of course, a long history of challenges to consequentialism. But this book needn’t address them all. The reason is that standard retributivists, as I describe them, are consequentialists as well. They don’t often specify their particular brand of consequentialism in great detail, but as a sociological matter, most retributivists I meet adopt some form of consequentialism when no deontological constraint (such as the prohibition against knowingly punishing in excess of desert) is at issue.
While it might seem unfair to saddle standard retributivism with whatever downsides consequentialism may have, there is a fundamental reason why punishment theorists who seek to address real-world matters tend to adopt aspects of consequentialism. Namely, to the extent they attempt to justify punishment practices in the here and now, retributivism without consequentialism tells us far too little to justify a punishment practice such as incarceration. While standard retributivists observe firm deontological constraints, they almost never provide a complete list of these constraints. Even if they did, the challenges of interpreting them would be enormous. Moreover, in the real world, we must assess the probability that a person is being used solely as a means to an end or the probability that a punishment is disproportionally excessive. Yet few offer detailed description of which deontological constraints we must observe and how to weight them in a world of uncertainty.
The affirmative case for consequentialism arises, in large part, from its relative completeness. Armed with an axiology—a way of valuing what is good and bad—pure consequentialism is quite complete. While debates about the appropriate consequentialist axiology rage on, standard retributivists face nearly-identical challenges (and not just because they are also consequentialists). In order to measure wrongdoing, for example, most retributivists consider the amount of harm an offender caused. Such measurements require them to decide whether to treat harms as bad experiences, dissatisfied preferences, or something else entirely. Since I conduct a comparative analysis of pure consequentialism and standard retributivism, I spend the better part of a chapter arguing that, if anything, pure consequentialism raises simpler axiological questions than standard retributivism because pure consequentialists needn’t give special weight to the complicated notion of desert, nor wrestle with its associated mysteries (for more on such mysteries, see here).
Once consequentialists have an axiology and their best probabilistic assessment of relevant empirical facts, they can tell us how to behave in a wide variety of circumstances. Importantly, they can say quite a bit about whether an instance of incarceration or a practice of incarceration is likely better or worse from a moral perspective than some alternative arrangement. The empirical issues are extraordinarily complicated, but at least we know how to address them. (If a patient must choose between two forms of cancer treatment and the scientific evidence is conflicting and complicated, we can still do our best to make a decision using relatively well-agreed upon methods of analysis.) With numerous choices, we may not know which option will lead to the greatest good, but we can often make choices for the greater good. It’s hard to ask for more in the here and now.
Imagine a car race with two competitors. If one car is missing too many parts to start, even its shabby competitor is superior. If standard retributivism is too incomplete to yield verdicts about incarceration, then pure consequentialism is superior in the here and now, even if it has its own blemishes. Hence, my claim in the book is one of superiority, not adequacy. I claim that pure consequentialism is superior to standard retributivism, not that pure consequentialism is necessarily an adequate theory to adopt in the here and now. An adequacy claim would require a much deeper defense of consequentialism than I provide in a relatively short book and would require us to look at many other approaches to punishment than the two I focus on. Still, to the extent that I address popular forms of consequentialism and retributivism, if I succeed in arguing that pure consequentialism is superior to standard retributivism, I may have provided you with at least some reason to increase your confidence in pure consequentialism’s adequacy as an approach to punishment in the here and now.
Punishment for the Greater Good is available here (discount code: ALAUTHC4) or through request from your library here. For review copies, click here or contact the author at adam.kolber@brooklaw.edu.
