Consequentialist Theories of Punishment (Part Two)

(Part One)

This is the second part in a brief series of posts looking at consequentialist justifications for punishment. The series works off some of the material contained in Michael Zimmerman’s book The Immorality of Punishment. As explained in part one, Zimmerman uses this book to defend two theses about the nature of legal punishment. The first — the partial unjustifiability thesis — holds that there is at least one serious moral reason (possibly more) that counts against punitive acts, practices and institutions. The second — the overall unjustifiability thesis — holds that the weight of moral reasons is highly likely to count against punitive acts, practices and institutions.

What I want to do here is see how those theses pan out when it comes to consequentialist theories of punishment. In other words, I want to see if there are serious moral reasons that count against such justifications of punishment. As we saw the last day, consequentialist theories of punishment typically justify punishment by pointing to its harm-reducing effects. This is thought to be achievable in one of three ways: (i) through punitive rehabilitation; (ii) through punitive incapacitation; and (iii) through punitive deterrence. In order for these practices to count as “punishment”, they must involve the intentional imposition of fitting harm.

And therein lies the problem. Key to the success of any one of these three theories is its ability to answer (in the affirmative) a contrastive question: Is a punitive form of rehabilitation/incapacitation/deterrence more likely to be successful than a non-punitive form? The question is necessary since non-punitive forms of rehabilitation/incapacitation/deterrence are possible and, if more likely to be successful, will be more morally justifiable than punitive ones. This is for simple moral reasons: it’s better not to harm someone than to harm them.

In this post, we will see how the rehabilitative theory of punishment fares when it comes face-to-face with the contrastive question. The analysis will be divided into two sections. First, we will look at educative approaches to rehabilitation and see whether they can justify punishment. Second, we will look at non-educative approaches to rehabilitation.


1. Rehabilitation via Moral Education
One way in which punishment is thought to achieve the goal of harm-reduction is through the moral education of the punishee. The reasoning is as follows. By engaging in conduct that causes harm (i.e. by engaging in criminal activity of some kind) an offender has signalled a lack of moral knowledge. Only someone who was not aware of the decisive moral reasons that counted against the harmful conduct would have engaged in it. Punishment can be used to make them aware of these reasons. Hence, punishment can be used to morally educate an offender, which in turn can reduce instances of harmful conduct.

To be sure, this reasoning is far from flawless. As Zimmerman notes, it only works if two preliminary conditions are met.

The Unawareness Condition: The punishee must be genuinely unaware of the moral reasons that counted against his conduct prior to punishment. If he learned them immediately after committing the offence and was deeply repentant, punishment would not be needed.

The Receptivity and Responsiveness Condition: The punishee must be both receptive and responsive to the educative effect of punishment. In other words, they must be capable of recognising and responding to moral reasons. Arguably, this condition is not met in the case of many offenders. For example, psychopaths are thought to be peculiarly unreceptive to moral reasons.

Zimmerman argues that it will be difficult to tell whether those conditions are actually met in many cases. But let’s assume that the two conditions can be met and identified and see what follows. Presumably, the rehabilitative defender of punishment would offer something like the following argument:

• (1) If punitive acts are more likely to achieve the end of moral education than alternative non-punitive acts, they can be (at least partly) justified in terms of their harm-reducing effect. 

• (2) Punitive acts are more likely to achieve the end of moral education than alternative non-punitive acts. 

• (3) Therefore, punitive acts can be (at least partly) justified in terms of their harm-reducing effect.

There are certain issues with the way in which premise (1) is formulated, but we’ll talk about those in a minute. The bigger problem, at least as Zimmerman sees it, is that premise (2) is false. Punitive acts are not more likely to achieve the end of moral education than alternative acts. At an empirical level, although some punishees may learn an important lesson from their incarceration (or other form of punishment), the high rates of recidivism suggest that this is not a general truth.

Furthermore, even in principle, it’s not clear exactly how intentionally harming people will make them learn their moral lessons. A common suggestion is that whenever the harm imposed matches the harm involved in the offence, the punishee will learn, not just in the abstract, but in their “gut” why their action was morally wrong. This has an air of plausibility about it, but problems remain. For starters, it would need to be demonstrated that this really is the case. In addition, many forms of punishment do not attempt to replicate perfectly the harm imposed by the offence, nor is it clear that it is desirable that they do so. For example, how do we replicate the harm imposed by a serial rapist? Would we wish to do so? Admittedly, there are some who think we should, but it seems like doing so would create a disturbing level of cruelty in the heart of the criminal justice system.

For these reasons, the argument from moral education seems to fail. But there is an escape for its defender. This is to argue that the version of the argument that I have presented misconstrues their claim. As I have phrased it, premise (1) is all about justifying individual punitive acts by reference to their educative effect. But this is not the argument that is being made. A proponent of educative rehabilitation could concede that individual punitive acts are unlikely to have the desired effect, and argue instead that the cumulative effect of a sustained and systematic series of such acts could have the desired effect. This would lead to the reformulation of premise (1):

• (1*) If a punitive act or sequence of punitive acts is more likely to achieve the end of moral education than alternative non-punitive acts (or sequences thereof), they can be (at least partly) justified in terms of their harm-reducing effect.

The problem is that even if the premise is reformulated in this manner, it is no more likely that the ensuing argument will be successful. If an offender is repeatedly punished, it seems more likely that they will grow bitter and resentful of the system that punishes them, or, perhaps worse, more battle-hardened and less worried about being punished for engaging in harmful conduct. Indeed, this seems to be borne out by the current status of system (and the level of recidivism involved (though one would like to see some figures on this just to be sure).


2. Non-Educative Rehabilitation
The argument we looked at spoke to the role that punishment could play in the moral education of the punishee, and to the role that this could play in rehabilitating character and reducing future harmful conduct. A more cynical rehabilitative theory might argue that punishment could be effective in changing behavioural dispositions, even if it did nothing to improve moral character. For example, it might argue for a form of punishment based on aversion therapy. In aversion therapy, a subject is repeatedly presented with a stimulus that is paired with an unpleasant or noxious event. The idea is that over time they will learn to associate the stimulus with the unpleasant event, and thereby become averse to it. It is often used in cases of addiction. This type of therapy would count as punishment since the unpleasant event would be intentionally harmful.

Could a pro-punishment argument, using this kind of rehabilitative practice as its basis, be made? Well, sure it could. We’d just have to adjust the premises from the moral education argument and make them speak about behavioural alteration instead. But we’d still run into the same problem, or so at least Zimmerman argues. The notion that we could successfully rehabilitate people using something like aversion therapy is purest fantasy, as best he can tell.

But this is just a “put-up or shut-up” argument. It merely challenges defenders of rehabilitation to show that their methods can be successful. There is always the danger that they actually do put up and thereby shut you up. So, unsurprisingly, Zimmerman argues that such methods are always likely to be unsuccessful. He does on the grounds that a rehabilitative punishment, like aversion therapy, relies on the mistaken assumption that criminal behaviour is attributable to some individual pathology that can be fixed or cured. This assumption is mistaken. It is far more plausible to think that criminal behaviour is attributable to a complex set of individual, genetic and environmental causes. Thus, a therapeutic intervention that only targets individual pathologies is always likely to fail.

I think this is difficult argument to evaluate. I would certainly concede to Zimmerman that the causal basis of behaviour is a complex set of individual and environmental factors, but I’m not sure it follows that rehabilitative interventions targetting the individual are thus unlikely to be successful. It seems plausible to me that even if there is a complex set of causal conditions behind any particular criminal act, preventative interventions could be targetted at either the individual and environmental level and still be successful. For example, heart disease might be attributable to a complex set of individual factors (no exercise, high fat and salt diet) and environmental triggers (sedentary pleasures and readily available high fat, high salt food). But even so, it’s possible that interventions targetted at one level or the other might play some role in reducing heart disease. The same could be true of interventions targetting criminal behaviour.

Where Zimmerman seems to be on firmer ground is in his contrastive evaluation of rehabilitative punishment. Even if something like aversion therapy could be successful, it would still need to be shown to be more effective than non-punitive forms of rehabilitative therapy. This seems like a far more difficult claim to prove.

To sum up, rehabilitative theories of punishment are unlikely to provide sufficient justification for punitive acts, practices or institutions. This is for two main reasons. First, it seems implausible, given what we currently know about the effects of punishment, to think that it can be effective in changing moral character or behavioural dispositions in the manner required. Second, it seems implausible to think that punitive acts and practices are more likely to have the required rehabilitative effect than non-punitive ones. Indeed, this conclusion is more in keeping with typical discussions of rehabilitation, which often paint it as an alternative to punishment, not as an instance of it.

That’s it for now. We’ll look at the other consequentialist theories of punishment the next day.