Harvard Law Review

March, 2001

 

Commentary

 

*1429 PUNISHING DANGEROUSNESS: CLOAKING PREVENTIVE DETENTION AS CRIMINAL

JUSTICE

 

Paul H. Robinson [FNa1]

 

 

 

 

Copyright ©  2001 Harvard Law Review Association; Paul H. Robinson

 

 

 

 

  Laypersons have traditionally thought of the criminal justice system as being in the business of doing justice: punishing offenders for the crimes they commit. [FN1]  Yet during the past several decades, the justice system's focus has shifted from punishing past crimes to preventing future violations through the incarceration and control of dangerous offenders.  Habitual- offender statutes, such as "three strikes" laws, authorize life sentences for repeat offenders. [FN2]  Jurisdictional reforms *1430 have decreased the age at which juveniles may be tried as adults. [FN3]  Gang membership and recruitment are now punished. [FN4]  "Megan's *1431 Law" statutes require community notification of convicted sex offenders. [FN5]  "Sexual predator" statutes provide for the civil detention of sexual offenders who remain dangerous at the conclusion of their criminal commitment. [FN6]  New sentencing guidelines increase the sentence of offenders with criminal histories because these offenders are seen as the most likely to commit future crimes. [FN7]  These reforms boast as their common denominator greater official control over dangerous persons, a rationale readily apparent from each reform's legislative history. [FN8]

 

  *1432 Although the individual legislative histories make clear that a preventive rationale has motivated each of these reforms, the system's general shift from punishment toward prevention has not been accompanied by a corresponding shift in how the system presents itself.  While increasingly designed to prevent dangerous persons from committing future crimes, the system still alleges that it is doing criminal "justice" and imposing "punishment." Yet it is impossible to "punish dangerousness."  To "punish" is "to cause (a person) to undergo pain, loss, or suffering for a crime or wrongdoing" [FN9]--therefore, punishment can only exist in relation to a past wrong. "Dangerous" means "likely to cause injury, pain, etc." [FN10]--that is, dangerousness describes a threat of future harm.  One can "restrain," "detain," or "incapacitate" a dangerous person, but one cannot logically "punish" dangerousness.

 

  Why the shift to preventive detention?  Why the wish to keep the old criminal "punishment" facade?  These are the starting points of inquiry in this Commentary.  It concludes that the trend of the last decade--the shifting of the criminal justice system toward the detention of dangerous offenders--is a move in the wrong direction.  The difficulty lies not in the laudable attempt to prevent future crime but rather in the use of the criminal justice system as the vehicle to achieve that goal.  The approach perverts the justice process and undercuts the criminal justice system's long-term effectiveness in controlling crime.  At the same time, the basic features of the criminal justice system make it a costly yet ineffective preventive detention system.

 

  Segregation of the punishment and prevention functions offers a superior alternative.  Punishment and prevention are fundamentally different; they rely on different criteria and call for different procedures. [FN11] Punishment, especially through imprisonment, happily produces a beneficial collateral effect of incapacitation.  If preventive detention is needed beyond the prison term of deserved punishment, it ought to be provided by a system that is open about its preventive purpose and is specifically designed to perform that function.

 

 

I. How Strong Is the Need for Increased Prevention?Strong Is the Need for

Increased Prevention?

 

  It is difficult to deny that a society ought to be able to defend itself against persons who will cause serious harm.  The goal of providing such protection underlies traditional civil commitment systems that detain persons who are dangerously mentally ill or who have contagious *1433 diseases or drug dependencies. [FN12]  In any case, political forces inevitably will press for protective measures if a perception of public vulnerability exists.

 

  Given the history of crime rates, the recent enthusiasm for protective reforms is predictable.  Despite recent declines, the violent crime rate remains more than three times higher than it was during the decade following World War II, when the baby-boomers, now the civic and political leaders, were growing up. [FN13]  Today's aggravated assault rate is nearly four times what it was almost forty years ago. [FN14]  News reports commonly celebrate that crime rates are back to the levels of the late 1970s, but fail to note that by that time the long unbroken string of annual crime increases had already tripled the rates of the 1950s. [FN15]  Given the widening epidemic of juvenile crime, the unknown future effects of a wave of crack babies, and a host of other both predictable and unpredictable changes, the current decline in crime rates may not continue.  Even if it did, the declining crime rate of the last eight years would have to continue unbroken for another three decades before crime levels returned to those enjoyed by baby-boomers as children. [FN16]

 

  Even a return to the low crime rates of the 1950s would leave Americans with reason to be dissatisfied.  As a result of the past decades of crime increases, people have seriously altered their lifestyles.  We no longer let our children walk home from school.  We dead bolt our doors, use "The Club" in our cars, and live in security-staffed apartment buildings and gated communities. [FN17] Current crime rates are high despite these precautions and would be even higher without them.  Recapturing the security of life we enjoyed a half century ago would necessitate not only a return to the crime rates of the 1950s but *1434 also the freedom of action enjoyed during that period.  The seeming impossibility of such a return highlights how much we have lost to crime since the 1950s.the Criminal Justice System for Preventive Detention

 

 

II. The Justice Problems

 

  From this perspective, it is understandable that today's citizens are demanding greater protection and that legislators are seeking new ways to provide it. [FN18]  But the use of the criminal justice system as the primary mechanism for preventing future crimes seriously perverts the goals of our institutions of justice.

 

  Lowering the age for adult prosecution, with its longer terms of imprisonment, is likely to increase societal protection.  Juveniles are committing an increasing number of serious crimes. [FN19]  But decreasing the age at which a juvenile can be prosecuted as an adult increases the number of cases in which a young offender lacking the capacity for moral choice is nonetheless held criminally liable.

 

  There is little dispute that many young offenders, especially those below the age of fifteen, lack the cognitive and control capacities of normal adults. Some may not appreciate the enormity of the consequences of their acts and others may lack normal behavior control mechanisms. [FN20]  If an adult offender is similarly dysfunctional, due to insanity or involuntary intoxication for example, an excuse defense is generally available. [FN21] Yet a young offender impaired in a similar way *1435 by immaturity has no defense or mitigation, because adult courts traditionally have not recognized an immaturity excuse. [FN22]  Courts have had no need to make such an excuse available in the past for the obvious reason that juvenile courts dealt with the cases involving youthful offenders.  The recent trend toward trying youths in adult courts has created the need for such an excuse defense, but none has been developed, perhaps because the defense would interfere with the goal of gaining control over dangerous offenders without regard to their blamelessness.

 

  A more common and more damaging distortion of justice derives from the use of "three strikes" and other habitual-offender statutes, and the use of prevention-oriented sentencing guidelines that dramatically increase sentences for offenders with prior criminal records.  These reforms affect nearly every case in which an offender has a prior criminal record.

 

  Shocking cases of long-term imprisonment for minor offenses are well known.  In Rummel v. Estelle, [FN23] for example, the defendant took $129.75 from a bar owner to fix the bar's air conditioner with no intention of actually doing so.  His conviction for fraud was his third, qualifying him for a term of life imprisonment without the possibility of parole under an early "three strikes" statute. [FN24]

 

  But problems are inherent not only in the shocking cases but in every case in which a habitual-offender statute or prior-record-based sentencing guideline applies.  In these cases, the sentence imposed exceeds the deserved punishment, albeit to a less dramatic extent than life imprisonment for minor check fraud. The imposition of that excess punishment is, of course, the motivating goal of such statutes: they significantly increase the sentence beyond the level deserved for the crime because a prior record may predict future offenses.  But the effect of such a policy is that the criminal justice system regularly imposes sentences that exceed the punishment deserved.  Sentencing guidelines that give great weight to prior criminal records and "three strikes" and related habitual-offender provisions commonly double, triple, or quadruple the punishment imposed on repeat offenders. [FN25]  An *1436 initial portion of the sentence may well be deserved, but what follows is a purely preventive detention portion that cannot be justified as deserved punishment.

 

  One can construct a theory that makes a prior criminal record relevant to deserved punishment, as Andrew von Hirsch has done. [FN26]  By committing an offense after a previous conviction, an offender might be seen as "thumbing his nose" at the justice system.  Such disregard may justify some incremental increase in punishment over that deserved by a first-time offender, but it seems difficult to justify the doubling, tripling, or quadrupling of punishment because of nose-thumbing. [FN27]  The recidivist nature of a second robbery is only one of *1437 many characteristics that determine blameworthiness. [FN28]  Lay intuitions may see the nose-thumbing as making the second robbery more condemnable than the first but not more condemnable than the second robbery itself, and certainly not twice as condemnable as the second robbery.  But note that, although nose-thumbing may justify a minor portion of the dramatic increases imposed for a prior record, the theory allows proponents of preventive detention to implement their program unobtrusively within a system of criminal punishment.

 

  Further, if such disrespect for law provided the impetus for these statutes, the aggravation of blameworthiness and increased punishment would apply to all offenses.  That is, if nose-thumbing is itself condemnable, then it ought to be condemnable in every context, not just in selected contexts.  Nose-thumbing through a second violent offense might be more condemnable than nose-thumbing through a second theft offense, but nose-thumbing through a second theft would hardly be irrelevant.  Yet the three strikes provisions typically apply only to a limited class of offenses--commonly violent offenses [FN29]--and typically account for only certain kinds of criminal history--again, commonly a history of violent offenses. [FN30]  It seems difficult to construct a desert theory of nose-thumbing disrespect that allows for such selective increases in punishment.  But note that applying habitual-offender schemes only to violent offenses does make sense under a prevention rationale, however, because these offenses most demand prevention.

 

  The criminal justice system's focus on dangerousness also causes, albeit less frequently, distortions of the reverse sort: failures of justice in which a person fails to receive the punishment he or she deserves.  This kind of error can occur both in the assignment of liability and in the assessment of the proper amount of punishment.  For example, the Model Penal Code provides a defense to inchoate liability if a person "presents [no] public danger" and the person's attempt was "inherently unlikely" to succeed. [FN31]  Such a defense may make sense for a system designed to incapacitate the dangerous person because incarcerating the nondangerous attempter is a waste of preventive resources.  But if the person believes his conduct will cause a criminal harm, the person deserves punishment whether or not the chosen method is likely to succeed.  For example, the HIV-positive son who attempts to kill his *1438 long-hated father by spitting on him [FN32] can escape liability if the killing method is impossible and he is not otherwise dangerous. [FN33]  But if the son's intention to kill his father unjustifiably is real and he has shown a willingness to carry out the intention fully, his blameworthiness is clear.

 

  Such failures of justice are more common in sentencing, at least in the discretionary systems that abounded two decades ago and that still exist in many jurisdictions.  The judge who focuses on prevention instead of desert [FN34] will give a minor sentence for a serious offense if the offender is no longer dangerous.  Thus, the recently discovered, elderly former Nazi concentration camp official can escape the punishment he deserves. [FN35]

 

  These conflicts between pursuing justice and incapacitating dangerous persons should come as no surprise.  Dangerousness and desert are distinct criteria that commonly diverge.  Desert arises from a past wrong, whereas dangerousness arises from the prediction of a future wrong.  A person may be dangerous but not blameworthy, or vice versa.  Consider, for example, a mentally ill offender.  A desert distributive principle acquits the dysfunctional person of all criminal liability because the person is not to blame for the offense; he deserves no punishment.  But an incapacitation principle would impose liability and require incapacitation because the offender is dangerous. [FN36]

 

  In a reverse set of cases, an incapacitation principle does not call for punishment of an offender even though the desert principle calls for conviction, as with the elderly Nazi official and the HIV-positive *1439 spitter. [FN37]  Because the person's conduct is harmless and the person is not otherwise dangerous, an incapacitation principle suggests that imposing criminal sanctions is a waste of resources. [FN38]  The desert principle, in contrast, takes the person's attempt to kill as evidence of blameworthiness deserving punishment.

 

 

III. The Inevitable Conflict Between Desert and Dangerousness as Distributive

Criteria

 

  The inherent conflict between incapacitation and desert has practical implications, as in the difference in the kinds of factors taken into account in assessing liability and determining sentences.  If incapacitation of the dangerous alone determined the distribution of criminal sanctions, prison terms would be set according to those factors that best predicted future crime.  The higher the likelihood of recidivism, the stronger the case for imprisonment and, often, the longer the sentence.  One of the best predictors of future criminality is employment history. [FN39]  Thus, unemployment for the two years preceding the crime could aggravate the grade of an offense or increase the imposed sentence.  An offender's age and family situation are also good predictors of future criminality, [FN40] and thus could also determine the offender's liability and sentence: younger offenders and offenders without fathers in the home would receive longer prison terms.  Indeed, if incapacitation of the dangerous were the only distributive principle, there would be little reason to wait until an offense were committed to impose criminal liability and sanctions; it would be more effective to screen *1440 the general population and "convict" those found dangerous and in need of incapacitation. [FN41]

 

  Yet openly relying on the factors relevant to an incapacitative principle would be offensive to a system of just punishment.  A person does not deserve more punishment for an offense because he has a poor employment history, is young, or has no father in his household. [FN42]  And certainly, no person deserves punishment before committing an offense.

 

  The incapacitative principle not only focuses on different criteria than the desert principle, but also wholly neglects factors central to the desert principle.  Even the nature of the crime committed may be of little relevance if the goal is prevention.  Consider, for example, the Model Sentencing Act, which was drafted in the early 1960s and considers *1441 only the purposes of rehabilitation and incapacitation of the dangerous.  The Report of Model Act proudly points out:

    The [Act] diminishes [differences in] sentencing according to the particular offense.  Under [the Act] the dangerous offender may be committed to a lengthy term; the non dangerous defendant may not.  It makes available, for the first time, a plan that allows the sentence to be determined by the defendant's make-up, his potential threat in the future, and other similar factors, with a minimum of variation according to the offense. [FN43]

 

  The point is that the traditional principles of incapacitation and desert conflict; they inevitably distribute liability and punishment differently.  To advance one, the system must sacrifice the other.  The irreconcilable differences reflect the fact that prevention and desert seek to achieve different goals.  Incapacitation concerns itself with the future--avoiding future crimes.  Desert concerns itself with the past--allocating punishment for past offenses.

 

 

IV. Denying the Conflict

 

  One of the most troublesome aspects of the conflict between incapacitation and desert is the denial that it exists.  People commonly believe that incapacitation and desert somehow can combine or reconcile in a way that allows both to achieve their objectives.  The Model Penal Code of the American Law Institute, for example, lists all of the traditional purposes of sentencing, including incapacitation and desert, and then directs judges to fashion sentences that most effectively further all of the purposes. [FN44]  The Code's commentary explains that if the purposes conflict in a particular case they should be "justly harmonized." [FN45]  Other writers have suggested that these competing interests *1442 are to be "balance[d]," [FN46] or "blended," [FN47] or "accommodate[d]." [FN48]

 

  But how can this be done?  When incapacitation and desert conflict, the principles suggest different sentences, and a judge or sentencing commission must choose between purposes.  Furthering one aim necessitates sacrificing the other.  Or, if a judge averages the sentences advocated by the two conflicting purposes, the resulting sentence may serve neither function effectively. [FN49]

 

  Norval Morris and others offer another argument to deny the existence of the conflict--a system may set sentences according to dangerousness without violating desert principles simply by avoiding any extreme disparity between levels of punishment and blameworthiness. [FN50]  This view conceives of desert as having only vague requirements, which operate at the extremes of disproportionality.  Under this view, desert requires no particular sentence; it merely sets the outer limits of a range of just punishments.

 

  But to those who study the demands of desert, its requirements are not so vague or flexible.  Von Hirsch, for example, notes that the principle of desert necessitates an ordinal ranking of cases [FN51]--justice requires that offenders of lesser blameworthiness receive less punishment than offenders of greater blameworthiness.  Given the finite range over which the amount of punishment can vary and the large number of distinctions commonly recognized between degrees of blameworthiness, the punishment deserved in a particular case falls into a narrow range.  The range is determined not by some special connection between that degree of blameworthiness and that amount of punishment, but by the need to distinguish a given case from the large number of other cases of distinguishable blameworthiness.  Empirical research *1443 supports this view. [FN52]  Small differences in facts often create a significant shift in shared lay perceptions of the punishment deserved. (Note that it is the amount of punishment, not the means of punishment, that is constrained by desert.  Thus, preventive concerns may properly guide the selection of a sentencing method without offending desert. [FN53])

 

 

V. The Utility of Desert

 

  The justice problems resulting from the conflict between incapacitation and desert are significant not only because doing justice is an important value in its own right--the nonconsequentialist, retributivist view--but also because doing justice can have important crime-prevention effects--the consequentialist, utilitarian argument.  As I have argued elsewhere, [FN54] the moral credibility of the criminal law, built on community perceptions that the criminal justice system distributes liability and punishment justly, gives the criminal law crime-control power.  If the criminal law has moral authority, it can stigmatize offenders and, for some, the fear of stigma will deter prohibited conduct.  More importantly, moral authority gives the criminal law persuasive power to label as morally condemnable conduct that was not previously seen as such.  That is, a criminal law with moral credibility can facilitate the internalization of norms that counsel against prohibited conduct.  It is this internalization of norms by individuals and their family and acquaintances that has the greatest effect in controlling conduct, more than threats of official liability and punishment.  Finally, a criminal law with moral authority can influence conduct by helping to shape community norms.  Norms relating to drunk driving and domestic violence, for example, have evolved in part because more severe criminal penalties and related reforms painted such conduct as more morally condemnable than previously thought.

 

  *1444 The strength of these crime-control powers of criminal law is a function of the criminal law's moral credibility.  A criminal justice system in the business of preventive detention, rather than administration of justice, can expect no more moral authority than that afforded doctors who determine whether a mentally ill person is sufficiently dangerous to be civilly committed.  Requiring the criminal justice system to distribute punishment according to predictions of future dangerousness rather than blameworthiness for past crimes can only undercut the system's moral credibility.  Citizens initially pleased by the added protection that preventive detention reforms provide, nonetheless may accurately perceive that the system is no longer in the business of doing justice.  As criminal liability is increasingly disconnected from moral blameworthiness, the criminal law can exercise less moral authority to change norms or to cause the internalization of norms.  In the long run, then, using the criminal justice system as a mechanism for preventive detention may undercut the very crime prevention goal that is offered to justify such use.

 

 

VI. Cloaking Preventive Detention as Criminal Justice

 

  It is ironic that the perversions of justice suffered in the name of prevention actually produce a seriously flawed prevention system.  These prevention difficulties arise primarily because of the perceived need to cloak preventive measures as doctrines of criminal punishment to make them appear consistent with a criminal justice system that imposes punishment.

 

  Why should this be so?  If reformers want to detain dangerous offenders, why not adopt a system that is open about its preventive detention nature and its intention to fill any preventive need remaining after criminal justice incarceration?  Most jurisdictions allow civil commitment of persons who are dangerous because of mental illness, drug dependency, or contagious disease. [FN55]  Why is there reluctance to detain preventively offenders who remain dangerous at the conclusion of their deserved criminal terms of imprisonment?

 

  The intense controversy surrounding the preventive detention legislation of the 1960s may help to explain this reluctance. [FN56]  Critics denounced *1445 the legislation as "Clockwork Orange" [FN57] and "'Alice in Wonderland' justice" in which the punishment precedes the offense [FN58] and as introducing a "police state" [FN59] and "fostering tyranny." [FN60]  Opponents described it as "intellectually dishonest," [FN61] characterized it as "one of the most tragic mistakes we as a society could make," [FN62] and feared that it "would change the complexion of American justice." [FN63]  Preventive detention was "simply not the American way." [FN64]

 

  A large part of the perceived problem with the 1960s preventive detention legislation was that it provided pretrial preventive detention.  In contrast, most current reforms provide preventive detention only after trial and conviction, an important difference. [FN65]

 

  *1446 Yet the primary criticism of pretrial preventive detention--that the sentence precedes the trial--can also be applied to the postconviction preventive detention reforms.  Detention for longer than the deserved term of imprisonment is justified as preventing predicted future crimes.  Such detention not only punishes an offense for which the detainee has not yet been convicted, but also punishes an offense that he has not yet committed.

 

  But the ability to punish the uncommitted crime, and thereby prevent it, is the genius of the current system's cloaking of preventive detention as criminal justice.  By obscuring the preventive nature of the liability and sentence, by making it appear not so entirely different from a criminal justice system of deserved punishment, the preventive detention controversy can be avoided entirely.

 

 

VII. The Practical Value of Creating Desert-Dangerousness Ambiguity

 

  The practical advantage of cloaking preventive detention as criminal justice lies in the opportunity it provides to bypass the logical restrictions on preventive detention.  First, if the justification for detention is dangerousness, then logically the government ought to be required periodically to prove the detainee's continuing dangerousness.  If the dangerousness disappears, so does the justification for detention.  However, if the detention is characterized as deserved punishment for a past offense, there is little reason to revisit the justification for the detention.  The factors relevant to determining deserved punishment may be weighed at the time of sentencing: the offender's conduct, state of mind, and capacities at the time of the offense and the resulting harm or evil.  Thus, characterizing preventive detention as deserved punishment obscures the need for periodic review.

 

  Second, if a person is detained for society's benefit rather than as deserved punishment, the conditions of detention should not be punitive.  The preventive detainee is not being punished but rather is suffering an intrusion of liberty for the benefit of society.  The mentally ill, drug-dependent, or contagious disease detainee logically ought to and often does enjoy better conditions than the person suffering punishment. [FN66]  In contrast, if confinement serves to impose deserved punishment, the offender has little justification for complaining about punitive conditions.  One of the points of imprisonment is, within the bounds of human dignity, to induce suffering.  By cloaking preventive *1447 detention as deserved punishment, the system avoids having to justify its failure to provide nonpunitive conditions of preventive detention.

 

  Third, prevention-justified restraint should logically be limited to the minimum required to ensure the community's safety.  If house arrest, an ankle bracelet, drug therapy, or other alternatives to incarceration provide adequate protection, then greater levels of restraint cannot be justified. [FN67]  No such minimum-restraint principle applies to deserved punishment. Indeed, Dan Kahan and others argue that imprisonment is a preferred form of punishment because of its expressive power of condemnation. [FN68]  Cloaking preventive detention as criminal justice, then, permits authorities to avoid demonstrating that detention is the least intrusive restraint adequate for protection.

 

  Finally, consistent with the preventive detention principle of minimum restraint, a detainee should be entitled to treatment if it can reduce the length or intrusiveness of the restraint.  No similar claim to treatment is available if the justification for incarceration is retributive.  The person incarcerated as deserved punishment has no greater claim to government-provided treatment than any other citizen.

 

  Thus, reformers benefit from all of these practical implications by cloaking preventive detention as criminal justice.  By continuing to present itself as "doing justice"--by obscuring the preventive nature of reforms with ambiguity as to their purpose--the system can provide preventive detention without the constraints that logically would attend an explicit preventive detention system.

 

 

VIII. Surreptitiously Discounting the Significance of Resulting Harm

 

  Diverting the criminal justice system from upholding justice to advancing preventive detention is not an entirely new phenomenon.  The seeds of this shift from desert to dangerousness were planted at least as early as the 1950s with the rehabilitation movement.  For example, the Model Penal Code, promulgated in 1962, generally grades inchoate offenses *1448 the same as substantive ones. [FN69]  Attempted rape has the same grade as rape, attempted arson the same as arson.  The judgment implicit in such grading clearly conflicts with the strongly held lay belief that resulting harm aggravates an offender's blameworthiness and calls for greater punishment. [FN70]  But the Code's grading approach makes sense if the goal is to maximize societal control over dangerous people.  The offender who fails to cause harm because police are able to interrupt him may be as dangerous as the offender who completes the offense.  The two are thus equal candidates for rehabilitation or, failing that, incapacitation. [FN71]  This approach is consistent with the *1449 Model Sentencing Act, which minimizes the significance of offense seriousness. [FN72]

 

  This approach--discounting the significance of resulting harm and offense seriousness in assessing punishment--became somewhat less attractive in the mid-1970s, when the limited ability of social and medical science to rehabilitate offenders became clear. [FN73]  Crime and the consequent need for criminal justice would not disappear through the power of clinical advances, as had been hoped. [FN74]  But an important step had been taken: the disconnect between criminal punishment and desert had been formally legitimized.

 

  Reformers soon realized that even if rehabilitation was unrealistic, at the very least incapacitation would prevent future crimes.  Hence, the modern ideas for reform developed--three strikes, lowering the age of eligibility for prosecution as an adult, and others.  If desert does not constrain the criminal justice system, then liability and punishment can be distributed in any way that the current crime-control utilitarian calculus suggests may reduce crime.

 

  Rather than openly recharacterize the system to reveal its nature as preventive detention, the reformers, then as now, appeared anxious to maintain the false image of a system of criminal punishment.  If the drafters believed that resulting harm should be irrelevant to grading, they could have simply eliminated all result elements from the Model Penal Code's offense definitions and defined all offenses in terms of conduct and accompanying mental state: "Engaging in conduct by which one intends to . . ." burn a building, falsify an official document, or injure another.  Why retain the result elements, implying that the Code considers resulting harm, only to negate the effect of the result elements by grading inchoate conduct the same as the completed offense?  One might speculate that the drafters saw value in maintaining the appearance, although not the spirit, of a criminal punishment system.

 

 

*1450 IX. The Preventive Detention Problems

 

  It is evident, then, that there are various ways in which the current criminal justice system surreptitiously provides preventive detention at the expense of just punishment.  Ironically, such cloaked preventive detention also seriously impedes the system's preventive effectiveness.  For example, instead of examining each offender to determine the person's actual present dangerousness, the current system uses prior criminal record as a proxy for dangerousness.  Prior record has some correlation with dangerousness and, with the assertion of the "nose-thumbing" theory, has plausible deniability as to its perverting justice. [FN75]  But prior record is only a rough approximation of actual dangerousness, and its use in preventive detention guarantees errors of both inclusion and exclusion.

 

  A scientist's ability to predict future criminality using all available data is poor; [FN76] using just the proxy of prior criminal history, a scientist's prediction is even less accurate.  It is often true that a person who has committed an offense will do so again.  But it is also frequently false--many offenders do not commit another offense. [FN77]  An explicit assessment of dangerousness would reveal that many second-time offenders are no longer dangerous, yet these offenders receive long preventive terms under three strikes statutes and criminal-history-based guidelines.  At the same time, an explicit assessment of dangerousness would reveal that many first-time offenders are dangerous; yet these offenders are not preventively detained under three strikes statutes and criminal-history-based guidelines. [FN78]

 

  *1451 Indeed, this particular cloaking device stands good prevention on its head.  Evidence suggests that criminality is highly age- related. [FN79]  Whether due to changes in testosterone levels or something else, the offending rate drops off steadily for individuals beyond their twenties.  The prior-record cloak leads us to ignore younger offenders' future crimes when they are running wild, and to begin long-term imprisonment, often life imprisonment under "three strikes," just when the natural forces of aging would often rein in the offenders.  Offenders with their criminal careers before them are not detained because they have not yet compiled their criminal resumes, whereas offenders with their criminal careers behind them are detained because they have the requisite criminal records.  Such a scheme produces a costly prevention system of prisons full of geriatric life-termers. Simultaneously, the scheme leads to ineffective prevention, because the system does little during the period in a criminal's life when the need for preventive detention is greatest.  A rational and cost-effective preventive detention system would more readily detain young offenders during their crime-prone years and release them for their crime-free older years.  Yet the need to cloak preventive detention with deserved punishment prompts the use of prior record as a substitute for actual dangerousness.

 

  An equally counterproductive aspect of the cloaked system is its mandating of fixed ("determinate") sentences immediately following a guilty verdict.  In determining the length of a deserved sentence, all of the relevant information is known at the time of sentencing--the nature of the offense and the personal culpability and capacities of the offender.  Thus, sentencing judges determining deserved punishment have little reason to impose any sentence other than a fully determinate one (that is, one that sets the actual release date) immediately after trial.  A system that instead allows a subsequent reduction of sentence, as by a parole board, undercuts deserved punishment.  Citizens become cynical that a just sentence will be undermined by early release. *1452[ FN80]  It is this cynicism that gives rise to demands for "truth in sentencing" and to the legislative response of establishing determinate terms and abolishing early release on parole. [FN81]

 

  Therefore, to maintain its justice cloak, the preventive system must follow this practice of imposing determinate sentences soon after trial.  But this practice is highly inappropriate for effective prevention.  It is difficult enough to determine a person's present dangerousness--whether he would commit an offense if released today.  It is much more difficult to predict an offender's future dangerousness--whether he would commit an offense if released at the end of the deserved punishment term in the future.  It is still more difficult, if not impossible, to predict today precisely how long the future preventive detention will need to last.  Yet that is what determinate sentencing demands: the imposition now of a fixed term that predicts preventive needs far in the future.

 

  A sentencing judge or guideline drafter is left to the grossest sort of speculation, inevitably doomed to setting either a term too long--thus unfairly detaining a nondangerous offender and wasting preventive resources--or a term too short--thus failing to provide adequate prevention.  In deciding between these two bad choices, decisionmakers commonly opt for errors of the first sort rather than the second, resulting in the recent increases in the terms of imprisonment.

 

  A rational preventive detention system would do what current civil commitment systems do: make a determination of present dangerousness in setting detention for a limited period, commonly six *1453 months, and then periodically revisit the decision to determine whether the need for detention continues. [FN82]

 

  Other inefficiencies resulting from the use of the cloak are found in the method of restraint.  A rational preventive detention system would follow a principle of minimum intrusion: a detainee would be held at the minimum level of restraint necessary for community safety.  If house arrest or regular medication would provide the same level of community safety as imprisonment, then the former choices would be preferred as less intrusive to the offender and less costly to society.  Implementing deserved punishment, in contrast, may often require a prison term to reaffirm the community's strong condemnation of the offense.  House arrest or regular medication may be unacceptable substitutes if they are perceived as trivializing the offense.  If preventive detention must operate under the cloak of criminal justice, it too often must follow the punishment preference for imprisonment even in situations in which prevention would be satisfied with less intrusive restraint.

 

  The preventive detention system hidden behind the cloak of criminal justice not only fails to protect the community efficiently but also fails to deal fairly with those being preventively detained.  As noted above, the inaccuracies created by the use of prior record as a substitute for actual dangerousness result in the unnecessary detention of a greater number of nondangerous offenders.  The inaccuracies created by the use of determinate sentences can have the same effect.  In cases in which a nonincarcerative sentence would provide adequate protection, the use of a prison term provides one more example of needless restraint.

 

  But the unfairness generated by the cloak of criminal justice extends to other aspects of the preventive detention system, such as the conditions of detention.  Punitive conditions are entirely consistent with a punishment rationale for the incarceration.  But if an offender has served the portion of his sentence justified by deserved punishment and continues to be detained for entirely preventive reasons, punitive conditions become inappropriate.

 

  Similarly, an offender being preventively detained should logically have a right to treatment, especially if such treatment can reduce the length or intrusiveness of the preventive detention--this constitutes a *1454 specialized application of the principle of minimum restraint.  If treatment can reduce the necessary individual sacrifice, [FN83] the offender ought to receive it.

 

 

X. Segregating Justice and Protection

 

  Real world problems commonly present us with conflicting interests that cannot be reconciled but can only be compromised.  The natural conflict between fair trials and a free press, for example, cannot be resolved; these competing interests must be balanced.  Each must be sacrificed to some extent to accommodate the other.  Society's interest in effective investigation of crime competes with individuals' interest in privacy, and Fourth Amendment analysis, the standard mechanism for resolving this competition, strikes a complex balance between the two.

 

  Fortunately, however, there is no need to compromise either justice or prevention to advance the other, for the conflict between justice and prevention can be avoided by simply segregating the two functions into two systems.  The first would be a criminal justice system that focused exclusively on imposing the punishment deserved for the past offense, and the second would be a post-sentence civil commitment system that considered only the protection of society from future offenses by a dangerous offender.

 

  The sticking point in this proposal is not in having a criminal justice system that is guided only by justice.  Most lay persons assume that the criminal justice system has always sought this goal.  The difficulty comes, instead, with the open acknowledgment of a system of preventive detention.

 

  There is some precedent for preventive detention.  As noted, all states currently have some form of civil commitment operating to protect society. [FN84]  Additional direct precedent exists in that many states currently have post-criminal-incarceration civil commitment of some criminal offenders, typically "sexual predators." [FN85]  Under these civil commitment systems, the government can attempt to detain an offender at the conclusion of his criminal term if the government can show continuing dangerousness. [FN86]

 

  *1455 Despite the precedent, there are understandable concerns about creating a broader system of explicit preventive detention: [FN87] the Gulag Archipelago potential for governmental abuse is real.  But if the alternative is the present system of cloaked preventive detention, the risk is worth taking.  An explicit system of post-criminal commitment would better serve both the community and potential detainees.

 

  To summarize the arguments above, under a segregated system, the community would be better off because such a system offers both more justice and increased protection from dangerous offenders.  Giving the criminal justice system a better chance of doing justice is valuable for its own sake.  It also creates greater moral credibility for the system, and thus greater long-term crime-control power.  An explicit preventive detention system offers better protection, because it can directly consider a person's present dangerousness and more accurately predict who is dangerous.  Such a system also enhances accuracy by allowing for periodic re-evaluations, in comparison with the present system's need to make a single prediction of dangerousness years in advance.  Greater accuracy leads to more detention of the dangerous, better protection, and less detention of the nondangerous, thus saving resources.

 

  A segregated system also benefits the potential detainees for many of the same reasons.  Better accuracy in prediction means less detention of nondangerous offenders.  Periodic re-evaluation leads to detention limited to periods of actual dangerousness.  Acknowledging the preventive nature of the detention also logically suggests a right to treatment, a right to nonpunitive conditions, and the application of the principle of minimum restraint, meaning greater freedom among those who are detained.

 

  Beyond the new limitations imposed on it, an open system of preventive detention ought to be preferred precisely because it is open rather than cloaked.  No one can guarantee that a legislature or court will not attempt to abuse its power.  But an open system makes it harder to abuse the system.  The openly preventive nature of the system makes it susceptible to closer scrutiny, which the present cloaked system escapes.  Instead of the current debates-- which typically reduce to disagreements about, for example, whether "three strikes" sentences are "too long"--the debate would shift to the many aspects of preventive detention that cry out for debate: What is the reliability of *1456 the predictions of dangerousness?  Is the threatened danger sufficient to justify the extent of intrusion on personal liberty?  Are there less expensive or less intrusive measures that would as effectively protect the community?  Under the current cloaked system, these issues escape examination and debate.

 

  Imagine a legislature considering an explicit preventive detention statute that would provide life preventive detention on a third conviction for a minor fraud offense, the disposition provided by the statute in Rummel.  Such legislation would be difficult to defend and would be unlikely to find support in any political quarter.  Indeed, imagine the Supreme Court's review of Rummel if Rummel were being preventively detained.  Life terms without the possibility of parole may be common and acceptable in a criminal justice system, in which horrible crimes deserve severe punishment.  But life commitment with no further dangerousness review for a property offense would be preposterous on its face in a civil preventive detention system.

 

  Some people will argue that it is simply not politically feasible in the United States today to create an explicit system of preventive detention, even one limited to dangerous felons about to be released from prison.  Less feasible, however, is political inaction in the face of recurring serious offenses that are preventable.  The inevitable pressure for protection will express itself in one form or another.  If the only choices are an open preventive detention system and a cloaked one, both the community and potential detainees ought to prefer the open system.  If there is a danger of governmental abuse of preventive detention, that danger is greatest when preventive detention is cloaked as criminal justice.

 

 

[FNa1]. Edna and Ednyfed Williams Professor of Law, Northwestern University.

 

 

[FN1]. Modern academics have become comfortable with using nondesert crime- control principles, such as deterrence and the incapacitation of dangerous people, to govern the distribution of criminal punishment.  Laypersons, however, generally do not share this perspective.  See, e.g., John M. Darley, Kevin M. Carlsmith & Paul H. Robinson, Incapacitation and Just Deserts as Motives for Punishment, 24 Law & Hum. Behav. 659, 659 (2000) (noting that empirical studies suggest that laypersons do not take account of correlates of future criminality in setting punishment, especially if protective mechanisms other than the criminal justice system are available, but that they instead look to blameworthiness); Cass R. Sunstein, David Schkade & Daniel Kahneman, Do People Want Optimal Deterrence?, 29 J. Legal Stud. 237 (2000) (discussing two experiments that suggest that people do not spontaneously think in terms of optimal deterrence, and that people would have objections to policies based on the goal of optimal deterrence); Kevin M. Carlsmith, John M. Darley & Paul H. Robinson, Why Do We Punish?  Deterrence and Just Deserts as Motives for Punishment 24 (Nov. 15, 2000) (unpublished manuscript, on file with the Harvard Law School Library) (noting empirical studies that imply that laypersons do not consider publicity or difficulty of detection in setting punishment).

 

 

[FN2]. See, e.g., 18 U.S.C. § 3559 (1994) (requiring life imprisonment on a third serious violent felony conviction); Mont. Code Ann. § 46-18-219 (1999) (requiring life imprisonment without the possibility of release after a second or third felony conviction, depending on the felonies committed); s ee generally John Clark, James Austin & D. Alan Henry, "Three Strikes and You're Out": A Review of State Legislation 9-10 (Nat'l Inst. of Justice: Research in Brief, NJC 165369, 1997) (noting that many states have expanded pre-existing repeat-offender statutes); Nat'l Conference of State Legislators, "Three Strikes" Sentencing Laws 24 (1999) (noting that between 1993 and 1999, twenty- four states and the federal government enacted "three strikes" laws and that nearly all states have some type of sentence enhancement applicable to habitual offenders).

  The protective rationale for these laws is evident in the legislative history of the federal three strikes statute.  After citing the "problem [of] a significant percentage of crimes ...committed by people who previously have committed crimes" and concluding that, to date, "the response of the criminal justice system to both violent crime and recidivism has been inadequate," the Report of the House of Representatives states that the purpose of the legislation is "to take the Nation's most dangerous recidivist criminals off the streets and imprison them for life."  H.R. Rep. No. 103-463, reprinted in H.R. 3981, 103d Cong., at 3-4 (codified at 18 U.S.C. § 3559 (1994)). Senate Majority Leader Trent Lott explained the need for the federal legislation by noting that "[t]here is no doubt that a small hardened group of criminals commit most of the violent crimes in this country" and that "[m] any of the people involved in these crimes are released again and again because of the 'revolving door' of the prison system."  139 Cong. Rec. 27,822-23 (1993).

 

 

[FN3]. Between 1992 and 1995, forty-one states passed laws making it easier to try juveniles as adults.  Melissa Sickmund, Howard N. Snyder & Eileen Poe- Yamagata, Nat'l Ctr. for Juvenile Justice, Juvenile Offenders and Victims: 1997 Update on Violence 30 (1997).  Twenty-nine states now allow the prosecution of ten-year-olds for at least one offense.  See Howard N. Snyder & Melissa Sickmund, Nat'l Ctr. for Juvenile Justice, Juvenile Offenders and Victims: A National Report 88 (1995).  Some of these states require juvenile court judges to agree to the transfer of juveniles to adult court, others leave the decision to transfer to prosecutorial discretion, and still others require the transfer for certain offenses.  See id. at 85-89; see generally Eric K. Klein, Dennis the Menace or Billy the Kid: An Analysis of the Role of Transfer to Criminal Court in Juvenile Justice, 35 Am. Crim. L. Rev. 371, 401-09 (1998) (discussing the problems that result when children are tried as adults).  According to a recent Justice Department report, "every state now has at least one provision to transfer juveniles to adult courts."  Kevin J. Strom, Profile of State Prisoners Under Age 18, 1985-97, at 1 (Bureau of Justice: Special Report, NCJ 176989, 2000).  As of 1997, twenty-eight states had statutes that automatically excluded certain types of offenders from juvenile court jurisdiction, fifteen states permitted prosecutors to file some juvenile cases in adult criminal courts directly, and forty-six states allowed juvenile court judges to send cases to adult courts at their discretion.  Id. at 2. As a result of such changes, the number of young people sent to prison rose from 18 per 1000 violent crime arrestees under age eighteen in 1985 to 33 per 1000 arrestees in 1997.  Id. at 5 tbl.4.

  Legislative histories provide further evidence of the protective rationale underlying these reforms.  The report for the 1994 California legislation, for example, explains the need for lowering the age of criminal prosecution from sixteen to fourteen by noting that "the public is legitimately concerned that crimes of violence committed by juveniles are increasing in number and in terms of the level of violence," and concluding that the legislation "is a rational response to the legitimate public desire to address what is a serious problem."  A.B. 560, 1993-1994 Leg., Reg. Sess. (Cal. 1994) (enacted).  The Congressional Research Service similarly summarizes the rationale for such state legislation: "locking up dangerous kids so that they will not commit further crimes."  Cong. Research Serv., Pub. No. 95-1152 GOV, Juveniles in the Adult Criminal Justice System: An Overview 5 (1995). Federal legislation that the House passed and that is pending in the Senate would reduce the age of presumptive adult prosecution to fourteen and would allow prosecution at thirteen for violent offenses and drug offenses.  The "Background and Need for the Legislation" section of the bill indicates that "[i]n America today no population poses a larger threat to public safety than juvenile criminals."  H.R. Rep. No. 105-86, at 14 (1997).

 

 

[FN4]. See, e.g., Nev. Rev. Stat. 193.168 (1999) (enhancing criminal penalties for felonies committed to promote criminal gang activities); Okla. Stat. Ann. tit. 21, § § 856(D)-(F) (West Supp. 2000) (creating a crime encompassing gang recruitment activities); Cal. Penal Code § 186.22(a) (West Supp. 2001) (providing special penalties for facilitating gang crime); s ee generally Bart H. Rubin, Note, Hail, Hail, The Gangs Are All Here: Why New York Should Adopt a Comprehensive Anti-Gang Statute, 66 Fordham L. Rev. 2033 (1998) (discussing the attributes of anti-gang statutes).  The California statute is part of the state's Street Terrorism and Enforcement Prevention Act, which was a response to "a state of crisis which has been caused by violent street gangs whose members threaten, terrorize, and commit a multitude of crimes against the peaceful citizens of their neighborhoods."  Cal. Penal Code § 186.21 (West Supp. 1998).

  This list of recent reforms focusing on dangerousness is not exhaustive.  Many death penalty provisions also use dangerousness explicitly as a ground for imposing the death penalty rather than life imprisonment.  See, e.g., Va. Code Ann. § 19.2-264.2 (Michie 2000); Wash. Rev. Code Ann. § 10.95.070(8) (West Supp. 1999); Wyo. Stat. Ann. § 6-2-102(h)(xi) (Michie 1999).  Sometimes lack of dangerousness is a mitigating factor in death penalty cases.  See, e.g., Md. Code Ann., Crimes and Punishments § 413(g)(7) (1996).  Occasionally criminal history is considered in sentencing instead of dangerousness.  See, e.g., Ark. Code Ann. § 5-4-604(31) (Michie 1997); Okla. Stat. tit. 21, § 701.12(1) (1991)

; Cal. Penal Code § 190.3(c) (West 1999); s ee also People v. Hawkins, 897 P.2d 574, 597 (1995) (finding that it was not error for a prosecutor to argue that the future dangerousness of the defendant was a factor weighing in favor of the death penalty).  Correctional officers are sometimes required to exclude allegedly dangerous offenders from certain release programs.  N.Y. Legis. Exec. Order 5.1 (1996).  Some shaming penalties are designed to "prevent future dangerous acts, rather than punish past action."  Art Hubacher, Every Picture Tells a Story: Is Kansas City's "John TV" Constitutional?, 46 U. Kan. L. Rev. 551, 587 (1998) (internal quotation marks omitted).

 

 

[FN5]. Federal legislation creates financial incentives for states to enact such sexual offender registration statutes.  42 U.S.C. § § 14071(g), (i). Most states have done so.  See, e.g., Nat'l Inst. of Justice, Sex Offender Community Notification 1 (Feb. 1997).

 

 

[FN6]. Washington was the first state to pass such a law.  See Wash. Rev. Code § 71.09 (1992).  Other states have since enacted similar laws.  E.g., Iowa Code § 901A.1 et seq.; Kan. Stat. Ann. § 59-29a01 (1994); Minn. Stat. § § 253B.18, .185 (1994); Wis. Stat. Ann. § 980 (West 1998).  The constitutionality of the Kansas statute was challenged in December, 1996.  See Kansas v. Hendricks, 521 U.S. 346, 350 (1997) (sustaining the act).  At that time, six states had such statutes--the other five being Arizona, California, Minnesota, Washington, and Wisconsin--and thirty-eight states, including New Jersey and New York, filed amicus briefs successfully urging the Justices to uphold the law.  Id. at 371.

  The promulgation of the Kansas statute was based on a finding that "sexually violent predators generally have anti-social personality features which are unamenable to existing mental illness treatment modalities[,] and those features render them likely to engage in sexually violent behavior[,] and that sexually violent predator' likelihood of engaging in repeat acts of predatory sexual violence is high."  Kan. Stat. Ann. § 59-29a01 (1994).

 

 

[FN7]. See, e.g., U.S. Sentencing Guidelines Manualch. 4,pt. A (1998-99);  id. at ch. 5, pt. A (providing guideline sentences as a function of "Offense Level" and "Criminal History Category"); Ariz. Rev. Stat. § 16-90- 801(b)(1) (1995); Del. Code Ann. tit. 11, § 6580(c)(1) (1995); Wash. Rev. Code Ann. § 9.94A.010(1) (West 1985).

  The rationale for heavy reliance upon criminal history in sentencing guidelines is its effectiveness in incapacitating dangerous offenders.  As the Guidelines Manual of the United States Sentencing Commission explains, "the specific factors included in [the calculation of the Criminal History Category] are consistent with the extant empirical research assessing correlates of recidivism and patterns of career criminal behavior."  U.S. Sentencing Guidelines Manual 289 (1999).

 

 

[FN8]. See supra notes 2-7.

 

 

[FN9]. Webster's New World College Dictionary 1180 (2d ed. 1959) (emphasis added).

 

 

[FN10]. Id. at 372.

 

 

[FN11]. See infra pp. 1437-41.

 

 

[FN12]. See Paul H. Robinson, Foreword: The Criminal-Civil Distinction and Dangerous Blameless Offenders, 83 J. Crim. L. & Criminology 693, 711-14 & nn.57-68 (1993) (listing related authorities).

 

 

[FN13]. Bureau of Justice Statistics, U.S. Dep't of Justice, Sourcebook of Criminal Justice Statistics--1998, at 260 tbl.3.114 (Kathleen Maguire & Ann L. Pastore eds., 1999) [hereinafter Sourcebook].

 

 

[FN14]. Id.

 

 

[FN15]. Id.

 

 

[FN16]. This takes into account a 7% decrease announced in preliminary data for 1999, FBI, U.S. Dep't of Justice, Preliminary Annual Uniform Crime Report-- 1999, at 1 (1999), producing a total eight-year decline of 30.1%.  See also Gary LaFree, Social Institutions and the Crime "Bust" of the 1990s, 88 J. Crim. L. & Criminology 1325, 1340-43, fig.3 (1998) (showing that although the rates of rape, robbery, and aggravated assault are decreasing, further significant declines are needed to reach the 1950' crime level).

 

 

[FN17]. A 1998 study reveals that 49% of the residents of the largest American cities avoided going out at night and 18% installed security systems. Steven K. Smith, Greg W. Steadman, Todd D. Minton & Meg Townsend, Criminal Victimization and Perceptions of Community Safety in 12 Cities, 1998, at 20-21 tbls.24 & 25 (NJJ 173940 1999), available at http://www.ojp.usdoj.gov/bjs/; see also Sourcebook, supra note 13, at 121 tbls.2.41 & 2.42.

 

 

[FN18]. Public opinion survey results rank crime as the second most serious problem facing the country, after "ethics, moral, family decline."  Sourcebook, supra note 13, at 96 tbl.2.1.  A July 1999 survey found that 52% of the respondents believed that crime had risen in the previous year.  Id. at 116 tbl.2.33.  Public concern about crime rates continues in many communities, even after the crime drop.  See Yvette Caig, Crime Down 10 Percent in Fort Worth in 1997, Fort Worth Star Telegram, Jan. 13, 1998, at 1 (Metro) ("Although the release of the year-end crime statistics is cause for applause, the news comes amid heightened concern about a string of homicides since the new year began."); Bob von Sternberg, Clearly, Life Is Good in Minnesota, Star Trib. (Minneapolis-St. Paul), Jan. 27, 2000, at 1A ("The single national issue that concerns the largest number of Minnesotans is crime, cited by 13 percent of those polled--ata time when the rates of violent crime are plummeting.").

 

 

[FN19]. Juveniles as a group are more dangerous persons today than a decade ago.  For example, in 1976, juveniles between the ages of fourteen and seventeen accounted for 10.6 offenders per 100,000 in terms of murders and non- negligent homicides; by 1995, the figure had more than doubled to 23.0 offenders per 100,000.  Sourcebook, supra note 13, at 340 tbl.3.132.

 

 

[FN20]. See, e.g., Arthur T. Jersild, Charles W. Telford & Jane M. Sawrey, Child Psychology 157 (7th ed. 1975); Glanville Williams, Criminal Law: The General Part 818 (2d ed. 1961) ("[I]t is only in a special sense that the child member of a delinquent gang can be said to know shoplifting or receiving stolen goods to be 'wrong.'  He knows that such conduct is frowned upon by the police, and perhaps by his parents; but he does not himself feel it to be wrong.").

 

 

[FN21]. Insanity, involuntary intoxication, and duress excuse a violator who has caused the harm or evil prohibited by an offense but who lacks the capacity to appreciate the wrongfulness of her conduct or to conform her conduct to the requirements of law.  See, e.g., Model Penal Code § § 2.08(4), 2.09(1), 4.01(1) (1962).  Because a person's lack of maturity can cause these same excusing conditions, an immaturity defense logically should be part of the criminal law's system of excuses.  For a general discussion of the conceptual analogy among excuses, see Paul H. Robinson, Criminal Law 477-94 (1997).

 

 

[FN22]. Instead, states provide for the transfer of jurisdiction to juvenile court for all defendants below a given age.  See generally 2 Paul H. Robinson, Criminal Law Defenses § 175 (1984 & Supp. 1998).

 

 

[FN23]. 445 U.S. 263 (1980).  The three prior fraud convictions that qualified Rummel for a life term involved a total of $229.11.  Id. at 265- 66.

 

 

[FN24]. Id. at 284-85.

 

 

[FN25]. Under three strikes statutes, for example, the criminal history often quadruples the sentence that would be imposed for the identical offense by the identical offender with no criminal history.  A twenty-five year old offender committing a felony that normally carries a ten-year sentence, for which less than ten years ordinarily would be served, can get mandatory life imprisonment without the possibility of parole, which may mean a sentence of forty-five years or more.  See Clark, Austin & Henry, supra note 2, at 7-9, exhibit 9; see also, e.g., Del. Code Ann. tit. 11, § 4214 (1995) (stating that a third felony conviction carries a life sentence for violations including kidnapping and aggravated robbery); 720 Ill. Comp. Stat. Ann. 5/33B-1 (West Supp. 1998) (stating that a third felony conviction carries a life sentence for violations including Class X felonies and criminal sexual assault).

  Even the less drastic habitual-offender statutes, which have been in use for some time, can have a substantial effect on the amount of punishment inflicted on an offender.  For example, the Model Penal Code provision, which is the structural model for many habitual-offender statutes, allows an "extended term of imprisonment" that essentially doubles the maximum authorized sentence for a repeat offender: the maximum sentence is increased from five years to ten years for a third-degree felony, from ten years to twenty years for a second-degree felony, and from twenty years to life imprisonment for a first-degree felony.  Model Penal Code § § 6.07, 7.03(3)- (4) (1962).

  Even in the absence of either three strikes or habitual-offender statutes, sentencing guidelines that tie a sentence in part to an offender's criminal history provide for a similar increase in punishment for dangerousness. Under the Guidelines Manual, for example, an individual who commits a level 10 offense receives a sentence of 6 to 12 months if he has no criminal record but receives a sentence of 24 to 30 months if he has a significant record; an individual who commits a level 19 offense receives a sentence of 30 to 37 months if he has no criminal record but receives a sentence of 63 to 78 months if he has a significant record; and an individual who commits a level 37 offense receives a sentence of 210 to 262 months if he has no criminal record but receives a sentence of 360 to life imprisonment if he has a significant record.  See U.S. Sentencing Guidelines Manual ch. 5, pt. A, sentencing tbl. (1997).

 

 

[FN26]. Andrew von Hirsch, Doing Justice: The Choice of Punishments 85  (1976).  But von Hirsch later withdrew much of his support for such a theory and relied instead on a different theory suggesting that a discount in punishment may be appropriate for first-time offenders.  See Andrew von Hirsch, Past or Future Crimes: Deservedness and Dangerousness in the Sentencing of Criminals 78-85 (1985) [hereinafter von Hirsch, Past or Future Crimes].

 

 

[FN27]. If a nose-thumbing theory does not work, one might argue that additional punishment can serve as a sort of lingering suspended sentence from the past offenses.  But that argument is inconsistent with the facts.  None of the reforms that look to past criminal history consider whether the sentences for past offenses were carried out in full or suspended.  In fact, none of the reforms consider prior sentences at all.  Two offenders with identical criminal histories who commit identical present offenses will receive identical treatment under recent reforms even if one offender has served many long prison terms and the other has served none.  Clearly, it is future dangerousness that is the concern; under the theory of the reforms, identical criminal history suggests identical dangerousness.

 

 

[FN28]. Von Hirsch agrees with this conclusion.  See von Hirsch, Past or Future Crimes, supra note 26, at 131-36 (1985).

 

 

[FN29]. See Clark, Austin & Henry, supra note 2, at 7-9, exhibit 9  (providing a table that lists the various offenses that states include in habitual-offender sentencing schemes).

 

 

[FN30]. See, e.g., Wash. Rev. Code § § 9.94A.030(23), 9.94A.030(27),  9.94.120(4) (1985); Md. Ann. Code art. 27, § 643B (1996).

 

 

[FN31]. Model Penal Code § 5.05(2) (1962).

 

 

[FN32]. For examples of such HIV-mistaken-effect attacks, see State v. Smith, 621 A.2d 493 (N.J. Super. Ct. App. Div. 1993).

 

 

[FN33]. Assume, for example, that the despised father subsequently dies from natural causes and the son is therefore no longer dangerous.  Alternatively, consider the son who poisons a hated mother with watermelon juice, falsely believing that such juice can cause a fatal allergic reaction.  The son is as blameworthy as an attempted murderer.  Even if watermelon allergies do not exist, and even if the despised mother has since died of natural causes so that the son is admittedly no longer a danger to anyone, allowing a defense in this hypothetical is a failure of justice that the Model Penal Code provision promotes.

 

 

[FN34]. In a 1981 study, forty-five percent of judges did not think that  "just deserts" was important.  See S. Rep. No. 98-225, at 41 n.18 (1983) (citing INSLAW/Yankelovich, Skelly & White, Inc., Federal Sentencing III-4 (1981)); see also Anthony Partridge & William B. Eldridge, Federal Judicial Center, The Second Circuit Sentencing Study: A Report to the Judges of the Second Circuit (1974).

 

 

[FN35]. The Nazi official may now be a productive member of society, presenting no need for incapacitation or rehabilitation.  Yet even if the circumstances of the Third Reich will never arise again, the offender is blameworthy, and the desert principle insists he receive punishment for his past offense.

 

 

[FN36]. Other utilitarian distributive principles also might impose liability.  Punishing such an offender reinforces the general prohibition against such offenses; that is, it serves a general deterrent purpose. Moreover, such an offender may need rehabilitation.

 

 

[FN37]. Consider also the practitioner of "voodoo" who tries to kill by placing a spell or by sticking needles into a doll.

 

 

[FN38]. It is this utilitarian reasoning that leads the Model Penal Code to provide a defense for an "inherently unlikely" attempt.  See supra p. 1437.

 

 

[FN39]. See, e.g., Don M. Gottfredson, Leslie T. Wilkins & Peter B. Hoffman, Guidelines for Parole and Sentencing 41-67 (1978) (including employment history in a list of nine factors best predicting future criminality); Peter W. Greenwood with Allan Abrahamse, Selective Incapacitation 105-06 (1982) (noting that employment history "is somewhat associated with ...offense rates").

 

 

[FN40]. Researchers have found age to be an effective predictor of future violence.  See, e.g., Joseph J. Cocozza & Henry J. Steadman, Some Refinements in the Measurement and Prediction of Dangerous Behavior, 131 Am. J. Psychiatry 1012 (1974).  Various aspects of an offender's family situation are also of predictive value.  See, e.g., Alfred Blumstein, David P. Farrington & Soumyo Moitra, Delinquency Careers: Innocents, Desisters, and Persisters, in 6 Crime and Justice: An Annual Review of Research 187, 198 (Michael Tonry & Norval Morris eds., 1985).  Greenwood identifies several other factors of predictive value: prior convictions of the instant offense type; incarceration for more than half of the preceding two years; conviction before the age of sixteen; time served in a state juvenile facility; drug use during the preceding two years; and employment during less than half of the preceding two years. Greenwood with Abrahamse, supra note 39, at 50.

 

 

[FN41]. One study suggests that rapists may be distinguished from nonrapists based on their penile erection response to certain stimuli.  Gene G. Abel, David H. Barlow, Edward B. Blanchard & Donald Guild, The Components of Rapist' Sexual Arousal, 34 Archives Gen. Psychiatry 895, 895 (1977).  If the predictive technique were sufficiently refined, a pure incapacitation distributive principle might find it an appropriate basis for distributing criminal liability.  See generally Sanford H. Kadish, The Decline of Innocence, 26 Cambridge L.J. 273 (1968).

  A similar analysis applies to the utilitarian crime-control principles of deterrence and rehabilitation.  If deterrence were the sole distributive principle for liability and punishment, potential offender' perceptions of the probability of apprehension would be highly relevant in determining the sanction needed to deter effectively.  Offenses with a perceived low probability of apprehension would require more severe punishment to compensate for that perception.

  Potential offenders' perceptions of the likelihood of conviction also can prove powerful.  Thus, it may be useful to drop those requirements of liability that prosecutors find difficult to prove, such as culpability requirements (for example, intention and recklessness), because such reforms would increase the perceived likelihood of conviction.

  A pure deterrence principle also would base liability on how widely the sanction is known.  Just as an advertising executive willingly pays more for an advertisement that will reach more people, the cost of a higher criminal sanction is a good investment if its deterrent message will reach more people.  Thus, greater news coverage of a case would aggravate the grade of the offense, lengthening the imposed prison term.

  If rehabilitation governed the distribution of punishment, a person's amenability to treatment would act as the primary variable.  Recall that fully indeterminate sentences were recently imposed in furtherance of the rehabilitative purpose.  The length of the sentence was determined by the length of time necessary for rehabilitation of the offender--a factor eluding predetermination.  The offender's term of imprisonment continued indefinitely until the offender was rehabilitated.  Thus, a minor offense might call for long-term incarceration and a serious offense might call for no incarceration, if long-term treatment were required to avoid recidivism in the first case and no treatment were required in the second.

  Indeed, with a pure rehabilitation principle, as with incapacitation, there is little reason to wait for an offense to occur.  Screening of the population would determine those people likely to commit future offenses absent rehabilitative treatment, followed by the imposition of liability and sanctions to compel the required treatment and thereby to avoid the anticipated crime.

 

 

[FN42]. Nor does a person deserve more punishment because his offense is one with a perceived low apprehension or conviction rate or because of aggressive news coverage--factors that increase the length of incarceration under a deterrence principle.  Similarly, a person does not deserve more punishment because the kind of rehabilitation program that might help him takes longer than other programs--a factor that increases the length of incarceration under a rehabilitative principle.

 

 

[FN43]. Council of Judges of the National Council on Crime and Delinquency, Model Sentencing Act: Second Edition, 18 Crime and Delinquency 335, 341 (1972).  There is disagreement as to whether a deterrence principle generates liability proportional to the seriousness of the offense.  Compare, e.g., Ernest van den Haag, Punishment as a Device for Controlling the Crime Rate, 33 Rutgers L. Rev. 706, 713 (1981) (suggesting that deterrence calls for proportionality), with Alan H. Goldman, Beyond the Deterrence Theory: Comments on van den Haag's "Punishment as a Device for Controlling the Crime Rate", 33 Rutgers L. Rev. 721 (1981) (criticizing van den Haag's theory for its asymmetry and for its failure to allow punishment for all who deserve it).

 

 

[FN44]. Model Penal Code § 1.02(2) (1962).

 

 

[FN45]. The Commentary explains the Code's rationale:

  The section is drafted in the view that sentencing and treatment policy should serve the end of crime prevention.  It does not undertake, however, to state a fixed priority among the means to such prevention, i.e., the deterrence of potential criminals and the incapacitation and correction of the individual offender.  These are all proper goals to be pursued in social action with respect to the offender, one or another of which may call for the larger emphasis in a particular context or situation.  What the Code seeks is the just harmonizing of these subordinate objectives, rather than the concentrationon some single target of this kind.  It is also recognized that not even crime prevention can be said to be the only end involved.  The correction and rehabilitation of offenders is a social value in itself, as well as a preventive instrument.  Basic considerations of justice demand, moreover, that penal law safeguard offenders against excessive, disproportionate or arbitrary punishment, that it afford fair warning of the nature of the sentences that may be imposed upon conviction and that differences among offenders be reflected in the just individualization of their treatment.

Model Penal Code § 1.02 cmt. at 2 (Tentative Draft No. 2, 1954).

 

 

[FN46]. Stanley A. Cohen, An Introduction to the Theory, Justifications and Modern Manifestations of Criminal Punishment, 27 McGill L.J. 73, 81 (1981).

 

 

[FN47]. Report of the Canadian Committee on Corrections--Toward Unity: Criminal Justice and Corrections 188 (1969) (commonly referred to as the Ouimet Report).

 

 

[FN48]. Cohen, supra note 46, at 73.

 

 

[FN49]. See generally Paul H. Robinson, Hybrid Principles for the Distribution of Criminal Sanctions, 82 Nw. U. L. Rev. 19 (1987); Paul H. Robinson, Why Does the Criminal Law Care What the Lay Person Thinks Is Just? Coercive Versus Normative Crime Control, 86 Va. L. Rev. 1939 (2000).

 

 

[FN50]. See Norval Morris, The Future of Imprisonment 73-76 (1974); Norval Morris & Marc Miller, Predictions of Dangerousness, in 6 Crime and Justice: An Annual Review of Research 1, 35 (1985).

 

 

[FN51]. See von Hirsch, Past or Future Crimes, supra note 26, at 40.

 

 

[FN52]. See, e.g., Paul H. Robinson & John M. Darley, Justice, Liability, and Blame: Community Views and the Criminal Law 229-71, 273 (1995) (reporting empirical studies of lay intuitions on a broad range of substantive criminal law issues, the results of which indicate substantial agreement among subjects).  Presented with a dozen scenarios describing similar but different cases, subjects in the studies generally agreed on the rank and even relative relation of the cases as to the amount of punishment deserved.  Some of the subjects imposed generally longer sentences, some generally shorter sentences, but both harsh and lenient sentencers typically agreed on the pattern of relative blameworthiness among the cases, even in instances in which the factual differences among the cases were slight.  Further, the agreement among subjects extended across gender, education, income, political affiliation, and most other significant demographic variables.  See id. at 223, 226.

 

 

[FN53]. Paul H. Robinson, Desert, Crime Control, Disparity, and Units of Punishment, in Penal Theory and Practice: Tradition and Innovation in Criminal Justice 93-105 (Antony Duff, Sandra Marshall, Rebecca Emerson Dobash & Russell P. Dobash eds., 1994).

 

 

[FN54]. See generally Paul H. Robinson & John M. Darley, The Utility of Desert, 91 Nw. U. L. Rev. 453 (1997).

 

 

[FN55]. See Robinson, supra note 12, at 711-14 & nn.57-68.

 

 

[FN56]. For a chronological list of preventive detention enactments, see Barbara Gottlieb, The Pretrial Processing of "Dangerous" Defendants: A Comparative Analysis of State Laws (Nat'l Inst. of Justice Study, 1984), reprinted in Report on Bail Reform Act of 1984, H.R. Rep. No. 98-121, app. A, at 90.  The first statutes appeared in Alaska and Delaware in 1967, Maryland and South Carolina in 1969, Vermont in 1967 and 1969, and the District of Columbia in 1970.  Despite the constitutional approval of pretrial preventive detention, United States v. Edwards, 430 A.2d 1321, 1343 (D.C. 1981), cert. denied, 455 U.S. 1022 (1982), many jurisdictions have refused to enact such a system.  The twenty-four states authorizing pretrial detention are listed in Gottlieb, supra, at 76.

 

 

[FN57]. Glen Elsasser, U.S. Defends Pretrial Jailings of Suspects Seen as Dangerous, Chi. Trib., Jan. 22, 1987, § 1, at 16 (quoting critics).

 

 

[FN58]. Elder Witt, Preventive-Detention Statute Is Upheld as Constitutional, Cong. Q. Wkly. Rep., May 30, 1987, at 1141 (quoting Professor Alan Dershowitz).

 

 

[FN59]. E.g., Ethan Bronner, Court Upholds Pretrial Jailing, Boston Globe, May 27, 1987, at 1 (quoting Justice Marshall); Gotti Being Held Under Tough Bail Reform Act, Las Vegas Rev. J., Dec. 17, 1990, at A7; Caryle Murphy, Law Allowing Denial of Bail Provokes Debate on Rights, Wash. Post, Aug. 1, 1986, at A1 (quoting Judge Jon O. Newman); Herb Robinson, Judge's Dilemma Delicate Balance in Bail Rulings, Seattle Times, May 21, 1986, at A14.

 

 

[FN60]. James J. Kilpatrick, A Cautious "Yes" on Preventive Detention, Seattle Times, June 3, 1987, at A12 (citing Justice Marshall).

 

 

[FN61]. Edwin M. Yoder Jr., A Gross Judicial Assault on Personal Liberties, Houston Chron., May 31, 1987, at 2.

 

 

[FN62]. Say 'No' to No Bail, Record, N. N.J., Nov. 19, 1986, at A22 (quoting New Jersey Senate President John Russo, a former prosecutor).

 

 

[FN63]. Richard Lacayo, First the Sentence, Then the Trial, Time, June 8, 1987, at 69.

 

 

[FN64]. Alan Dershowitz, 'Wonderland' Court Trashes Our Safeguards, Chi. Sun-Times, June 7, 1987, at 11.

 

 

[FN65]. Not all of the dangerousness reforms can claim this distinction.  If dangerousness rationales set the minimum conduct requirements for criminal liability--as in expanding the scope of attempt and conspiracy to include persons thought to be dangerous--the system encompasses people who would not otherwise have been judged sufficiently blameworthy to merit criminal conviction, but who now will be criminally convicted and detained.  It is true that detention requires proof of certain past conduct, not just the prediction of future conduct, but if the past conduct is not itself condemnable but is taken as a predictor of future criminality, the distinction between pretrial preventive detention and the new "criminal conduct" is less clear.

  Indeed, in some respects, one can argue that pretrial preventive detention is less objectionable than prevention-based expansions of criminal liability and punishment.  The data suggest that after pretrial preventive detention, the detainee is almost always convicted of an offense.  One would expect as much because the strength of the prosecution's case is one of the factors the judge must consider in determining whether to permit pretrial preventive detention.  For example, under the federal pretrial preventive detention scheme, the judge must consider not only the offense charged and the defendant's dangerousness, but also "the weight of the evidence against the person."  18 U.S.C. § 3142(g)(2) (1984).  At sentencing, the convicted offender receives credit toward the sentence for the time served during pretrial detention.  In the end, then, a convicted offender is not materially harmed by the pretrial detention (although there may still be a risk of injustice and a breach of traditional due process).  At least theoretically, an offender serves only the time deserved.  The person caught in the prevention-based expansion of liability and punishment, in contrast, is subject to post-trial detention based solely on his predicted dangerousness.

 

 

[FN66]. See, e.g., Glanville Williams, Criminal Law: The General Part 829- 36 (2d ed. 1961) (discussing the practice and rationale for nonpunitive conditions of detention for civilly committed juvenile delinquents).

 

 

[FN67]. See generally Norval Morris & Michael Tonry, Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System 3, 176-220 (1990) (discussing a full range of sanctions between imprisonment and probation).

 

 

[FN68]. See, e.g., Dan M. Kahan, Social Influence, Social Meaning, and Deterrence, 83 Va. L. Rev. 349, 362-63, 384 (1997) (suggesting that the form of punishment conveys a social meaning: "Imprisonment is an extraordinarily potent gesture of moral disapproval; because of the symbolic importance of individual liberty in American culture, there is never a doubt that society means to condemn someone when it takes that person's freedom away."); Dan M. Kahan, What Do Alternative Sanctions Mean?, 63 U. Chi. L. Rev. 591, 594-605 (1996) (discussing the "expressive dimension" of punishment and the significance of imprisonment for moral condemnation).  Of course, the more incarceration is used for preventive detention rather than deserved punishment for a past offense, the less clearly imprisonment will convey condemnation of a past wrong.

 

 

[FN69]. Model Penal Code § 5.05(1) (1962).  The drafters explain:

  The theory of this grading system may be stated simply.  To the extent that sentencing depends upon the antisocial disposition of the actor and the demonstrated need for a corrective sanction, there is likely to be little difference in the gravity of the required measures depending on the consummation or the failure of the plan.

Model Penal Code § 5.05(1) cmt. at 490 (1985).  The drafters exempt inchoate offenses to commit a first-degree felony because of the overriding deterrent purpose.  Id. at 489-90.

 

 

[FN70]. See Robinson & Darley, supra note 52, at 14-28, 33-42 (1995).

 

 

[FN71]. As the drafters explain, "[t]he primary purpose of punishing attempts is to neutralize dangerous individuals."  Model Penal Code § 5.01 cmt. at 323 (1985); see also id. at 298 (stating that "the proper focus of attention is the actor's disposition"); id. at 331 (advocating the broadening of liability to permit the earlier apprehension of dangerous persons).

  For another example of the Model Penal Code's concern with dangerousness, consider § 1.02 of the Code, which includes as one of "[t]he general purposes of the provisions governing the definition of offenses: ...[t]o subject to public control persons whose conduct indicates that they are disposed to commit crimes."  Model Penal Code § 1.02(1)(b) (1962).

  In their definitions of offenses, the drafters reveal a shift in focus from desert to dangerousness.  For example, the Model Penal Code alters the common law definition of attempt from requiring that a person come within some proximity of completing an offense--the so-called "proximity" test, of which the common law had several variations--to requiring only that a person have taken a "substantial step" toward the offense.  It no longer matters whether the actor actually came close to committing a specific offense.  All that is relevant is that the actor at some point externalized his intention to commit an offense, thereby revealing his dangerousness.  The proximity tests were rejected because they focused on the nature of the actor's conduct rather than on "the proper focus of attention ...the actor's disposition [to commit a crime]."  Id. § 5.01 cmt. at 298.

  Similar substantive criminal law reformulations abound in the Code.  The Code's definition of conspiracy was altered so that it no longer requires that the actor in fact agree with another that one of them will commit an offense, but requires only that the actor thought he had agreed with another.  Compare, e.g., Archbold v. State, 397 N.E.2d 1071, 1073 (Ind. Ct. App. 1979) (requiring actual agreement), with People v. Schwimmer, 411 N.Y.S.2d 922, 928 (1978) (holding that a defendant can be found guilty of conspiracy even though the coconspirators did not have the requisite mens rea).  See generally Paul H. Robinson, Criminal Law 648 (1997) (discussing the bilateral agreement required at common law for conspiracy and the unilateral requirement of modern codes).  This change logically follows from the drafters' view that conspiracy's purpose is "as a basis ...for the corrective treatment of persons who reveal that they are disposed to criminality."  Model Penal Code § 5.03 cmt. at 97 (Tentative Draft No. 10, 1960).  There may be no real danger of an offense being committed if an actor only mistakenly believes he has conspired with another (an undercover officer, for example), but his agreement shows that he is the type of person who would be willing to join in such an agreement.  Similarly, the Code drops the common law's unconvictable perpetrator defense to complicity; it is no longer relevant that the perpetrator had no intention of committing the crime.  The accomplice is equally dangerous whether the perpetrator actually intended to commit the crime or not.  Still another example is the Code's reformulation of the objective requirements of complicity that allows liability for the full substantive offense on proof that a person tried but failed to assist.  Model Penal Code § 2.06(3)(a)(ii) (1962) (assigning liability when an individual "aids or agrees or attempts to aid").  That the failed assister in fact contributed nothing to the offense does not reduce his display of dangerousness in trying to assist.

 

 

[FN72]. See supra p. 1441.

 

 

[FN73]. By 1974 Robert Martinson concluded in his survey article: "I am bound to say that these data, involving over two hundred studies and hundreds of thousands of individuals as they do, are the best available and give us very little reason to hope that we have in fact found a sure way of reducing recidivism through rehabilitation."  Robert Martinson, What Works?--Questions and Answers About Prison Reform, 35 Pub. Int. 22, 49 (1974).

 

 

[FN74]. See, e.g., Council of Judges, Nat'l Council on Crime & Delinquency, Model Sentencing Act (2d ed. 1972) art. 1, § 1, reprinted in 18 Crime and Delinquency, 335, 344 (1972) ("The purpose of penal codes and sentencing is public protection.  Sentences should not be based upon revenge and retribution.").

 

 

[FN75]. See supra pp. 1435-36.

 

 

[FN76]. For a review of the available studies, see Stephen J. Morse,  Blame and Danger: An Essay on Preventive Detention, 76 B.U. L. Rev. 113, 126 n.39 (1996).  Morse concludes that "the ability of mental health professionals to predict future violence among mental patients may be better than chance, but it is still highly inaccurate, especially if these professionals are attempting to use clinical methods to predict serious violence."Id. at 126.

 

 

[FN77]. See Office of the Legislative Auditor, State of Minnesota, Recidivism of Adult Felons, at http:// www.auditor.leg.state.mn.us/ped/1997/pe9701.htm (reporting that 55% of all felony offenders in the study were not convicted of a subsequent offense during three years following their initial arrest and finding that homicide offenders had one of the lowest recidivism rates); see also Bureau of Justice Statistics Special Report, Recidivism of Prisoners Released in 1983 (1989).

 

 

[FN78]. The chronic spouse abuser who turns to obsessive violence when the battered spouse leaves may have no criminal history--battering spouses are often able to persuade their victims not to press criminal charges--yet the abuser may present a clear and immediate danger.  Similarly, a stalking and threat offense, depending on its circumstances, may suggest a high risk of serious danger.  See, e.g., John Douglas & Mark Olshaker, Obsession 266 (1998).  Yet most such offenses would not trigger the dangerousness add-on provisions that are components of recent reforms.  That is, even if the circumstances and nature of the offense suggest a life-threatening level of violence, a system that looks to criminal history rather than to dangerousness will have no grounds to detain the perpetrator.

 

 

[FN79]. For example, only 15% of those arrested for crime in 1994 were in their forties or older, although that age group made up 40% of the U.S. population.  Bureau of Justice Statistics, Sourcebook of Criminal Justice Statistics--1995, at 397 tbl.4.4 (1996).  In contrast, persons in their thirties made up 25.3% of arrests but accounted for only 16.9% of the population.  Id.  Persons between the ages nineteen and twenty-nine made up 37.1% of arrests but only 15.9% of the population.  Id.  Homicide arrest rates suggest an even greater drop-off in criminality with age in 1993: 11.9 of 100,000 males in the thirty-five to forty-four age bracket were arrested for homicide.  Id. at 423 tbl.4.18.  Of those aged twenty-five to twenty-nine, the rate was more than two and a half times higher, 30.0 per 100,000.  Id. For those between twenty-one and twenty-four, the rate was almost five times higher, 56.8.  Id.  Of those between eighteen and twenty, the rate was almost eight times higher, 91.3.  The trend of the last several decades has been toward even less criminality by middle-aged persons.  In 1970, the homicide arrest rate for males between thirty-five and forty-four was two- thirds higher than it was in 1993--19.5 per 100,000 versus 11.9 per 100,000. Id.  Yet the current reforms will detain a greater number of middle-aged offenders for a longer period of time.

 

 

[FN80]. The public probably infers that early release is common from the plethora of well-publicized anecdotes in which predictably dangerous criminals are released earlier than their sentences dictate, only to quickly return to their criminal ways.  See, e.g., James Wootton, Truth in Sentencing--Why States Should Make Violent Criminals Do Their Time, 20 U. Dayton L. Rev. 779, 779-80 (1995) (citing the case of a nun murdered in 1993 by a convicted killer who had been sentenced to eighteen to twenty years in 1979, subsequently escaped and after his eventual recapture was released on parole in 1991); Editorial, Honesty at Prison Gate, Grand Rapids Press, July 12, 1998, at D2; Richard P. Jones, Thompson Signs Bill Ending Parole: State Lawmaker Calls for 3,000 New Prison Beds To Handle Inmate Increase, Milwaukee J. Sentinel, June 16, 1998 (reporting that when the parole bill was signed, 58% of all new prisoners received sentences of four years or less; with parole, they were released in two years or less); Time To Face the Truth on Prisons, Chi. Trib., Aug. 28, 1995, § 1, at 14 (describing the public as "continually outraged by stories of violent offenders who serve half their time and commit other heinous acts when released").

 

 

[FN81]. One of the prime motivations for the federal Sentencing Reform Act of 1984, which among other things abolished the United States Parole Commission, was an attempt to reestablish credibility with an emphasis on "truth in sentencing" that determinate sentences bring.  See, e.g., U.S. Sentencing Comm'n, Sentencing Guidelines and Policy Statements 1.2 (1987) (stating that "Congress first sought honesty in sentencing... to avoid the confusion and implicit deception" arising out of the then-extant indeterminate sentencing system).

 

 

[FN82]. See, e.g., Idaho Code § 66-337(a) (Michie 2000) (requiring department directors to examine a patient's need for commitment at the end of the first ninety days and every one hundred twenty days thereafter); R.I. Gen. Laws § 40.1-5.3-4(f) (1997) (permitting courts to commit dangerous persons, but requiring courts to review such orders every six months); S.D. Codified Laws Ann. § 27A-10-14 (Michie 2000) (requiring a board to review a patient who has been committed for mental illness at least once every six months for the first year and at least once every twelve months thereafter).

 

 

[FN83]. For example, a number of encouraging studies have recently suggested that comprehensive treatment of pedophilia has a ninety percent or better success rate.  See Robert E. Freeman-Longo, Reducing Sexual Abuse in America: Legislating Tougher Laws or Public Education and Prevention, 23 New Eng. J. on Crim. & Civ. Confinement 303, 323 (1997).

 

 

[FN84]. See supra p. 1444.

 

 

[FN85]. See supra p. 1431.

 

 

[FN86]. Unlike the proposal made here, there is no indication that the current "sexual predator" legislation excludes reliance on dangerousness in setting the criminal commitment.  It is less a segregated system than a cloaked system followed by a preventive detention system.

 

 

[FN87]. The post-sentence civil commitment proposal might be held unconstitutional.  Although the permissible scope of civil commitment has recently been expanded slightly, it still appears to require not only a finding of dangerousness, but also some additional factor, such as mental abnormality.  Kansas v. Hendricks, 521 U.S. 346, 358 (1997) (citing Heller v. Doe, 509 U.S. 312, 314-15 (1993); and Allen v. Illinois, 478 U.S. 364, 366 (1986)).  But this would be an odd result: barring the civil commitment with periodic review of dangerousness of a dangerous murderer, but permitting, as the Court did in Rummel, the life imprisonment without parole of a petty fraud offender.