Recidivist premiums are sentence enhancements given to individuals for their most recent criminal offenses because they have continued to commit crimes after having previously been punished for other offenses. Though considerable variation in such enhancements exists, they can be found, in some form or other, in most legal jurisdictions throughout the world (Bagaric 2010: 10; Roberts 2008a: 468). However, the ubiquity of this sentencing practice is hard to square with its insecure theoretical foundations. Though numerous defenses of recidivist premiums have been advanced, most have been found wanting (Frase 2012; Bagaric 2010; Roberts 2008a). Some sentencing theorists react to the unimpressive arguments for recidivist premiums by advocating their elimination (Bagaric 2010). Other theorists doubt that the publicās appetite for such premiums can be so blithely ignored (Roberts 2008b). It seems fair to say that the publicās support of recidivist premiums, though apparently strong, is somewhat undifferentiated in the sense that it does not translate into highly specific recidivist sentencing policies. This, in turn, suggests that public demands for such policies might be tutored in order to produce recidivist sentencing practices that are more constructive and fair than is often the norm.
The discussion that follows is divided into three sections. In the first, I briefly survey the kinds of recidivist premiums that can be found in existing legal jurisdictions and discuss the theoretical defenses that have been proffered for at least some of them. Like others, I find the arguments given on behalf of recidivist premiums unconvincing, though I agree with those theorists who are reluctant to simply urge their elimination. In the second section, I advance a hypothesis concerning the publicās support of recidivist premiums. When sentencing practices are commonplace, the theoretical bases of which are suspect, it makes sense to ask whether something else, besides theory, explains them. I suggest that certain basic facts about human social cooperation might explain the tendency to enhance the sentences given to repeat offenders. Individuals might be rather crudely evaluated by their fellow citizens as either being on board with the rules of the group or prepared to disobey them. When individuals are perceived as being unwilling to abide by the rules of the group, the punishments meted out to them steadily escalate with an eye toward limiting their access to the group for longer and longer periods of time. By contrast, offenders who appear remorseful signal their allegiance to group norms. In response, such offenders are typically granted some sentence mitigation, though they have flouted the rules of the community. Importantly, the theoretical basis for such mitigation has likewise been challenged. Yet, even if we must honor the largely instinctive reactions of the public to what the conduct and attitudes of offenders signal about rule-compliance, little else follows about sensible sentencing practices. This is where the lessons that have emerged from the scholarship of recidivist premiums might be drawn upon in crafting better sentencing policies.
In the third section, I apply these lessons, arguing that the public sentiments that engender recidivist premiums should be steered away from disproportionate and costly versions of them. Not only might the public be convinced to give young, first-time offenders some mitigation of their sentences, the public might be led to see how progressively enhanced recidivist premiums make little sense in light of the fact that most offenders naturally desist from further offending as they age. Also, if evidence of rule-compliance is crucial, the public should support early release of offenders who show signs of being rehabilitated. Furthermore, the retention of lengthy sentence enhancements for a narrowly defined group of particularly violent recidivists might serve to reassure the public that a more general easing off of recidivist premiums will not come back to haunt it.
One recurring theme throughout the discussion involves what I term the āsocial deprivation narrative.ā Many if not most criminal offenders come from socially deprived backgrounds (Clear 2007: 61). Their lives are already difficult, which makes following the rules society enforces more challenging for them than for other citizens. Further, their encounters with the criminal justice system harm them in a variety of well-documented ways (Haney 2006; Western 2006). Our responses to evidence of rule non-compliance are often insensitive to these complications. But the interesting question is whether our punitive responses might be tempered by a better understanding of social deprivation and the corrosive effects of legal punishment. I contend that there is reason to believe that they might be, though our patience with the rule-defiant will inevitably have its limits.
Part one: The theory of recidivist premiums
Sentencing policies for repeat offenders can be conceived of in terms of two extremes: At one extreme, we might adopt what are termed āflat-rateā sentences for such offenders, by which each crime that they commit is punished according to an estimation of its seriousness with no enhancement for previous offenses (Ashworth 1995: 152). In effect, a flat-rate approach elides an offenderās criminal history in sentencing her for her most recent offense. At the other extreme, we might adopt āescalatingā sentences, wherein repeat offenders receive sentences for their current crimes that reflect the seriousness of those crimes plus an enhancement based on the length of their criminal histories. The longer their criminal histories, the greater the enhancement could, in theory, become. Indeed, the criminal history enhancement could exceed if not dwarf the sentence for the current offense. Again, few, if any, legal jurisdictions have flat-rate sentencing for recidivists. Few also have the more extreme forms of escalating sentencing, though some jurisdictions in the United States come close and their āhabitual offenderā or āthree strikesā sentencing provisions produce sentences in which the premiums for recidivism exceed, by a considerable margin, the sentences assigned offenders for their most recent convictions (Greenberg 2002: 238). Most legal jurisdictions fall somewhere in-between flat-rate and the more draconian forms of escalating sentencing.
To simplify things, I will focus on retributive and crime reduction rationales for recidivist premiums. Other theories of sentencing exist, but retribution and crime reduction are invariably the predominant concerns of the legislators, sentencing commissions, or judges who determine sentencing policies. Retributive, or desert-based, rationales for recidivist premiums seem especially unpersuasive. The reason for this is succinctly captured by Mirko Bagaric (2010: 13; see also Fletcher 1978). Recidivist premiums seem to punish offenders repeatedly for crimes for which they have already been punished. Suppose that offender S commits a crime at time T1 and at T2 receives the deserved punishment for that offense. Suppose also that S then commits another crime at T3 and at T4 receives not only punishment for his second crime but also an enhancement based on his previous offense. How can we avoid the conclusion that S is being punished twice for his crime at T1? The answer might be that his offending at T3 is somehow made worse (that is, more culpable) by his offending at T1, and therefore he is deserving of harsher punishment at T4 than his second offense, all by itself, would warrant. It could be argued, for instance, that his continued offending reveals Sās bad character, which renders him more blameworthy for his subsequent offenses; his punishment at T4 reflects that. Alternatively, it could be argued that in spite of his having been put on notice that his conduct at T1 was unacceptable, his continued offending shows that he is defiant. S has not absorbed the lesson conveyed by the punishment for his first offense about the importance of following the legal rules. His enhanced sentence at T4 is deserved because his defiance makes S more blameworthy.
Neither of these rationales for recidivist premiums seems convincing. Retributivists should balk at punishing people for their bad character. You can have a thoroughly bad character, but as long as you do not break the law, you do not deserve to be legally punished (Bagaric 2010: 15). When you do break the law, you deserve punishment for that infraction, based on its seriousness. Crime seriousness is typically held by retributivists to be a function of the harm inflicted or risked by an offense and the culpability of the offending agentāin particular, whether she inflicted or risked harm deliberately, recklessly, or negligently (von Hirsch 1993: 29). Bad character sentence enhancements punish people for who they are, not for what they have done and how blameworthy they were in doing it. Also, would each subsequent offense demonstrate that an individual had even worse character and therefore is eligible for increasing enhancements? It would seem so, but then the problem is that, at some point, recidivist sentencing enhancements will exceed the sentences for individualsā most recent offenses. Such an outcome should be worrisome to retributivists for two reasons: First, sentences for repeat offenders might exceed cardinal proportionality limits, that is, become so harsh that they are non-comparatively unjust (von Hirsch 1993: 18). For instance, a 25-year sentence for a petty theft is unjust no matter how lengthy an individualās criminal history. Second, sentences for repeat offenders might exceed ordinal proportionality limits. This will occur if the sentences assigned to repeat offenders for their most recent crimes wind up being significantly harsher than the sentences they ought to be assigned for considerably more serious offenses (Roberts 2008b: 475; Frase 2012: 183). A petty thief with a long criminal history should not be punished as if he were guilty of having committed homicide.
Similarly, people can have defiant attitudes toward the criminal law, regarding it as little more than an oppressive regime aimed at protecting the rich and powerful at the expense of the poor and weak, yet as long as they otherwise remain within its strictures, they do not merit punishment. It is not apparent why the attitudes individuals have toward the criminal law suddenly become relevant to determining their sentences once they have committed one or more crimes. Also, the defiance rationale seems incapable of explaining why recidivist premiums should continue to increase with each subsequent offense. Even if it makes sense to punish second-time offenders more than first-time ones, due to the formerās defiance of the rules, once such an enhancement has been added to an offenderās sentence, why keep increasing it for each subsequent offense he commits? Defiance seems all-or-nothing, and so unlike character, which can range from bad to worse. This suggests that defiance-based sentence enhancements should not increase with each subsequent offense. If in spite of this they do increase, then the likelihood that such enhancements will run afoul of cardinal and ordinal proportionality limits comes into play.
A different approach to recidivist premiums has been urged by Andrew von Hirsch and several of his co-authors over the years. Von Hirsch claims that first- and perhaps some second-time offenders are less culpable for their criminal acts than third- or fourth-time offenders; hence first- or second-time offenders might be reasonably punished less than their crimes would otherwise merit (Wasik and von Hirsch 1994: 410; von Hirsch and Ashworth 2005: 148; von Hirsch 2010). First-time (and some second-time) offenders might have simply ālapsedā from otherwise law-abiding lives, a common enough form of human frailty that should be acknowledged with reduced punishment. However, if individuals persist in offending, it is no longer plausible to view their offending as a lapse and so they deserve full (but not enhanced) punishment. The āprogressive loss of mitigationā (or PLM) approach is not designed to support harsh recidivist premiums of the kinds found in many jurisdictions. For many theorists, this is a strong point in its favor. Against the PLM approach, it has been argued that not all first-time offenses appear to be ālapses.ā They are deliberately planned crimes by adults (Frase 2012: 182); as such it is not clear why the punishment of such offenders should be reduced. Another problem is that we should hesitate to give people too much credit for their past law-abidingness. Maybe they have just been lucky, never encountering sufficiently compelling temptations to offend.
It could be argued that we punish more harshly those whose bad characters or defiance have been revealed not because they are more blameworthy but because we reasonably fear that they will go on to commit other crimes and wish to deter or incapacitate them from doing so. At first glance, a crime reduction approach to defending recidivist premiums appears to be on firmer ground. Individuals who continue to offend after having been punished for earlier offenses would seem to be in need of stronger incentives, in the form of enhanced sentences, to desist from future offending. The longer their offending persists, the stronger those incentives might have to become. Or if enhanced sentences do not work to reduce crime via deterrence, at least incapacitating repeat offenders for progressively longer periods of time might seem to be in order. After all, one of the best predictors of future offending is past offending (Frase 2012: 178). At the extreme, individuals with lengthy criminal records might be justifiably incapacitated under habitual offender or āthree strikesā provisions designed to imprison them for many years, if not for the rest of their lives.
Critics of crime reduction rationales for recidivist premiums point out that the evidence linking longer sentences with desistance from crime is singularly unimpressive (Frase 2012: 189; Doob and Webster 2003). Even as the āpriceā of crime goes up, there seems little detectable decrease in the demand for it. As for recidivist premiums and incapacitation, there is no gainsaying the fact that when individuals are locked up in prisons, their abilities to reoffend are limited. However, the evidence strongly suggests that by their mid-thirties, most individuals discontinue their criminal careers all on their own (Nagin 1998: 364). This āaging outā phenomenon means that substantial recidivist premiums assigned once offenders have reached their late twenties will produce few incapacitation effects but will be enormously costly. Habitual offender or āthree strikesā provisions might uselessly and expensively keep offenders imprisoned into old age. Also, at least for certain kinds of offenses, removing individuals from society by imprisoning them does little to reduce offending in communities, since āreplacementsā quickly emerge to take up their roles in drug or property theft rings (Nagin 1998: 365).
Overlaying all of these difficulties with the standard rationales for recidivist premiums is what might be termed a āsocial deprivation narrative.ā The fact is that many criminal offenders come from socially deprived backgrounds and will return to them once they have served whatever sentences are assigned to them for their crimes. The diminished opportunities such individuals faced before they were convicted are constricted further once they acquire criminal records (Western 2006). Moreover, imprisonment often warps their personalities in predictable and worrisome ways, especially in societies (like the United States) in which efforts to rehabilitate offenders are half-hearted or non-existent (Haney 2006). It is thus hardly surprising that many of the individuals who have acquired criminal records continue offending; their past records leave them with relatively few viable alternatives for earning income or gaining social status. Add to this the well-known fact that it is generally easier to police the crimes of the poor than those of the rich or middle class (Stuntz 1998). This means that the profile of offenses that the criminal justice system processes is skewed in ways that make the poor look like worse rule-breakers than they probably are, at least compared with their fellow citizens who are better off.
This social deprivation narrative bears on the theoretical debate about recidivist premiums in a number of ways. For one thing, it casts doubt on the notion that recidiv...
