So the notion of individual deterrence seems to be of little value in justifying our penal practices. But there is another, perhaps more promising category of deterrent effect: general deterrence. This is the idea that offenders are punished, not to deter the offenders themselves, but pour encourager les autres.3 General deterrence theory is often cited to justify punishments, including those imposed on particular offenders. One faintly ludicrous example is a 1983 case4 where the Court of Appeal said that a particular sentence would ‘indicate to other people who might be minded to set fire to armchairs in the middle of a domestic row that if they do, they were likely to go to prison for as long as two years’.
Now, there can be little doubt that the existence of a system of punishment has some general deterrent effect. When, during the Second World War, the German occupiers deported the entire Danish police force for several months, recorded rates of theft and robbery (though not of sexual offences) rose spectacularly (Christiansen, 1975; Beyleveld, 1980: 159). And if, for instance, on-the-spot execution were to be introduced for parking on a double yellow line, there might well be a significant reduction in the rate of illegal parking. But short of such extreme situations, it seems that what punishments are actually inflicted on offenders makes little difference to general deterrence. For example, in Birmingham in 1973 a young mugger was sentenced to a draconian 20 years’ detention amid enormous publicity, and yet this sentence made no difference to the incidence of mugging offences in Birmingham or in other areas (Baxter and Nuttall, 1975; Beyleveld, 1980: 157). Similarly, studies have found little if any evidence that jurisdictions with harsh levels of sentencing benefit as a result from reduced crime rates (von Hirsch et al., 1999: ch. 6).
This does not mean that deterrence never works, but it does mean that its effects are limited and easy to overestimate. There are several reasons for this. First, most people most of the time obey the law out of moral considerations rather than for selfish instrumental reasons (Tyler, 1990; Paternoster et al. 1983). Second, people are more likely to be deterred by the likely moral reactions of those close to them than by the threat of formal punishment (Willcock and Stokes, 1968). Again, potential offenders may well be ignorant of the likely penalty, or believe they will never get caught. Some recent research has found that those who carry out robberies on enterprises such as banks and building societies tend to be dismissive of their chances of being caught even when they already have been caught and sent to prison, and as a result most do not think twice about the kind of sentence they might get (Gill, 2000). Much the same seems to be true of burglars (Bennett and Wright, 1984: ch. 6). Or the offender may comn-iit the crime while in a thoughtless, angry or drunken state. There is some good evidence that general deterrence can be improved if potential offenders’ perceived likelihood of detection can be increased5 but little to suggest that severer punishments deter any better than more lenient ones.
These truths were officially recognized by the then Conservative government in 1990, before Mr Howard’s announcement that ‘prison works’. The 1990 White Paper Crime, Justice and Protecting the Public (Home Office, 1990a:para. 2.8) stated: ‘There are doubtless some criminals who carefully calculate the possible gains and risks. But much crime is committed on impulse, given the opportunity presented by an open window or unlocked door, and it is committed by offenders who live from moment to moment... It is unrealistic to construct sentencing arrangements on the assumption that most offenders will weigh up the possibilities in advance and base their conduct on rational calculation. Often they do not.’
All of this suggests that, while general deterrence might form the basis of a plausible general just fication for having a system of punishment, it is more difficult to argue that the amount of punishment imposed by our system is justifiable by deterrent considerations. In terms of its deterrent effects, it seems almost certain that the English penal system is engaging in a massive amount of ‘overkill’. As we saw in Chapter 1 (especially Table 1.2), England has more prisoners proportionate to its population than most other countries in Western Europe. For example, contrast England with Finland, which in 1999 had 55 prisoners per 100,000 population compared with England and Wales’s 125. Unlike England, Finland in recent years has as a deliberate matter of policy sought to reduce its prison population (Törnudd, 1993), without noticeably poor effects on its crime rate which has risen at a similar rate to that of other European countries. Similarly, a significant reduction in the West German prison population in the 1980s did not lead to an increase in major crime or make the streets less safe (Feest, 1988; Flynn, 1995).
A utilitarian deterrence theorist ought to conclude from this that the English penal system is an immoral one. Jeremy Bentham (1970: 179) himself propounded the principle of ‘frugality’ (or ‘parsimony’), which states that punishments should be no more severe than they need to be to produce a utilitarian quantity of deterrence. ‘Overkill’ causes unnecessary suffering to the offender, and all suffering is bad unless it prevents a greater amount of suffering or brings about a greater quantity of pleasure.6 So, although for a utilitarian deterrence might justify having a penal system, it does not justify the one we actually have. We shall argue later that the same is true for our
preferred approach based on human rights.
Reform
Reform (or ‘rehabilitation’)7 is the idea that punishment can reduce the incidence of crime by taking a form which will improve the individual offender’s character or behaviour and make him or her less likely to re-offend in future. Reform as the central aim of the penal system was a highly popular notion in the 1950s and 1960s, when penological thought was dominated by ‘the rehabilitative ideal’. Many proponents of reform (of a kind known as ‘positivists’: see later in this chapter) have favoured a particularly strong version of this ideal called the ‘treatment model’. This views criminal behaviour not as freely willed action but (either metaphorically or literally) as a symptom of some kind of mental illness which should not be punished but ‘treated’ like an illness.
For some advocates of rehabilitation, optimism about reforming offenders has included the sentence of imprisonment, with incarceration being seen not so much as a retributive or deterrent punishment but as an opportunity to provide effective reformative training and treatment. For most rehabilitationists, however, the conventional wisdom has long been that ‘prison doesn’t work’ in reforming offenders, and so cannot be justified in these terms. Figures showing high rates of re-offending following release from custody are often quoted as bearing this out; for example a Home Office study of prisoners released in 1994 found that 56 per cent of offenders released from prison in 1994 (and 75 per cent of young males released from custody) were reconvicted within two years. After seven years, 73 per cent of released prisoners have been reconvicted (Kershaw, 1999). Statistics such as these led the government to state in the White Paper which preceded the Criminal Justice Act 1991 (Home Office, 1990a: para. 2.7) that imprisonment ‘can be an expensive way of making bad people worse’.
Although once dominant in penal discourse, the ideal of reform became discredited in the early 1970s, partly as a result of research results which suggested that penal measures intended to reform offenders were no more effective in preventing recidivism than were punitive measures. The received wisdom about reform came to be that ‘nothing works’, that ‘whatever you do to offenders makes no difference’, although this was always an exaggeration. It is true that in the 1970s extensive reviews of research in the United States (Lipton et al., 1975) and in Britain (Brody, 1976) found it to be generally the case that different penal measures had equally unimpressive outcomes in terms of re-offending. Similarly, recent studies (for example Kershaw, 1999) have found that, when account is taken of the differing characteristics of offenders8 sentenced to custody and various types of community sentence, the type of sentence they receive seems to make no discernible difference to whether they reoffend or not.9 However, studies from the 1970s onwards — including those most often quoted as evidence that ‘nothing works’ — have also found examples of reformative programmes which seem to work to some extent with certain groups of offenders (see Palmer, 1975). The generalized conclusion (associated with the American Robert Martinson)10 that ‘nothing works’ became widely accepted — not so much because it had been shown to be true, more because the disappointment of the high hopes invested in reform led to an over-reaction against the rehabilitative ideal.
In the past few years there has been something of a revival of the reformative approach. The new attitude — sometimes associated with the ‘Strategy B’ approach to criminal justice (see the Introduction) — has been that ‘something works’, that systematic experimentation, research and monitoring can identify effective methods of dealing with offenders. A burgeoning literature (for example Gendreau and Ross, 1987; McGuire, 1995) asserts that certain programmes and techniques have already been proved to have a significant effect in reducing reoffending. The current Labour government (and especially Home Secretary Jack Straw) has expressed particular interest in eliciting evidence as to ‘what works’ to reduce offending and applying the results of research evidence in practice: an ‘evidence-based’ policy of trying to increase the effectiveness of the criminal justice system. (Although, as we shall see, especially in Chapter 11, it is arguable that government policy is still based more upon ideological and political considerations than upon any dispassionate consideration of the evidence.)
The claims which are now made for the effectiveness of reformative measures are usually more modest than those which were put forward during the period of rehabilitative optimism. Few nowadays hold to the ‘medical’ or ‘treatment’ models of punishment, or claim that science can provide a cure for all criminality. Reform tends now to be seen not as ‘treatment’ which is imagined to work independently of the will of the offender, but as measures which enable or assist rather than force offenders to improve their behaviour - what has been called ‘facilitated change’ rather than ‘coerced cure’ (Morris, 1974: 13—20). Many of the currently fashionable programmes attempt to improve the cognitive and reasoning skills of offenders (for example, Ross et al., 1988), often by confronting them with the consequences and social unacceptability of their offending in the hope that they will as a result, decide to change their attitudes towards breaking the law. This ‘cognitive behavioural’ approach (which we discuss further in ChapterS) does not deny the offender’s free will, rather it appeals to it. It follows that reform can never be guaranteed to work (as, of course, research well and truly confirms). But it may still be well worth trying, even though we remain sceptical about some of the more enthusiastic recent claims for the effectiveness of reform. The empirical evidence may have destroyed the reformative aim as a plausible general justification of the penal system, but reform remains a reductivist aim which it may well be right to pursue within a system of punishment.
Incapacitation
Prison works, according to Conservative Home Secretary Michael Howard in 1993, not only by deterrence, but also because ‘it ensures that we are protected from murderers, muggers and rapists’ — apparently a reference to the reductivist mechanism known as ‘incapacitation’. Incapacitation simply means that the offender is (usually physically) prevented from re-offending by the punishment imposed, either temporarily or permanently. The practice in some societies of chopping off the hands of thieves incapacitates in this way (as well as possibly deterring theft). Similarly, one of the few obviously valid arguments in favour of capital punishment is that executed offenders never reoffend afterwards. Lesser penalties can also have some incapacitatory effects. Disqualification from driving does something to prevent motoring offenders from repeating their crimes. Attendance centres may be used to keep hooligans away from football matches. And imprisonment normally ensures that the offender is deprived of the opportunity to commit at least some kinds of offence for the duration. Not all crimes, by any means: many thefts and assaults (on staff and other inmates) take place in prison, as do drug offences, while headlines such as ‘Bootlegger ran £23m empire from prison’ (The Guardian, 2 December 1999) exemplify some of the other criminal opportunities open to the incarcerated felon. But it is true that offences such as domestic burglary and car theft become somewhat more difficult when you are locked up in prison.
Life imprisonment is one sentence which is specifically used in many cases for the purposes of incapacitation. A ‘life imprisonment’ sentence would be more precisely described as a potentic?lly lifelong prison sentence, since most ‘lifers’ are eventually released; but the life sentence means that they will not be released as long as it is believed that they pose an unacceptable risk of serious reoffending (see further Chapter 8). Life sentences may be imposed, and lifers kept in prison~ even though this exceeds what would be a normal length sentence proportionate to the seriousness of the offence.
It is certainly a plausible claim that incapacitation could be a justification (or partial justification) for punishments such as disqualification from driving and attendance centre orders. As to whether and how far incapacitation can serve to justify imprisonment, one key issue is the factual question of how effectively prison reduces crime in this way. Although only rough estimates are possible, the best calculations suggest that the incapacitation effects of imprisonment are only modest. This is largely because most ‘criminal careers’ are relatively short, so that by the time offenders are locked away they may be about to give up crime or reduce their offending anyway. The most authoritative estimate, by the former head of the Home Office Research and Planning Unit, Roger Tarling (1993: 154), is that ‘a change in the use of custody of the order of 25 per cent would be needed to produce a 1 per cent change in the level of crime’. On the other hand, the prison population could be substantially reduced without creating a massive crime wave: if the numbers in prison were cut by 40 per cent, this could be expected to lead to an increase in criminal convictions of only 1.6 per cent (Brody and Tarling, 1980).
Nor is there much evidence that incapacitatory sentences can be targeted with any great success or efficiency on more selected groups of repeat offenders who are particularly likely to reoffend (Tarling, 1993: 154—160; Hagell and Newburn, 1994). Nor can we accurately predict which offenders are likely to commit particularly serious crimes if they do reoffend (Levi, 1997):
our powers of prediction are simply not up to the job, whether we use impressionistic guesswork, psychological testing, statistical prediction techniques or any other method. If we do try to pick out individuals in any of these ways and subject them to extra-long sentences on the basis of our predictions, we will be imprisoning a large number of people who would not in fact reoftend; typically at least twice as many as those who actually would offend again. And even if it was possible to target potential recidivists or those likely to commit grave crimes, this would run into the ethical objection that we were punishing people not for what they have done but for what they might do in the future — punishment for imaginary crimes in the future rather than real ones in the past — which might not be fundamentally wrong in principle to a utilitarian, but is a serious objection for most moral codes, including retributivism and human rights theory.
It seems unlikely, then, that incapacitation can provide a general justification for our present practice of imprisonment, let alone justify increasing our use of imprisonment,” or for introducing any new incapacitatory measures. Nevertheless, the current trend in both England and the United States is for governments to create new sentences explicitly aimed at achieving incapacitation.’2 Most US jurisdictions now have so-called ‘three strikes and you’re out’ laws, whereby repeat offenders are automatically jailed for life for a third offence. Under these laws, people have literally been sent to prison for life without parole for offences such as stealing a slice of pizza, which was the third offence of the unfortunate Jerry Williams in California in 1995 (The Guardian, 13 October 1995). In England, provisions introduced in the Criminal Justice Act 1991 (sections 1(2)(b) and 2(2)(b)) permit extra-severe custodial sentences for violent and sexual offenders if the court thinks this is necessary ‘to protect the public from serious harm’. And the Crime (Sentences) Act 1997 imported the ‘three strikes and you’re out’ principle, with semi-mandatory prison sentences for adult domestic burglars and drug dealers convicted for a third time, and semi-mandatory life imprisonment for serious violent offenders who offend a second time (see Chapter 4).
RETRIBUTIVISM
The retributivist principle — that wrongdoers should be punished because they deserve it — is in some ways the complete antithesis of reductivism. Where reductivism is forward-looking, retributjvism looks backwards in time, to the offence. It is the fact that the offender has committed a wrongful act which deserves punishment, not the future consequences of the punishment, that is important to the retributivist. Retributivism claims that it is in some way morally right to return evil for evil, that two wrongs can somehow make a right.
Retributivism is sometimes combined with reductivism to produce a compromise theory. Often these compromise theories state, in effect, that punishment is only justified if it is both deserved and likely to have deterrent effects (for example von Hirsch, 1976: chs 5 and 6).
If people are to be punished because they deserve it, it is natural to say that they should also be punished as severely as they deserve — that they should get their just deserts. Retributivism thus advocates what is known as a tariff a set of punishments of varying severity which are matched to crimes of differing seriousness: minor punishments for minor crimes, more severe punishments for more serious offences.13 The punishment should fit the crime in the sense of being in proportion to the moral culpability shown by the offender in committing the crime. The Old Testament lex talionis (an eye for an eye, a life for a life, etc.) is one example of such a tariff, but only one: a retributive tariff could be considerably more lenient than this, as long as the proportionate relationship between crimes and punishments was retained.
This is a point which needs stressing, because it is a common mistake —certainly among our own students – to assume that retributivists are those who advocate the harshest punishments, and to equate retributivism with a draconian, ‘Strategy A’ approach to criminal justice. In fact, it is often the case that retributivists (for example, those who follow the ‘justice model’ of ounishment we discuss later in this chapter) actually favourre)ative)y lenient punishment. (But punishment which is ultimately justified by the fact that it is deserved and proportionate.) On the other hand, some notable exponents of Strategy A — such as Michael Howard — have attempted to justify their harsh penal policies by appeals to their supposed effectiveness in controlling crime by reductivist mechanisms such as deterrence and incapacitation. The mistake is understandable, and there may be a certain psychological truth behind it. Maybe many advocates of Strategy A are primarily motivated by a hatred of criminals and a wish to see them ‘get what they deserve’ rather than by a desire to pursue rational steps to reduce crime, whatever their proclaimed motives. But retributivism is not inherently harsher than other philosophies, and indeed it has certain attractive features to those of a humane disposition.
One of these attractive features is its consonance with what is generally acknowledged to be one fundamental principle of justice: that like cases should be treated alike. (‘Like’ for retributivists means alike in the intuitively appealing sense of ‘similarly deserving’.)
Another attractive feature of retributivism is that there is a natural connection between the retributive approach and the idea that both offenders and victims have rights. Reductivist theory (at least in its utilitarian form) has always found it difficult to encompass the notion of rights, even when it comes to providing entirely innocent people with a right not to be punished. (For if we could achieve the desired reductive consequences by framing an innocent person, and if these effects are all that is needed to justify punishment, what would be wrong with punishing the innocent?) Retributivism has no such problem, since it follows automatically from the retributive principle that it must be wrong to punish non-offenders. Nor may we punish criminals to a greater extent than their crimes are felt to deserve (for example in the hope of reforming or incapacitating them or deterring others): under the retributivist principle offenders have a right to go free once they have ‘paid their debt to society’. Life imprisonment for stealing a pizza would be ruled out as disproportionate, for example. Retributivisrn thus fits in well with our common-sense intuitions which insist that it is indeed morally relevant whether the person punished has behaved well, badly or very badly in the past. Probably for this reason, it has proved a remarkably resilient idea. Although for many years retributivism was regarded (at least in academic circles) as outmoded and even atavistic, it has enjoyed a major revival in recent years, notably in the form of the ‘justice model’ discussed later in this chapter.
But retributivism is not without its own philosophical difficulties. One problem is how to justify the retributive principle itself. It may accord with some of our moral gut reactions, which seem to tell us that wrongdoers should be made to suffer. But maybe these reactions are merely irrational vindictive emotions (akin to vengeance) which, if we are to be moral, we ought to curb rather than indulge. A related objection is that it is not immediately clear how the retributivist principle relates to any general notion of what is right or wrong. At least utilitarian reductivisrn has the virtue that it can be derived from the general moral and political theory of utilitarianism.
Some theorists have attempted to counter these objections by reference to the ‘social contract’, a theory which provides a general account of political obligation (see especially Murphy, 1979). The idea is that all citizens are bound together in a sort of multilateral contract which defines our reciprocal rights and duties. The terms of this contract include the law of the land, which applies fairly and equally to all of us. The lawbreaker has disturbed this equilibrium of equality and gained an unfair advantage over those of us who have behaved well and abided by the rules. Retributive punishment restores the balance by canceling out this advantage with a commensurate disadvantage. It thus ensures that wrongdoers do not profit from their wrongdoings, and is justified because if we failed to punish lawbreakers it would be unfair to the law-abiding.
This ‘modern retributivism’ was highly influential for a time, although it was always far from universally accepted and it eventually became discredited even in the eyes of some of its foremost former advocates (Murphy, 1992:24—5, 47—8; von Hirsch, 1986: ch. 5; and 1993: ch. 2). But even if we assume that it is sound at an abstract philosophical level, it would be extremely dubious to assert that this theory can justify our present practices of punishment or anything like them. The main difficulty is that the theory only applies if our society is a just one in which all citizens are genuinely equal; otherwise there is no equilibrium of equality for punishment to restore. If — as appears to be the case — detected offenders typically start from a position of social disadvantage (which means that the obligation to obey the law weighs more heavily upon them than on others) then punishment will tend to increase inequality rather than do the opposite. In fact, this was exactly the conclusion once reached by the modem retributivist Jeffrie Murphy (1979:95), who stated that ‘modem societies largely lack the moral right to punish’)4 Even if such a sweeping conclusion is not warranted, retributivists should be strongly critical of many aspects of our penal system. Not least among these are the lack of consistency in sentencing practices (see Chapter 4) and the use of incapacitatory sentences, which mean that offenders are to a great extent not dealt with in proportion to their just deserts. So despite its resilient popularity, retributivism remains an implausible justification for our actual practices of punishment.