Intermediate Sanctions in Sentencing Guidelines . MENU TITLE: Intermediate Sanctions in Sentencing Guidelines Series: NIJ Issues and Practices Published: May 1997 101 pages 223,747 bytes ------------------------- Figures, charts, forms, and tables are not included in this ASCII plain-text file. To view this document in its entirety, download the Adobe Acrobat graphic file available from this Web site or order a print copy from NCJRS at 800-851-3420. ------------------------ U.S. Department of Justice Office of Justice Programs 633 Indiana Avenue N.W. Washington DC 20531 Janet Reno Attorney General U.S. Department of Justice John C. Dwyer Acting Associate Attorney General Laurie Robinson Assistant Attorney General Jeremy Travis Director, National Institute of Justice Nancy E. Gist Director, Bureau of Justice Assistance ------------------------------ Justice Information Center World Wide Web Site http://www.ncjrs.org ------------------------------ U.S. Department of Justice Office of Justice Programs National Institute of Justice ------------------------------ Intermediate Sanctions in Sentencing Guidelines by Michael Tonry Subcontractor to: Abt Associates Inc. May 1997 ------------------------------ Issues and Practices in Criminal Justice is a publication series of the National Institute of Justice. Each report presents the program options and management issues in a topic area, based on a review of research and evaluation findings, operational experience, and expert opinion on the subject. The intent is to provide information to make informed choices in planning, implementing, and improving programs and practice in criminal justice. ------------------------------ National Institute of Justice Jeremy Travis Director Cheryl Crawford Program Monitor ------------------------------ Prepared for the National Institute of Justice, U.S. Department of Justice by Abt Associates Inc., under contract #OJP-94-C-007. Points of view or opinions stated in this document are those of the authors and do not necessarily represent the official position or policies of the U.S. Department of Justice. ------------------------------ The National Institute of Justice is a component of the Office of Justice Programs, which also includes the Bureau of Justice Assistance, the Bureau of Justice Statistics, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. ------------------------------ NCJ 165043 ------------------------------ Table of Contents Foreword Acknowledgments Executive Summary Sentencing Guidelines Intermediate Sanctions Combining Guidelines with Intermediate Sanctions Zones of Discretion Punishment Units Exchange Rates Categorical Exceptions Likely Future Developments Chapter 1: Intermediate Sanctions in Sentencing Guidelines The Case for Intermediate Sanctions Guidelines for Use of Intermediate Sanctions Endnotes Chapter 2: Intermediate Sanctions General Impediments to Effective Intermediate Sanctions Recidivism Prison Beds Cost Savings Experience with Intermediate Sanctions Boot Camps Intensive Supervision House Arrest and Electronic Monitoring Day-Reporting Centers Community Service Monetary Penalties Day fines Planning for the Future Endnotes Chapter 3: Sentencing Guidelines Effects on Sentencing Practices Changing Sentencing Decisions Reducing Disparities Disparity Reduction in the States Racial and Gender Differences Tying Policy to Resources Endnotes Chapter 4: Incorporating Intermediate Sanctions in Sentencing Guidelines Obstacles Efforts at Integrating Intermediate Sanctions Into Sentencing Guidelines Interchangeability Between Prison and Nonprison Sentences Residual interchangeability Limited interchangeability Bounded interchangeability District of Columbia Pennsylvania North Carolina Practical application Substitution of Penalties Punishment units Exchange rates Categorical Exceptions Delaware's Voluntary Continuum of Sanctions Interchangeability Among Nonincarcerative Punishments Authority Endnotes Chapter 5: Problems and Prospects Building on the Past Zones of Discretion Punishment Units Exchange Rates Categorical Exceptions Looking to the Future Jurisprudential Principles that Can Frame the Design of Guidelines for Intermediate Sanctions Purposes at sentencing The principle of parsimony Suggestions for Incorporating Intermediate Sanctions in Sentencing Guidelines The guidelines grid should contain four to six zones of discretion Guidelines should include dispositive presumptions Guidelines should authorize judges to declare and be guided by the relevant purposes of sentencing in every case Guidelines should establish policies and presumptions concerning Categorical exceptions Endnotes ----------------------------- Foreword New intermediate sanctions have been developed in every State since 1980, and nearly half the States have, have had, or are developing sentencing guidelines. Looking back just 27 years to 1970, both developments are dramatic; then, few States had programs that would today be considered intermediate sanctions, and not one had sentencing guidelines. From a late-1990's perspective, neither intermediate sanctions nor guidelines can be considered novel. What is novel, however, is that policymakers in many States have begun to recognize that intermediate sanctions and guidelines, taken together, may assist each to achieve its primary purposes. Until the early 1990's, intermediate sanctions and sentencing guidelines developed separately. Intermediate sanctions were developed at State and local levels to achieve various purposes, including cost savings, diversion of offenders from jail or prison, reduction in recidivism rates, and provision of midlevel punishments to midlevel offenders. Sentencing guidelines were developed at State levels, usually with purposes of reducing unwarranted disparities, reducing scope for racial and gender bias, and sometimes coordinating sentencing policies and prison resources. Measured in terms of their stated purposes, sentencing guidelines have been the more successful innovation. Evidence from a number of States indicates that sentencing guidelines have reduced unwarranted disparities (in general and in relation to race and gender), have enabled policymakers to make statewide changes in sentencing policy, and have permitted States to coordinate sentencing and prison-use policies. The evaluation evidence concerning intermediate sanctions has been less reassuring. Major evaluations of boot camps, intensive-supervision probation, community service, house arrest, and work-release programs show that many new programs do not achieve reductions in recidivism, corrections costs, or prison use. These results occur in part because of two common program characteristics: high failure rates for technical violations and high rates of net widening, that is, sentencing offenders who would otherwise have received less punitive probation sentences to new intensive programs. If the net-widening problem can be solved, however, intermediate sanctions can serve as cost-saving sanctions that reduce demand for prison beds without significantly diminishing public safety. Policymakers in a number of States--notably Massachusetts, North Carolina, Ohio, and Pennsylvania-have begun to design coordinated sentencing and intermediate sanctions policies. Because guidelines until recently dealt only with prison and jail sentences, promoters have urged that they be extended to bring greater consistency to probation sentences and intermediate sanctions. Because of the net-widening problem, guidelines have come to be seen as a device for establishing enforceable policies governing judges' sentences to intermediate sanctions. And, because many States are concerned about the fiscal ramifications of recent increases in sentence lengths for violent crimes, the combination of sentencing guidelines and intermediate sanctions has been seen as a cost- effective means to direct violent offenders to appropriate prison sentences and many nonviolent offenders to appropriate community sanctions. Delaware, North Carolina, and Pennsylvania have implemented guidelines that encompass sentences to confinement, probation, and intermediate sanctions. Late in 1996, several other States were at work on such systems. A number of different approaches have emerged. This report describes and assesses these approaches and others under consideration, so that States attempting such coordinated approaches in the future can build on the experiences of their predecessors. Jeremy Travis Director National Institute of Justice and Nancy Gist Director Bureau of Justice Assistance ------------------------------ Acknowledgments Most of what I know about sentencing policy developments I have learned from practitioners who have been working for several decades to make sentencing in the United States better. This report draws heavily on published materials from State and Federal sentencing commissions but also benefited from careful comments provided by many current and former State officials who read earlier drafts. These include Sandra Pearce and Robin Lubitz of North Carolina's Sentencing and Policy Advisory Commission; Kathleen Bogan and David Factor, former heads of the Oregon Criminal Justice Council that created Oregon's guidelines; Debra Dailey, long- time director of Minnesota's commission; John Kramer, long-time director of Pennsylvania's Commission on Sentencing and now staff director of the U.S. Sentencing Commission, and his able Pennsylvania deputy Cynthia Kempinen; Richard Van Wagenen and John Steiger, director and research director of Washington State's Sentencing Guidelines Commission, and also Roxanne Lieb, previously executive director of that commission, and David Boerner, a long-time participant in and commentator on sentencing policy developments in Washington. In addition, Dale Parent, probably the best-informed student of sentencing and community corrections policy developments in the United States, provided wise counsel and good technical advice throughout the writing of this report. Enormous thanks are due to Kate Hatlestad, who has worked with me on countless editing and writing projects over a decade and without whose graphic sense, remarkable patience, and attention to detail much of my work, and this report in particular, would be poorer things than they are. ------------------------------ Executive Summary More States have adopted and are developing sentencing guidelines than ever before, and intermediate sanctions continue to proliferate. Both assertions may surprise people who are not actively involved in these developments. Programs in the States get much less media and scholarly attention than do Federal developments. Owing to the unpopularity of the Federal sentencing guidelines and the near absence of intermediate sanctions in the Federal courts, a person who knew only of Federal developments could be excused for believing that both are failed or passe innovations of the 1980's. In the States, however, both guidelines and intermediate sanctions are thriving. Guidelines were in effect in more States in early 1997 than ever before, and both the number of intermediate sanctions programs and the number of people supervised in them grow every year. A principal reason both are thriving is that in important respects they can accomplish the goals policymakers set for them. A second reason is that policymakers in many States are worried about the fiscal consequences for State budgets of recently enacted mandatory minimum sentence laws, "three- strikes" laws, and general increases in severity of sentences for violent offenders. Legislators in a number of States, notably North Carolina, Ohio, and Pennsylvania, have enacted laws that will increase use of prison sentences and lengthen terms for violent offenders, while reducing use of prison sentences for nonviolent offenders and diverting them into intermediate sanctions instead. In all of these States, funds have been authorized both to build more prisons and to pay for more community- based programs. Coordinating sentencing policies expressed in guidelines with operation of intermediate sanctions may be the way to make ambitious new punishment policies workable and affordable. Sentencing Guidelines Consider guidelines first. State guidelines received considerable national attention in the 1980's and much less since. Yet there are many more guidelines systems in operation in the 1990's than in the 1980's, and they are typically more effective. Guidelines come in two broad forms: presumptive and voluntary. Presumptive guidelines, as the words suggest, establish rebuttable presumptions about appropriate sentences in individual cases. Judges can impose some other sentence by "departing," but must then give reasons for the departure which are subject to appellate review if a party appeals. Voluntary guidelines create no presumptions. They are in effect suggestions that the judge may accept if he or she wishes to do so. Although as many as 10 States adopted voluntary guidelines in the late 1970's and the 1980's, the few that were evaluated were shown to have few or no effects on sentencing patterns and most were abandoned or fell into desuetude. Delaware adopted voluntary guidelines in 1987 which remain in effect. Florida established voluntary guidelines in the early '80's and later made them presumptive. More recently, Arkansas and Missouri adopted voluntary guidelines. Only a few States initially adopted presumptive guidelines-Minnesota in 1980, Pennsylvania in 1981, and Washington in 1984-but they were adjudged reasonably effective at reducing disparities, diminishing scope for gender and racial bias, and improving coordination between sentencing policy and corrections resources. Newer presumptive schemes have since taken effect in Oregon, Kansas, North Carolina, and Ohio. Commissions early in 1997 were at work on guidelines in Maryland, Massachusetts, Michigan, Montana, Oklahoma, and South Carolina (and probably in other States of which I am unaware). A principal criticism of guidelines systems is that they are too limited in scope. The successful Minnesota, Washington, and Oregon guidelines in the 1980's governed decisions of who was sent to prison, and for how long, but set no standards for imposition of jail sentences, intermediate sanctions, or standard probation. Since fewer than 25 percent of convicted felons in many States are sentenced to State prison, those early guidelines systems were far from comprehensive. Why, the argument goes, if guidelines can reduce disparities and make sentencing more predictable, should they not apply to all sentences? Intermediate Sanctions The story concerning intermediate sanctions is similar-more attention and excitement in the 1980's than today but more, and more sophisticated, activity today. Corrections programs less restrictive than total confinement but more so than probation are not new. Halfway houses, curfews, and intensive probation programs existed in the 1950's and 1960's, they were conceptualized as rehabilitative programs. With the collapse of confidence in the ability of corrections programs to rehabilitate offenders, these programs lost credibility and support. During the 1970's, community service, intensive probation, and restitution programs were established in many jurisdictions; they were conceptualized as alternatives to imprisonment. There was little evidence that alternatives reduced recidivism rates, and there was much evidence that they resulted in "net widening," used by judges as alternatives to standard probation rather than imprisonment. Alternatives too soon lost credibility and support. In the 1980's, a series of new "intermediate sanctions" appeared and quickly spread. They included various forms of intensive probation, house arrest, electronic monitoring, boot camps, day-reporting centers, and day fines. Except for day fines, all these sanctions can be run as "front-end" or "back-end" programs. Entry into front-end programs is controlled by judges; corrections officials control entry into back-end programs, often in connection with early-release systems. Intermediate sanctions were typically conceptualized as punishments located on a continuum between prison and probation and were supposed to be more intrusive and burdensome than standard probation. Proponents sometimes promised that the new punishments would cost less than jail or prison, reduce prison crowding, and cut recidivism rates. Although major evaluations of day-reporting centers and day fines had not been published by the end of 1996, evaluations of intensive probation, house arrest, electronic monitoring, and boot camps were available, and they did not confirm over-enthusiastic proponents' predictions. Evaluated front-end programs typically experienced recidivism rates for new crimes neither higher nor lower than those of other sanctions for comparable offenders (but often much higher rates of technical violations and revocations), but because of extensive net widening and high rates of technical violations and revocations, they often cost more than confinement and worsened prison crowding. Back-end programs had similar recidivism- rate experiences but because corrections officials' control of entry prevented net widening, they were more effective at achieving cost savings and reducing prison-population pressures. Because intermediate sanctions have multiple purposes, the evaluation findings do not deprive them of credibility and support. First, implementation evaluations show that intermediate sanctions can deliver much more intrusive and burdensome punishments than standard probation; that is why technical violation and revocation rates are high. From a retributive perspective, intermediate sanctions can be much more punitive than probation and can be scaled in severity to the seriousness of the crime. Second, national evaluations of intensive probation and boot camps suggest (but do not prove) that intermediate sanctions with strong treatment components can improve treatment effectiveness and thereby reduce recidivism rates. Third, experience with back-end programs shows that intermediate sanctions can save money and prison resources if ways can be found to eliminate or greatly diminish net widening. Thus,intermediate sanctions can be used to save money and prison use, without significant sacrifices in public safety. The trick is to reduce net widening in front-end programs. In the American legal system, judges decide who is not sentenced to prison. Since that power is unlikely to be taken away, some way needs to be devised to set enforceable standards for sentences other than to imprisonment. Sentencing guidelines may be the answer. Combining Guidelines With Intermediate Sanctions North Carolina and Ohio have adopted new guidelines systems incorporating standards for the use of intermediate sanctions. Pennsylvania in 1994 overhauled its 13-year-old guidelines to do the same thing. The Massachusetts sentencing commission in 1996 presented a proposal for a similar set of guidelines to the Massachusetts legislature. Commissions are at work on similar plans in several other States, and the pressures of rising prison populations and corrections budgets are likely to lead more States to similar efforts. The early evidence from North Carolina suggests that guidelines incorporating intermediate sanctions can work. The North Carolina guidelines cover all felonies and misdemeanors and attempt to increase use of prison sentences for violent crimes. They also attempt to reduce prison use for nonviolent crimes by directing judges to sentence more offenders to intermediate sanctions. Both things happened in 1995, the guidelines' first full year of operation. Eighty-one percent of violent felons received prison sentences, up from 67 percent in 1993. Twenty-three percent of nonviolent felons were sent to prison, down from 42 percent in 1993. For all imprisoned felons, the mean predicted time to be served increased from 16 to 37 months. Notwithstanding North Carolina's apparent success, it is small wonder that earlier guidelines dealt only with prison (and occasionally jail) sentences. A number of serious impediments prevented the development of guidelines with broader scope. First, judges in many States fiercely resisted the very idea of guidelines, and overcoming that resistance for prison guidelines was challenge enough. In some States, including New York, Maine, Connecticut, and South Carolina, judicial resistance could not be overcome and no guidelines were adopted. Second, guidelines cannot realistically set standards for nonconfinement sentences, nor can judges be expected to follow them, unless credible programs exist to which offenders can be sentenced. Until recently, few States had extensive community corrections programs, especially outside the big cities. A number of States have now begun to provide community corrections funding to counties that makes operation of well-managed intermediate sanctions feasible; many States as yet have not. Third, nonconfinement guidelines present more complex issues than do prison guidelines. For serious violent crimes, and for chronic offenders, the current crime and the past criminal record are in most cases the primary considerations relevant to sentencing. Guidelines grids that array crime categories along one axis and criminal history along the other can efficiently encapsulate the major criteria for those cases. Sentencing for less serious crimes and offenders entails other considerations for many judges: might drug or sex- offender treatment be more effective than confinement, what are the likely collateral effects of imprisonment on the offender and his family, and are there special circumstances of the offense or the offender's characteristics that make one kind of sentence more appropriate than another? The two- axis grid by itself is not a very efficient way to address these and other offender-specific considerations. Incorporation of intermediate sanctions into sentencing guidelines is in its earliest days. There are, nonetheless, a number oftechniques that have been developed and ideas that have been examined. They are discussed briefly in this executive summary and at length in the body of this report. Zones of Discretion Most guidelines commissions that have tried to expand their guidelines' coverage to include nonconfinement sentences have altered the traditional guidelines format to include more zones of discretion. The first guidelines in Minnesota, Pennsylvania, and Washington divided their grids into two zones. One contained confinement cells setting presumptive ranges for prison sentences, and the other contained nonconfinement cells that gave the judge unfettered discretion to impose any other sentence, often including an option of jail sentences of up to one year. Minnesota's guidelines, for example, contained a bold black line that separated the confinement and nonconfinement zones. New North Carolina, revised Pennsylvania, and proposed Massachusetts guidelines, by contrast, have four or more zones. The details vary but they follow a common pattern. Sentences other than those authorized by the applicable zone are departures for which reasons must be given which are subject to review on appeal. One zone contains cells in which only prison sentences are presumed appropriate. A second might contain cells in which judges may choose between restrictive intermediate sanctions, such as residential drug treatment, house arrest with electronic monitoring, and a day- reporting center, and a prison sentence up to a designated length. A third might contain cells in which judges may choose among restrictive intermediate punishments. A fourth might authorize judges to choose between restrictive intermediate sanctions and a less restrictive penalty like community service or standard probation. A fifth might authorize sentencing choices only among less restrictive community penalties. Punishment Units A second approach that Oregon adopted and several other States considered is to express punishment in generic "punishment units" into which all sanctions can be converted. A hypothetical system might provide, for example, for the following conversion values: o One year's confinement: 100 units o One year's partial confinement: 50 units o One year's house arrest: 50 units o One year's standard probation: 20 units o 25 days' community service: 50 units o 30 days' intensive supervision: 5 units o 90 days' income (day fines): 100 units o 30 days' electronic monitoring: 5 units That is by no means a complete list; such things as drug testing, treatment conditions, and restitution might or might not be added. The values could be divided or multiplied to obtain values for other periods (for example, 75 days' confinement equals 20 units). If guidelines, for example, set "120 punishment units" as the presumptive sentence for a particular offender, a judge could impose any combination of sanctions that represented 120 units. One year's confinement (100 units) plus 60 subsequent days' intensive supervision (10 units) on electronic monitoring (10 units) would be appropriate. So would a 90-unit day fine (100 units) plus one year's standard probation (20 units). So would 25 days' community service (50 units) and six months' intensive supervision (30 units), followed by two years' standard probation (40 units). In practice, the punishment unit approach has proven too complicated to be workable. Oregon's original guidelines had two zones of discretion and in every cell in the nonconfinement zone specified the maximum number of punishment units for cases falling in that cell. However, detailed conversion rates were not established. All forms of confinement were given the same weight, and 16 hours' community service was made equivalent to 1 day's confinement. The commentary to the Oregon guidelines indicated that the provision of custody units was a foundation for later elaboration of conversion rates. The elaboration never happened. Pennsylvania likewise considered including the punishment unit concept in its revised 1994 guidelines but abandoned the idea as unworkable. Exchange Rates Another approach is simply to specify equivalent custodial and noncustodial penalties and to authorize judges to impose them in the alternative. Washington's commission did this in a modest way and later proposed a more extensive system, which the legislature did not adopt. Partial confinement and community service were initially authorized as substitutes for presumptive prison terms on the basis of 1 day's partial confinement or 3 days' community service for 1 day of confinement. The partial confinement/confinement exchange is probably workable (for short sentences; house arrest, assuming that to count as partial confinement, is seldom imposed for more than a few months), but the community service exchange rate is not. Like the punishment unit proposals, so far the exchange-rate approaches have been unable to overcome the psychological and political pressures to make "equivalent" punishments as objectively burdensome as prison, which limits their use to the most minor offenses and offenders. Under Washington's 3-days'-community-service-equals-1- day's-confinement policy, that range would permit community service in place of 3 to 10 days' confinement if existing successful programs were used as models. The difficulty is that community service programs, to be credible, must be enforced, and experience in this country and elsewhere instructs that they must be short. That is why the best-known American program in New York set 70 hours as a standard, and the national policies in England and Wales, Scotland, and the Netherlands set 240 hours as the upper limit. Those programs were designed to be used for offenders who otherwise would receive prison sentences of up to 6 months. Exchange rates are limited in their potential uses for the same reason punishment units are. For so long as prevailing views require that imprisonment be considered the normal punishment and that substitutes for imprisonment be comparably burdensome and intrusive, exchange rates are unlikely to play a significant role in sentencing guidelines. Categorical Exceptions Categorical exception policies, focusing not on the sanction but on the offender, are permissive. They authorize, but do not direct, judges to disregard otherwise applicable sentencing ranges if offenders meet specified criteria. One example is Rule 5.K.1 in the Federal guidelines that empowers judges to depart from guidelines if the prosecution files a motion proposing such a departure because the defendant has provided "substantial assistance [to the government] in the investigation or prosecution of another person." Once the motion is made, the judge is free from guidelines presumptions about appropriate sentences. The Federal categorical exception concerning substantial assistance, however, has no special relevance to intermediate sanctions. Only one State, Washington, has developed extensive categorical exception policies. Under the First- Time Offender Waiver, judges may disregard otherwise applicable guidelines in sentencing qualifying offenders and, guidelines commentary indicates, "The court is given broad discretion in setting the sentence. " Available alternatives include up to 90 days'jail or 2 years' probation and financial penalties, compulsory treatment, and community service. To be eligible, the offense must be a first conviction for a nonviolent, nonsexual offense (some drug offenders are also ineligible). In 1993, 2,139 offenders (of 7,224 eligible) were sentenced under the first-time offender exception. Washington's Special Sex Offender Sentencing Alternative authorizes judges to suspend prison sentences for most first-time sex offenders. To qualify, the offender must agree to two examinations by certified sex-offender treatment specialists and to preparation of a treatment plan. Offenders whose otherwise applicable presumptive sentence does not exceed 8 years are eligible. Following a decision that the offender is amenable to treatment, the judge may suspend the presumptive sentence and impose a community sentence that includes sex-offender treatment, up to 90 days in jail, community supervision, various financial obligations, and community service. In 1993, of 940 eligible offenders, 400 received special sex- offender departures. No other State has attained as much experience with use of categorical exceptions to sentencing guidelines (Washington also has a "work ethic [boot] camp" program that permits substitution of 4 to 6 months' boot camp for 22 to 36 months in prison). The idea, however, has potentially broad application to guidelines systems. Likely Future Developments Past experience suggests that some of the devices used to date are likely to be useful tools in incorporating intermediate sanctions into guidelines, and that others are not. At least in America in the 1990's, punishment units and exchange rates appear to be at dead ends. The most ambitious efforts to implement either concept have had negligible scope. Zones of discretion and categorical exceptions, however, do have roles to play. Use of zones of discretion has permitted policymakers to specify categories of offenses and offenders for which only particular kinds of sanctions are presumptively appropriate (only imprisonment, or only intermediate sanctions, or only less intrusive community penalties). Little guidance has as yet been provided to judges in choosing between imprisonment and other sanctions or among intermediate sanctions. Categorical exceptions are the most promising tools available for providing that guidance. Future sentencing commissions will probably develop current ideas in new ways. None of the commissions that have adopted a zones-of-discretion approach, for example, has attempted to provide guidance to judges on how to choose among authorized intermediate sanctions or community penalties or between intermediate sanctions and authorized confinement or community sanctions. This could easily be done by setting policies that particular kinds of sanctions are appropriate for particular kinds of offenders: an obvious example would be a policy that residential drug treatment be presumed appropriate for a drug-dependent chronic property offender. Depending on how convinced the commission was about the wisdom of the policy, it could be made presumptive (and thus require a "departure" with reasons given for any other sentence) or only advisory. Use of categorical exceptions likewise could be fine-tuned. The Federal and Washington State examples given above, for example, are permissive, entirely within the judge's discretion. A State might, however, want to make some categorical exceptions permissive and others presumptive. A first-time offender exception, like Washington's, might be permissive, while the Federal "substantial assistance" exception might be made presumptive. More States will be facing the kinds of issues discussed in this report. Most States have in recent years enacted laws mandating greatly lengthened sentences for violent offenders and for some drug and repeat offenders. Under the incentive of Federal funds for prison construction, many States now require that violent offenders serve at least 85 percent of those longer sentences. Forecasts of enormous resulting increases in prison operating costs led the North Carolina legislature to adopt guidelines intended to carry out those policies for violent offenders but also to divert many nonviolent offenders from prison to less expensive intermediate sanctions. Many States will face the same financial choices, and some at least are likely to try to follow the path that North Carolina and Pennsylvania have charted. Together the suggestions offered in this report for incorporating intermediate sanctions into sentencing guidelines may appear complicated, but that is a misimpression. Singly or together they constitute modest incremental steps toward creating comprehensive sentencing systems that incorporate confinement and nonconfinement sanctions and attempt to achieve reasonable consistency in sentencing while allowing judges to take account of meaningful differences between cases. ------------------------------ Chapter 1 Intermediate Sanctions in Sentencing Guidelines Policies concerning intermediate sanctions and sentencing guidelines are beginning to converge, which is desirable for substantive reasons. Guidelines are intended to bring greater fairness and predictability to sentencing but can do so for only a small percentage of cases if their application continues to be limited to imprisonment decisions. Likewise, intermediate sanctions are intended to provide a graduated series of sentencing options that save money and prison beds without introducing unacceptable risks to public safety but can do so only if the programs are used for the offenders for whom they are designed; often, they are used for less serious offenders. Sentencing guidelines and intermediate sanctions are two of the most significant criminal justice policy developments in recent decades. Half the States have adopted or considered statewide guidelines; and in early 1997, sentencing commissions were at work in more than 20 States. Intermediate sanctions have proliferated since 1980. All States have established intensive supervision probation and most have initiated use of electronic monitoring; many States have created boot camps and day-reporting centers, and a few have experimented with day fines and well-run community service programs. Although the first voluntary sentencing guidelines took effect in Denver in 1975 and the first presumptive guidelines in Minnesota in 1980, until recently guidelines systems have focused almost entirely on who goes to prison and for how long. Punishments other than incarceration have a longer history but, conceptualized as intermediate sanctions rather than as rehabilitative or prison- diversion programs, date only from the early 1980's. Although intermediate sanctions can manage lower-risk offenders as effectively as confinement, which is more costly, the success of "front-end" programs depends on their being used for the kinds of offenders for whom they were designed. However, if they are used for lower-risk offenders than program developers intended, programs which were designed to save public monies may end up increasing overall costs. Developers of sentencing guidelines in many States have decided that guidelines should be extended to encompass nonincarcerative sentences. Moreover, developers and evaluators of intermediate sanctions in many States recognize the mismatch problems common to front-end programs. Although control over entry to some programs (notably boot camps) has shifted from judges to prison officials in some States, it is inconceivable that judges' control over entry to most or all intermediate sanctions will be shifted. Sentencing guidelines are the likeliest device to help structure judges' discretion concerning use of nonincarcerative sentencing options. Because intermediate sanctions and sentencing guidelines developed independently, this report describes the past 20 years of their respective policy and research developments separately. Also discussed are the modest efforts, to date, to combine sanctions and guidelines; this "tying in" is essential if either is to achieve its primary purposes. o This chapter offers an introduction to the whole. o Chapter 2 describes the development of intermediate sanctions, including boot camps, intensive supervision probation, house arrest, fines and day fines, day-reporting centers, and community service. The evaluation literature shows that all of these programs have the potential to reduce costs without unduly endangering public safety if they are used for appropriate groups of offenders, but that they are often used inappropriately. o Chapter 3 describes the development of sentencing guidelines and summarizes evaluation findings showing that guidelines can (1) make sentencing more consistent and fairer; (2) reduce racial, gender, and other unwarranted disparities; and (3) bring sentencing policies and corrections resources into balance. o Chapter 4 discusses the limited efforts to date to incorporate intermediate sanctions into sentencing guidelines, which has become an objective of policymakers in many States. Several different approaches have been tried. None has yet been demonstrated to be successful, though some are promising. o Chapter 5 suggests next steps policymakers might consider. The Case for Intermediate Sanctions Intermediate sanctions have been seen as a way both to reduce the need for prison beds and to provide a continuum of sanctions that satisfies the "just deserts" concern for proportionality in punishment. During the 1980's, intermediate sanctions were often oversold as being able simultaneously to divert offenders from incarceration, reduce recidivism rates, and save money. Like most propositions that seem too good to be true, intermediate sanctions fell short of such high expectations. During the experimentation of the 1980's, it has become clear that (1) well-run programs can achieve some of their goals; (2) some conventional goals are incompatible; and (3) the availability of new sanctions may tempt judges and other officials to use them for offenders other than those for whom they were created. The goals of diverting offenders from prison and providing tough, rigorously enforced sanctions in the community have proven largely incompatible. A major problem is that close surveillance necessarily uncovers more technical violations than do less intensive sanctions. However, revocation rates for new crimes are seldom higher for offenders in evaluated programs than for comparable offenders in other programs. Similarly, there is no reason to conclude that offenders in evaluated programs commit technical violations at higher rates than their counterparts in other programs. When offenders in intensive programs breach a curfew or stop performing community service or get drunk or violate a no-drug-use condition, the chances of discovery are high. When this happens, many managers believe they must take punitive action-often revocation and resentencing to prison- to maintain the program's credibility in the eyes of judges, the media, and the community. Many evaluations have shown that intermediate sanctions are likely to save money or prison beds only if they are used primarily for offenders who otherwise would serve prison terms, yet many practitioners resist using new programs for such offenders. This results, in part, because some participants in every program will fail, and commit new crimes, and some practitioners are reluctant to be seen as responsible for these crimes. Judges misuse intermediate sanctions partly because they believe new community penalties are more appropriate than either prison or probation for some offenders. Forced by limited options from which to choose between prison and probation, judges will often choose probation, albeit with misgivings, because prison is perceived to be too severe or too disruptive for the offender and his or her family. When house arrest or intensive supervision becomes an option, these penalties may appear more appropriate than either probation or prison. This not-uncommon pattern is often pejoratively characterized as "net widening," but the epithet oversimplifies the problem. The notion that proportionality in punishment and the creation of a continuum of sanctions are desirable makes understandable some judges' preferences to divert offenders from probation to more intrusive sanctions. However, from the perspective of those who created these programs in order to save money and prison space by diverting offenders from prison, judges' actions defy the programs' rationales and obstruct achievement of their goals. Guidelines for Use of Intermediate Sanctions Probably the most important lesson learned from 15 years' experience with intermediate sanctions is that they are seldom likely to achieve their goals unless a means can be found to set and enforce policies governing their use. Two complementary methods are available for establishing enforceable policies to govern use of intermediate sanctions. First, discretion can be shifted from judges and prosecutors to corrections officials. "Back-end" programs to which offenders are transferred from prison or to which they are released early have been more successful at saving money and prison space than "front-end" programs. For this reason, in several States control over entry to boot camps has been shifted from judges to prison officials. Similarly, parole guidelines in some jurisdictions have been more effective in reducing parole release disparities than have some sentencing guidelines in reducing sentencing disparities.[1] Presumably parole guidelines are successful because decision processes in bureaucracies are placed in fewer peoples' hands and are more readily regulated by management controls than are decisions made by autonomous, politically selected judges. Second, sentencing guidelines, which in some jurisdictions have reduced disparities in terms of who goes to prison and for how long, can be extended to govern choices among intermediate sanctions and between intermediate sanctions and prison or probation. Some States have taken tentative steps in this direction, and many others are considering such guidelines extensions. These alternative approaches, shifting discretion from judges to corrections officials and regulating the discretion of judges, are not incompatible, and neither by itself is sufficient. Shifting discretion to corrections officials is feasible for some kinds of programs, but not others. If boot camps are intended to be used for prison-bound offenders and to achieve savings in money and prison beds, having prison officials select participants from the current prison population seems more likely to achieve and maintain these goals than having judges select offenders at sentencing. By definition, removing offenders from the prison population "diverts" them from prison, and savings may be realized. Likewise, strong arguments can be made for correctional control over admission to house arrest, intensive- supervision probation, and day-reporting centers that are designed to operate as early- or graduated-release mechanisms. Such programs are, by their nature, corrections options, designed to achieve corrections officials' objectives, and admission to them would seldom be determined by judges. Shifting front-end discretion to corrections officials is not likely to be a viable option in other sets of circumstances: 1. In jurisdictions that want to use a continuum of sanctions to achieve a proportionate sentencing system in which the severity of punishment is scaled to the seriousness of the crime, initial determinations of offense seriousness (and hence sanction severity) inevitably will fall to judges. A sentencing system might be imagined in which judges would be responsible for case processing only through adjudication with a corrections agency making all further sanctioning decisions, but that seems unlikely. If intrusive sanctions (like intensive supervision and house arrest), coupled with treatment, financial, and other conditions, have a place in a graded continuum of sanctions, judges must retain discretion over their use. 2. Some sanctions by their nature are inconsistent with correctional control. Day fines and community service are commonly used in a number of European countries as sanctions for moderately serious crimes committed by offenders for whom incapacitation is not a relevant sentencing purpose. The penalties are intended to be burdensome and punitive but at the same time to avoid the costs of imprisonment to the State and the collateral consequences of imprisonment to offenders and their families. If U.S. jurisdictions establish credible day-fine, community service, or other programs for such cases, correctional control of admission would contradict the programs' premises. 3. Some programs are designed to be used in place of imprisonment. In the 1970's, many community- based penalties were perceived by some people to be solely or primarily prison alternatives, and were commonly called "alternatives to imprisonment." In the 1990's, the term "intermediate sanctions" is more common, and they are considered mid-level punishments for moderately serious crimes. Nonetheless, for some categories of offenders, many programs are intended solely as alternatives to incarceration. The choice between incarceration and an alternative on a case-by-case basis is inexorably a judicial function. Thus, there appears no choice but to retain judges' discretion to sentence offenders to intermediate sanctions. Sentencing guidelines are the only mechanism now available for structuring judges' discretion concerning the use of intermediate sanctions. Although the earliest guidelines in Minnesota set presumptions concerning only prison sentences, more recent systems have tried to take into account other sanctions. The most ambitious efforts to date, in Pennsylvania, North Carolina, and Massachusetts, identify categories of cases for which only imprisonment is presumptively applicable, others for which judges may choose between imprisonment and intermediate sanctions, still others in which judges may choose between intermediate sanctions and less restrictive community penalties, and some for which only community penalties are presumptively appropriate. Even in these States, the guidance given to judges is not detailed, and the restrictions on their choices between confinement and intermediate sanctions, and among intermediate sanctions, typically are slight. Nonetheless, the approaches in Pennsylvania, North Carolina, and Massachusetts represent substantial advances beyond the earlier efforts of other States. Further progress will depend on policymakers' willingness to appropriate the funds necessary to create credible intermediate sanctions. The most visionary guidelines will fail if the sanctions they specify do not exist or are poorly managed and ineffective. The past 15 years' experience demonstrates that adequately funded, well-managed programs can achieve realistic goals. The rest of this report summarizes experience to date concerning the development of sentencing guidelines and efforts to regulate their use. Endnotes 1. Arthur D. Little, Inc., An Evaluation of Parole Guidelines in Four Jurisdictions. Report prepared for the National Institute of Corrections (Washington, DC, and Cambridge, MA: Arthur D. Little, Inc., 1981). Also see chapter 3 of Alfred Blumstein, Jacqueline Cohen, Susan E. Martin, and Michael Tonry, eds., Research on Sentencing: The Search for Reform, vol. 1 (Washington, DC: National Academy Press, 1983). ------------------------------ Chapter 2 Intermediate Sanctions Intermediate sanctions are important for three reasons: (1) they make possible sentencing systems in which the severity of punishments can be scaled to the seriousness of crimes; (2) they can provide mid-level punishments for offenders for whom incarceration is unnecessarily severe and ordinary probation is inappropriately slight; and (3) they provide an effective way to deal with violations of conditions; that is, a more restrictive intermediate sanction can be imposed rather than either ignoring the violation or revoking the sentence and locking up the violator. Well-run programs offer levels of supervision that permit credible monitoring of offenders' behavior and compliance with conditions. Especially important, in light of widespread drug dependence among offenders, these programs offer a means to back up required participation in drug-treatment programs. And, in programs that are used for the offenders for whom they are designed, these goals can be accomplished at less cost than incarceration and with comparable recidivism rates. In the evaluation literature, there are tantalizing but far from conclusive suggestions that participants in some intermediate sanctions achieve higher levels of participation in treatment programs and lower levels of recidivism than do comparable offenders in other programs.[1] There are two important caveats to the preceding advantages. First, they assume the existence of well-run programs. For reasons of policy or finance, some States have very few intermediate sanctions; California is a prominent example.[2] In every jurisdiction, some programs are better run than others, and it should go without saying that adequately funded and well-managed programs are more effective than those that are not. Second, and more fundamentally, the advantageous picture assumes that programs are used for the kinds of offenders for whom they are designed. Makers of sentencing and corrections policies generally have specific target groups of offenders in mind when they create a new program. In the 1970's, when intermediate sanctions were generally called "alternatives to incarceration" and were intended to divert offenders from prison, the target group was property and drug offenders and violent offenders who posed little threat to public safety. In the 1980's and 1990's, intermediate sanctions have been designed for offenders whose crimes warrant mid-level punishments and whose risks and needs assessments suggest that they can be dealt with appropriately in the community. In both periods, evaluation research has repeatedly shown extensive "net widening" on the part of judges. Net widening is a serious problem because it is beyond the control of corrections managers. New programs can be created if policymakers are prepared to appropriate the necessary money, and management problems can be fixed. Corrections managers, however, have no control over the discretionary decisions of judges. If judges simply disagree with a program's targeting policies, they will sentence the "wrong offenders" to the program, and the program will fail to achieve its goals. This chapter summarizes the substantial body of evaluation research on the operations and effects of intermediate sanctions programs. The first section gives a brief overview of problems that make reductions in recidivism, costs, and prison use difficult to achieve. The second section summarizes experience to date with the implementation and evaluation of various intermediate sanctions, including boot camps, intensive supervision, house arrest and electronic monitoring, day-reporting centers, community service, and monetary penalties. The evaluation literature raises doubts about the effectiveness of intermediate sanctions at achieving the goals their promoters have commonly set, but this does not mean that there are no effective programs. Only a handful have been carefully evaluated, and many of these were altered after their evaluations. Many experienced practitioners believe that their programs are effective, and some no doubt are. The evaluation literature does not "prove" that programs cannot succeed; it only demonstrates that many have not. General Impediments to Effective Intermediate Sanctions In retrospect, it was naive for proponents of new programs to assure skeptics that recidivism rates would decline, costs fall, and space pressures in prisons diminish. Net-widening pressures and management problems interact in complex ways to frustrate the achievement of those goals. Recidivism From influential evaluations of community service, intensive supervision, and boot camps, to mention only a few, comes the consistent finding that offenders given intermediate sanctions have similar recidivism rates for new crimes as comparable offenders receiving other sentences.[3] Failure and revocation rates for violation of other conditions are generally higher for offenders sentenced to intermediate sanctions than for comparable offenders under less intensive supervision. Both findings may be interpreted as good or bad. The finding of no effect on rates of new crime may be seen by many as good if the offenders involved have been diverted from prison and the new crimes are not very serious. Sentences to prison are much more expensive to administer than sentences to house arrest, intensive supervision, or day- reporting centers, and if the latter are no less effective at reducing subsequent criminality, they can provide nearly comparable public safety at greatly reduced cost. For offenders shifted from standard probation to an intensive sanction, however, the fact that more intensive sanctions do not reduce commission of new crimes raises different issues. If ordinary probation is no less effective at preventing new crimes than is a new intermediate sanction costing three times more, the case for sentencing offenders to the new program instead of probation cannot be made on cost-effectiveness terms. However, that does not mean that no case can be made: A "just deserts" argument can be made that intermediate sanctions deliver a more intrusive and burdensome punishment than probation, which can be appropriately proportioned to the offender's guilt. Although plausible, this argument shifts the rationale from utilitarian claims about crime and cost reductions to normative claims about the quality of justice. The finding that participants in intermediate sanctions have higher rates of violation of technical conditions provokes a similar set of concerns. Most observers agree that the increased violation and revocation rates result from the greater likelihood that violations will be discovered in intensive programs, not from greater underlying rates of violation. From an accountability perspective, the higher failure rates are good: offenders should comply with conditions, and consequences should follow when they do not. The contrary view is that the higher failure rates expose the inappropriateness of conditions (for example, prohibitions of drinking or drug use, or expectations that offenders will conform to middle- class behavioral standards they have never observed before) that many offenders will foreseeably breach and that do not involve harm to others. Many offenders have great difficulty achieving conventional, law-abiding patterns of living, and many stumble along the way. It might be argued that a social-work approach to community corrections should expect and accept the "stumbles" (so long as they do not involve significant new crimes) and hope that through them, with help, the offender will learn to become law-abiding. From this perspective, low-intensity programs may be favored because they reveal fewer violations, and high- intensity programs may be questioned because they necessarily reveal more. Prison Beds If all offenders in a new program were diverted from prison, an overall revocation rate for technical and new-crime violations as high as 50 percent would not be an insurmountable problem. The net savings in prison beds would be the number of persons diverted multiplied by the average time they would otherwise spend in prison, less the number of persons revoked for violations multiplied by their average term to be served. Unless the average time to be served after revocation substantially exceeded the average time that would have been served if not diverted, bed savings are inevitable. The combination of net widening and elevated rates of technical violations and revocations makes the calculation harder and makes prison-bed savings difficult to achieve. For front-end programs to which offenders are directly sentenced by judges, a 50 percent rate of prison diversion is commonly considered a success. Consider how the numbers work out: The 50 percent diverted from prison save prison beds, but the 50 percent diverted from probation do not, as they would not otherwise have occupied prison beds. If half of the offenders on probation suffer revocation and imprisonment, they represent a new demand for beds and a higher demand than would have otherwise existed because more of their technical violations will be discovered and acted on. Whether a particular program will save prison beds depends on why offenders' participation is revoked and in what percentage of cases, and whether they are sent to prison and for how long. But 50 percent is a high assumed diversion rate for a front-end program. If the true rate is 30 percent or 20 percent, net reductions in the demand for prison beds are unlikely. Cost Savings The third often-claimed goal of intermediate sanctions is to save money. The interaction of recidivism and demand for prison beds makes dollar savings unlikely except in the best cases. If a majority of program participants are diverted from probation rather than from prison, and if technical violation and revocation rates are higher in the intermediate sanction than in the programs to which offenders would otherwise be assigned, the chances of net cost savings are slight. For boot camps, for example, assuming typical levels of participant noncompletion and typical levels of postprogram revocation, Dale Parent has calculated that "the probability of imprisonment has to be around 80 percent just to reach a break-even point-that is, to have a net impact of zero on prison bed- space."[4] Cost analyses must, however, look beyond diversion rates, revocation rates, and prison beds. At least three other considerations are important: transaction costs, marginal costs, and savings to the community at large. First is the issue of transactions costs. Net-widening programs that shift probationers to intensive supervision and then shift some of those to prison cost the State more because they use up additional prison space. In addition, they create new expenses for probation offices, prosecutors, courts, and corrections agencies responsible for administering each of these transfers. Correctional cost-benefit analyses often ignore cost ramifications for other agencies, but the other agencies must either pay additional costs or refuse to cooperate. Second is the problem of marginal costs. Especially in the 1980's, promoters of new programs commonly contrasted the average annual costs per offender of administering a new program (for example, $4,500) with the average annual cost of housing one prisoner (for example, $18,500) and claimed substantial potential cost savings. This ignores the complexities presented by net widening and raised revocation rates, but it also ignores a more important problem of scale. For an innovative small program of 50 to 100 offenders (many were and are of this size or smaller), the valid comparison is with the marginal, not the average, costs of housing diverted offenders. Unless a prison or a housing unit is closed or not opened because the system has 50 fewer inmates, the only savings are incremental costs for food, laundry, supplies, and other routine items. The major costs of payroll, administration, debt service, and maintenance are little affected. In a system with 5,000, 15,000, or 50,000 inmates, the costs saved by diverting a few hundred offenders are scarcely noticeable. Third is the issue of savings to the larger community, which result from crimes avoided by incapacitating offenders. If believable values could be attached to crimes that would be averted by imprisonment but that would occur if offenders were sentenced to intermediate sanctions, they would provide important data for considering policy options. Unfortunately, this subject has as yet received little sustained attention, most of which has been ideological and polemical. Some conservative writers have claimed that increased use of imprisonment is highly cost-effective.[5] Liberal scholars have responded by showing the implausibility of assumptions made in such calculations. Zimring and Hawkins, for example, showed that, on the assumptions made in Zedlewski's analysis about the number of crimes prevented for each inmate confined, the 237,000-person increase in the prison population that occurred between 1977 and 1986 should "have reduced crime to zero on incapacitation effects alone . . . on this account, crime disappeared some years ago."6 The same is true of all the conservative cost-benefit analyses mentioned previously. No one who has worked with the justice system will be surprised by the observation that the system is complex and that economic and policy ramifications ripple through it when changes are made to any one of its parts. Sometimes this truism has been overlooked, much to the detriment of programs on behalf of which oversimplified claims were made. Georgia, for example, operated a pioneering front- end intensive supervision program that, at one time, was claimed to have achieved remarkably low recidivism rates (for new crimes) and to have saved Georgia the cost of building two prisons. It was later realized that many of those sentenced to ISP were low-risk offenders convicted of minor crimes who otherwise would have received probation. Although serving initially as an exemplar of successful intensive supervision for probationers and parolees (ISP) programs that save money and reduce recidivism rates, Georgia's ISP program now serves as an exemplar of net-widening programs that increase system costs and produce higher rates of revocation for violations of technical conditions.[7] Experience With Intermediate Sanctions An evaluation literature on intermediate sanctions has accumulated, and lessons have been learned. The following subsections discuss research on boot camps, ISP, house arrest and electronic monitoring, day-reporting centers, community service, and monetary penalties. The discussion of each emphasizes the more substantial evaluations and literature reviews. In some cases, for example concerning ISP, day-reporting centers, and boot camps, relatively recent and detailed literature reviews are available for readers who want more information.8 In other cases, for example concerning fines and community service, the best literature reviews are more dated; relatively little research on those subjects has been conducted in the United States in recent years, and these articles, despite their dates, cover most of the important research.[9] Boot Camps The emerging consensus must be discouraging to proponents of boot camps. Although promoted as a means to reduce recidivism rates, corrections costs, and prison crowding, most boot camps have no discernible effect on subsequent offending and increase costs and crowding.[10] Many have been front-end programs that have drawn participants from among offenders who otherwise would not have been sent to prison. In many, one-third to one-half of participants fail to complete the program and are sent to prison as a result. In most programs, close surveillance of graduates after release produces technical violation and revocation rates that are higher than those of comparable offenders in less intensive programs. Still, the news is not all bad. Back-end programs to which imprisoned offenders are transferred by corrections officials for service of a fixed-term boot camp sentence in lieu of a longer conventional sentence do save money and prison space, although they too often experience high failure, technical violation, and revocation rates. Most of what we know about the effects of boot camps on participants comes from a series of studies by Doris MacKenzie and colleagues at the University of Maryland, from a U.S. General Accounting Office survey of research and experience in 1993, and from an early descriptive overview of boot camps commissioned by the National Institute of Justice.[11] One tentative finding concerning possible positive effects of rehabilitative programs on recidivism merits emphasis. Although MacKenzie and her colleagues concluded overall that boot camps do not by themselves result in reduced recidivism rates, they found evidence in Illinois, New York, and Louisiana of "lower rates of recidivism on some measures," which they associated with strong rehabilitative emphases in those States.[12] An earlier article describes a "somewhat more positive" finding that graduates under intensive supervision "appear to be involved in more positive social activities (e.g., work, attending drug treatment) [after release] than similar offenders on parole or probation."[13] Boot camps illustrate most vividly of all intermediate sanctions the ways in which net widening, rigorous enforcement of conditions, and high revocation rates can produce the unintended side effects of increased costs and prison use in programs intended to reduce both. Figure 1, from Parent's work, shows the effects of different assumptions of prison diversion and post-program revocation and reincarceration on prison beds in a hypothetical 90-day, 200-bed facility. Other assumptions regarding failure rates within the program and lengths of confinement in lieu of boot camp and after revocation, based on averages documented in MacKenzie's eight-State assessment, were built into the analysis. The diagonal lines show the effects of different post-program reincarceration rates. At the lower 15 percent rate (broken line), boot camps create a net demand for additional beds if less than half those in the program would otherwise have gone to prison. At the more realistic 40 percent reincarceration rate (solid line), at least 80 percent of participants must have been diverted from prison if beds are to be saved.[14] If a primary goal is to reduce prison use, the policy implications are straightforward. First, boot camps should recruit offenders who have a very high probability of imprisonment. This means that participants should usually be selected by corrections officials from among prisoners rather than by judges from among sentenced offenders. Second, boot camps should minimize failure rates by reducing in-program failures and post-release failures. This means that misconduct should be dealt with within the boot camp whenever possible rather than by transfer to a regular prison and that misconduct after release should be dealt with within the supervision program whenever possible rather than by reincarceration. Third, participants should be selected from among prisoners who otherwise would serve a substantial term of imprisonment. Transfer of prisoners serving 9-month terms to a 180-day boot camp is unlikely to reduce costs and system crowding. Transfer of prisoners serving 2- or 3-year minimum terms, or longer, is likely to reduce both. Intensive Supervision ISP was initially the most popular intermediate sanction, has the longest history, and has been the most extensively evaluated. ISP has been the subject of the only multi-site experimental evaluation involving random allocation of eligible offenders to ISP and to whatever the otherwise appropriate sentence would have been.[15] ISP programs probably exist in every State. A General Accounting Office survey in 1989 identified programs in 40 States and in the District of Columbia.[16] Programs can be organized by State or county correctional agencies and can be located in parole, probation, and prison departments; as a result they are easy to miss in national mail and phone surveys. Contemporary ISP programs have caseloads ranging from 2 officers for 25 probationers to 1 officer for 40 probationers. More frequent contact leads to closer surveillance, which makes it likelier that misconduct is discovered and punished. Because of closer surveillance, low- to mid-risk offenders can be diverted from prison to less-costly ISP, without unduly jeopardizing public safety. Because of the frequency of contact, subjection to unannounced urinalysis tests for drugs, and rigorous enforcement of restitution, community service, and other conditions, ISP is more punitive than conventional probation. Evaluation findings in ISP programs parallel those for boot camps. Front-end programs in which judges control placement tend to draw more heavily from offenders who would otherwise receive less restrictive sentences than from offenders who would otherwise have gone to prison or jail. The multisite ISP evaluation by RAND, in which jurisdictions agreed in advance to cooperate with a random assignment system for allocating offenders to sanctions, was unable to evaluate front-end ISP programs when judges refused to accept the outcomes of the randomization system.[17] Back-end programs draw from prison populations, and accordingly are more likely to 0save money and prison beds. Like the boot camp evaluations, the ISP evaluations have concluded that ISP offenders do not have lower recidivism rates for new crimes than do comparable offenders receiving different sentences, but typically experience higher technical violation and revocation rates. As with supporters of boot camps, early proponents argued that ISP, while reducing recidivism rates and rehabilitating offenders, would save money and prison resources; evaluations suggest that the combination of net widening, high revocation rates, and processing costs for revocations makes most cost-savings claims improbable.[18] There is one tantalizing positive finding that parallels a positive boot camp finding: ISP did succeed in some sites in increasing participants' involvement in counseling and other treatment programs. The drug treatment literature demonstrates that participation, whether voluntary or coerced, can reduce both drug use and crime by drug-using offenders. Because Drug Use Forecasting data indicate that one-half to three-fourths of arrested felons in many cities test positive for drug abuse, ISP may hold promise as a device for getting addicted offenders into treatment and keeping them there.[19] Here, too, the policy implications are straightforward. Because recidivism rates for new crimes are no higher for ISP participants than for comparable imprisoned offenders, ISP is a cost- effective prison alternative for offenders who do not present unacceptable risks to public safety. Cost savings are likely to depend, however, on finding a way to ensure that ISP is used for target offenders. ISP may offer a promising tool for facilitating treatment for drug-using and other offenders. House Arrest and Electronic Monitoring The lines that distinguish community penalties begin to blur after ISP. House arrest, often called home confinement, has as a precursor the curfew condition traditionally attached to many probation sentences and may be ordered as a sanction in its own right or as a condition of ISP.[20] Most affected offenders, however, do not remain in their homes around the clock, but instead are permitted to go out to work or to participate in treatment, education, or training programs. House arrest is sometimes, but not necessarily, backed up by electronic monitoring. House arrest programs have grown and proliferated. The largest is in Florida, where more than 13,000 offenders were on house arrest in 1993. Programs coupled with electronic monitoring-a subset-existed nowhere in 1982, in 7 States in 1986, and in all 50 States in October 1990. In 1986, only 95 offenders were subject to monitoring, a number that rose to 12,000 in 1990 and to a daily count of 30,000 to 50,000 in 1992 and 1993.[21] No published evaluations of house arrest match the scale or sophistication of the best studies of boot camps or ISP. One analysis of Florida's program concluded that it drew more offenders from among the prison-bound than from the probation-bound.[22] However, this conclusion is based on two dubious analyses. The first looked to see whether offenders on house arrest should, under Florida's sentencing guidelines, have been sentenced to confinement. This assumes, however, that the guidelines significantly constrained the choices Florida judges made; the best evidence indicates that they did not.[23] The second analysis compared characteristics of probationers, house arrest offenders, and prisoners, and concluded that those on house arrest more closely resembled prisoners than probationers. Unfortunately, there is no way to be sure the comparison group was actually comparable. A case study of the development, implementation, and evolution of a back-end program in Arizona cautions that house arrest programs are likely to share the prospects and problems of intermediate sanctions generally. Originally conceived as a money-saving system for early release of low-risk offenders from prison, the program eventually incurred greater costs. Participants had to be approved by the parole board, which proved highly cautious and released few eligible inmates. The rate of revocation for technical violations (34 percent of participants) was twice that for ordinary parolees.[24] No larger-scale evaluations have been conducted to date. House arrest coupled with electronic monitoring has been the subject of many small studies and a linked set of three evaluations in Indianapolis.[25] Two recent literature reviews stress the scantiness of the evidence on prison diversion, recidivism, and cost-effectiveness. On recidivism, Renzema notes that most of the "research is uninterpretable because of shoddy or weak research designs."[26] The most comprehensive research review observes, "We know very little about either home confinement or electronic monitoring."[27] There seems little reason to believe, therefore, that house arrest is any less vulnerable to net widening than is ISP or that it is likely to achieve different findings on recidivism. Day-Reporting Centers Day-reporting centers developed earlier and more extensively outside the United States. The earliest U.S. day-reporting centers-places in which offenders spent their days under surveillance and participating in treatment and training programs, and slept elsewhere at night-date from the mid-1980's. The English precursors, originally called day centers and now probation centers, began operation in the early 1970's. Most of our knowledge of U.S. day-reporting centers comes from descriptive writing; no published literature provides credible findings on important empirical questions. The English programs date from the creation of four "day-training centers" established under the Criminal Justice Act of 1972 that were charged to provide intensive training programs for persistent petty offenders whose criminality was believed to be rooted in general social inadequacy, and from the creation of ad hoc day centers for serious offenders that were set up by a number of local probation agencies. The training centers were adjudged unsuccessful and were soon closed. The probation-run day centers, however, thrived after enabling legislation was enacted in 1982, numbering at least 80 by 1985 and serving thousands of offenders by the late 1980's.[28] Programs vary, with some emphasizing control and surveillance, some operating as therapeutic communities, and most offering a wide range of mostly compulsory activities. The maximum term of involvement is 60 days, and some programs have 30- day or 45-day limits. A major Home Office study concluded that "most centers unequivocally saw their aim as diversion from custody," that more than half of the participating offenders had previously been imprisoned, and that 47 percent had six or more prior convictions.[29] The results were seen as so promising that the Criminal Justice Act of 1991 envisioned a substantial expansion. In the United States, a 1989 National Institute of Justice survey identified 22 day-reporting centers in 8 States. Another survey conducted in mid-1994 identified 114 day-reporting centers in 22 States; most opened after 1990. The best-known (or the best-documented) centers were established in Massachusetts-in Springfield (Hampton County Sheriff's Department) and in Boston (the Metropolitan Day-Reporting Center)-and both were based in part on the model provided by the English day centers.[30] The 1994 survey showed significant differences between newer and older programs. Centers started before 1992 were more likely to be operated by private vendors, to be back-end programs that received offenders from prisons or jails, and to give greater emphasis to providing treatment and services. Newer centers were generally operated by public agencies, were front-end programs receiving clients under pre-trial release or by direct sentence from the courts, and devoted less emphasis to treatment.[31] As with the English centers, U.S. programs vary. Programs range in duration from 40 days to 9 months, and program content varies widely. Most require development of hour-by-hour schedules of each participant's activities, some are highly intensive with 10-or-more supervision contacts per day, and a few include 24-hour electronic monitoring.[32] The 1994 survey showed generally high negative termination rates, averaging 50 percent and ranging from 14 to 86 percent in the programs surveyed.[33] Unfortunately, no substantial evaluations have been published.[34] Community Service Community service is the most underused intermediate sanction in the United States. Used in many countries as a mid-level penalty to replace short prison terms for moderately severe crimes, community service in the United States is used primarily as a probation condition or as a penalty for minor crimes like motor vehicle offenses. This is unfortunate because community service is a burdensome penalty that meets with widespread public approval,[35] is inexpensive to administer, and produces public value; also, it can to a significant extent be scaled to the seriousness of crimes. The first large-scale community service programs were established in England in the early 1970's, followed by pilot programs in Scotland in the late 1970's and in the Netherlands in the early 1980's.[36] All three efforts led to programs that have been fully institutionalized as penalties falling between probation and imprisonment in those countries' sentencing tariffs. With the exception of one major U.S. study, the most ambitious evaluation research has been carried out elsewhere. In England and Wales, Scotland, and the Netherlands, community service orders (CSOs) were statutorily authorized with the express aim that they serve as an alternative to short-term incarceration. The U.S. study examined a pilot community service program in New York City that was intended to substitute community service for jail terms of up to 6 months.[37] CSOs in England and Scotland are regarded as more intrusive and punitive than probation and as an appropriate substitute for imprisonment.[38] CSOs can involve between 40 and 240 hours of work supervised by a community service officer, and failure to participate or cooperate can result in revocation. It is generally estimated that half of those sentenced to community service would otherwise be sentenced to prison and half to less severe penalties.[39] Reoffending rates are believed and generally found to be neither higher nor lower than those of comparable offenders sent to prison.[40] The story in the Netherlands is similar. Evaluations reached the expected conclusion that recidivism rates were no worse but that judges were using CSOs both for offenders who otherwise would have received prison sentences and those who otherwise would have received suspended sentences (with the proportions unknown).[41] The only well-documented U.S. community service project, operated by the Vera Institute of Justice, was established in 1979 in the Bronx and eventually spread to Manhattan, Brooklyn, and Queens. The program was designed as a credible penalty for repetitive property offenders who had previously been sentenced to probation or jail and who faced a 6-month or longer jail term for the current conviction. Offenders were sentenced to 70 hours' community service under the supervision of Vera foremen. Participants were told that attendance would be closely monitored and that nonattendance and noncooperation would be punished. A sophisticated evaluation concluded that recidivism rates were unaffected by the program, that prison diversion goals were met, and that the program saved taxpayers' money.[42] For offenders who do not present unacceptable risks of future violent (including sexual) crimes, a punitive sanction that costs much less than prison to implement, that promises comparable reoffending rates, and that presents negligible risks of violence by those who would otherwise be confined has much to commend it. Monetary Penalties Monetary penalties for nontrivial crimes have yet to catch on in the United States; however, this does not deny that millions of fines are imposed every year. Studies conducted as part of a 15-year program of fines research coordinated by the Vera Institute of Justice showed that fines are nearly the sole penalty for traffic offenses and, in many courts, are often imposed for misdemeanors;[43] in many courts, most fines are collected. Although ambiguous lines of authority and absence of institutional self-interest sometimes result in haphazard and ineffective collection, courts that wish to do so can be effective collectors.[44] Although monetary penalties are not widely used in the United States except for minor offenses, convicted offenders in some jurisdictions are routinely ordered to pay restitution and, in most jurisdictions, are routinely ordered to pay growing lists of fees for probation supervision, for urinalyses, and for use of electronic monitoring equipment. A survey of monetary exactions from offenders carried out in the late 1980's identified more than 30 separate charges, penalties, and fees that were imposed by courts, administrative agencies, and legislatures. These commonly included court costs, fines, restitution, and payments to victim compensation funds. Often, they also included a variety of supervision and monitoring fees.[45] The problem, as George Cole and his colleagues reported when summarizing the results of a national survey of judges' attitudes about fines, is that, "at present, judges do not regard the fine alone as a meaningful alternative to incarceration or probation."[46] This U.S. inability to see fines as serious penalties stands in marked contrast to the legal systems of other countries. In the Netherlands, the fine is legally presumed to be the preferred penalty for every crime, and Section 359(6) of the Code of Criminal Procedure requires judges to provide a statement of reasons in every case in which a fine is not imposed. In Germany in 1986, 81 percent of all sentenced adult criminals were ordered to pay a fine, including 73 percent of those convicted of crimes of violence. In Sweden in 1979, fines constituted 91 percent of all sentences. In England in 1980, fines were imposed in 47 percent of convictions for indictable offenses (roughly equivalent to U.S. felonies); these included 45 percent of convicted sex offenders, 24 percent of burglars, and 50 percent of those convicted of assault.[47] European monetary penalties for serious crimes take two forms. The first is the day fine, in use in the Scandinavian countries since the turn of the century and in Germany since the 1970s, which scales fines both to the defendant's ability to pay (some measure of daily income) and to the seriousness of the crime (expressed as the number of daily income units assessed).[48] The second is the use of the fine as a prosecutorial diversion device; that is, in exchange for paying the fine (often the amount that would have been imposed after conviction) the criminal charges are dismissed. Only the day fine has attracted much attention in the United States. Some of the efforts to establish day-fine systems are discussed below. First, though, some discussion of the remarkable success of prosecutorial diversion programs seems warranted. In Sweden, prosecutors routinely invite defendants they intend to charge to accept a fine calculated on day-fine principles in exchange for dismissal of the charges. Nearly 70 percent of fines are imposed in this way.[49] Under Section 153a of the German Code of Criminal Procedure, which has been in effect since 1974, the prosecutor, if "convinced of the defendant's guilt," may propose a conditional dismissal under which the defendant agrees to pay a fine. If the charges are serious, the judge must approve the arrangement (approval is seldom withheld). The defendant need not confess guilt. Two hundred forty thousand cases were resolved by conditional dismissal in 1989, constituting a 16 percent reduction in indictments that otherwise would have been filed.[50] In the Netherlands, the 1983 Financial Penalties Act authorized prosecutors to resolve criminal cases by means of an arrangement comparable to the German conditional dismissal. Defendants charged with crimes bearing up to 6-year prison sentences are eligible. The prosecution is terminated but can be reinstated if the defendant commits a new crime within 3 years. The prosecutorial diversion program has been credited with keeping the number of criminal trials stable between 1980 and 1992, despite a 60 percent increase in recorded crime. Two-thirds of criminal cases are settled out of court by prosecutors.[51] Day fines. Despite the substantial successes of fines as part of prosecutorial diversion programs in many countries, in the United States the day fine has received principal attention as a penal import from Europe. The results to date are at best mildly promising. The initial pilot project was conducted in Staten Island, New York, in 1988-89, under the auspices of the Vera Institute of Justice. Judges, prosecutors, and other court personnel participated in the planning, and implementation was remarkably successful. Most judges cooperated with the new voluntary scheme, judges followed the system, the average fine imposed increased by 25 percent, the total amount ordered on all defendants increased by 14 percent, and 70 percent of defendants paid their fines in full.[52] The Staten Island findings, while not unpromising, are subject to two important caveats. First, the participating court had limited jurisdiction and handled only misdemeanors; the use of day fines for felonies thus remains untested. Second, applicable statutes limited total fines for any charge to $250, $500, or $1,000, depending on the misdemeanor class, and thus artificially capped fines at those levels and precluded meaningful implementation of the scheme in relation to other than the lowest-income defendants. A second modest pilot project was conducted for 12 weeks in 1989 in Milwaukee, and four projects funded by the Bureau of Justice Assistance operated for various periods between 1992 and 1994 in Maricopa County (Phoenix), Arizona; Bridgeport, Connecticut; Polk County, Iowa; and Coos, Josephine, Malheur, and Marion Counties in Oregon. Only the Phoenix program remains in operation. The Milwaukee project applied only to noncriminal violations, resulted in reduced total collections, and was abandoned. The Phoenix project, known as FARE (Financial Assessments Related to Employability), was conceived as falling between unsupervised and supervised probation. The Iowa pilot included only misdemeanants. The Oregon projects included misdemeanants and probationable felonies (except Marion County, the largest, which covered only misdemeanants). Only in Connecticut did the pilot cover a range of felonies and misdemeanors.[53] A RAND Corporation evaluation of the Arizona, Connecticut, Iowa, and Oregon projects was funded by the National Institute of Justice. It concluded that the Phoenix project was somewhat successful in achieving its modest goals (principally in terms of revenue collection and in reduced supervision levels for fined low-risk probationers; it was not intended to substitute for confinement sentences).[54] Most Western justice systems rely heavily on financial penalties. In coming decades, U.S. jurisdictions are likely to continue their experiments with such penalties and to assign them important sanctioning roles. For the present, however, their lack of demonstrated effectiveness and acceptance will likely limit the role they can play. Planning for the Future The core findings of research on intermediate sanctions are positive, though some may not initially recognize that. That the original claims for intermediate sanctions-reduced recidivism, lower costs, less prison use-have not been widely demonstrated is no reason to abandon intermediate sanctions but should spur policymakers, correctional officials, and judges to devise ways to ensure that they are used as intended. For offenders convicted of mid-level crimes (and for some people convicted of first-time but out-of- character serious crimes) who do not pose unacceptable risks of violence, an intermediate sanction in place of prison will save money and avoid some of the collateral consequences of incarceration for offenders' families and public welfare budgets. The key is to ensure that programs are used for the right kinds of offenders and that overly rigid enforcement of conditions does not ratchet minor offenders into very punitive sanctions. Chapter 3 summarizes research on sentencing guidelines, which are the only realistic mechanism currently available to structure the discretion of judges in deciding whether and when to sentence offenders to intermediate sanctions. Endnotes 1. For intensive supervision of practitioners and parolees (ISP), see Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. For boot camps, see Doris Layton MacKenzie, 1994, "Boot Camps: A National Assessment," Overcrowded Times 5(4):1, 14-18. 2. Joan Petersilia, 1995, "How California Could Divert Nonviolent Prisoners to Intermediate Sanctions," Overcrowded Times 6(3):4-8. 3. Douglas McDonald, Punishment Without Walls: Community Service Sentences in New York City (New Brunswick, NJ: Rutgers University Press, 1986). Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. Doris Layton MacKenzie and C. Souryal, Multi-Site Evaluation of Shock Incarceration, Report to the National Institute of Justice (College Park, MD: University of Maryland, Department of Criminology and Criminal Justice, 1994). 4. Dale Parent, 1994, "Boot Camps Failing To Achieve Goals," Overcrowded Times 5(4): 9. 5. For examples, see Edwin Zedlewski, Making Confinement Decisions, Research in Brief (Washington, DC: National Institute of Justice, 1987); John J. DiIulio, 1990, "Crime and Punishment in Wisconsin," Wisconsin Policy Research Institute Report 3(7):1-56; John J. DiIulio and Anne M. Piehl, "Does Prison Pay?" (Unpublished manuscript, Princeton, NJ: Princeton University, Center for Domestic and Comparative Policy Studies, 1991); William Barr, The Case for More Incarceration (Washington, DC: U.S. Department of Justice, Office of Policy Development, 1992); Ted R. Miller, Mark A. Cohen, and Brian Wiersema, Victim Costs and Consequences: A New Look, National Institute of Justice Research Report (Washington, DC: U.S. Department of Justice, 1996). 6. Franklin E. Zimring and Gordon Hawkins, The Scale ofImprisonment (Chicago: University of Chicago Press, 1991). 7. Billie Erwin, Evaluation of Intensive Probation Supervision in Georgia, Report to the National Institute of Justice (Atlanta: Georgia Department of Corrections, 1987). Billie Erwin and Lawrence Bennett, New Dimensions in Probation: Georgia's Experience With Intensive Probation Supervision, Research in Brief (Washington, DC: National Institute of Justice, 1987). Todd R. Clear and James M. Byrne, "The Future of Intermediate Sanctions: Questions to Consider," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 321. 8. U.S. General Accounting Office, Intermediate Sanctions: Their Impacts on Prison Crowding, Costs, and Recidivism Are Still Unclear (Gaithersburg, MD: U.S. General Accounting Office, 1990). Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. Dale Parent, Jim Byrne, Vered Tsarfaty, Laura Valade, and Julie Esselman, Day Reporting Centers, National Institute of Justice Issues and Practices Report (Washington, DC: U.S. Department of Justice, 1995). U.S. General Accounting Office, Prison Boot Camps: Short-Term Prison Costs Reduced, But Long-Term Impact Uncertain (Washington, DC: U.S. General Accounting Office, 1993). Doris Layton MacKenzie, 1994, "Boot Camps: A National Assessment," Overcrowded Times 5(4):1, 14-18. Roberta C. Cronin and Mei Han, Boot Camps for Adult and Juvenile Offenders: Overview and Update, National Institute of Justice Research Report (Washington, DC: U.S. Department of Justice, 1994). 9. Sally Hillsman, "Fines and Day Fines," in Crime and Justice: A Review of Research, vol. 12, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1990). Ken Pease, "Community Service Orders," in Crime and Justice: A Review of Research, vol. 6, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1985). 10. Dale Parent, 1994, "Boot Camps Failing To Achieve Goals," Overcrowded Times 5(4):8-11. Doris Layton MacKenzie, 1994, "Boot Camps: A National Assessment," Overcrowded Times 5(4):1, 14-18. Roberta C. Cronin and Mei Han, Boot Camps for Adult and Juvenile Offenders: Overview and Update, National Institute of Justice Research Report (Washington, DC: U.S. Department of Justice, 1994). 11. Doris Layton MacKenzie and J. W. Shaw, 1990, "Inmate Adjustment and Change During Shock Incarceration: The Impact of Correctional Boot Camp Programs," Justice Quarterly 7(1):125-50. Doris MacKenzie, 1993, "The Impact of Shock Incarceration on Technical Violations and New Criminal Activities," Justice Quarterly 10:463-87. Doris Layton MacKenzie, 1993, "Boot Camp Prisons 1993," National Institute of Justice Journal 227:21-28. Doris Layton MacKenzie, 1994, "Boot Camps: A National Assessment," Overcrowded Times 5(4):1, 14-18. Doris Layton MacKenzie and C. Souryal, Multi-Site Evaluation of Shock Incarceration, Report to the National Institute of Justice (College Park, MD: University of Maryland, Department of Criminology and Criminal Justice, 1994). U.S. General Accounting Office, Prison Boot Camps: Short-Term Prison Costs Reduced, But Long- Term Impact Uncertain (Washington, DC: U.S. General Accounting Office, 1993). Dale Parent, Shock Incarceration: An Overview of Existing Programs (Washington, DC: National Institute of Justice, 1989). 12. Doris Layton MacKenzie, 1994, "Boot Camps: A National Assessment," Overcrowded Times 5(4):16. 13. Doris MacKenzie, 1993, "The Impact of Shock Incarceration on Technical Violations and New Criminal Activities," Justice Quarterly 10:465. 14. Dale Parent, 1994, "Boot Camps Failing To Achieve Goals," Overcrowded Times 5(4):8-11. 15. Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. 16. U.S. General Accounting Office, Intermediate Sanctions: Their Impacts on Prison Crowding, Costs, and Recidivism Are Still Unclear (Gaithersburg, MD: U.S. General Accounting Office, 1990). James M. Byrne and April Pattavina, "The Effectiveness Issue: Assessing What Works in the Adult Community Corrections System," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 17. Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. 18. Joan Petersilia, Arthur J. Lurigio, and James M. Byrne, "Introduction: The Emergence of Intermediate Sanctions," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), ix-x. 19. Doris Layton MacKenzie and J. W. Shaw, 1990, "Inmate Adjustment and Change During Shock Incarceration: The Impact of Correctional Boot Camp Programs," Justice Quarterly 7(1):125-50. Joan Petersilia and Susan Turner, "Intensive Probation and Parole," in Crime and Justice: A Review of Research, vol. 17, ed. Michael Tonry (Chicago: University of Chicago Press, 1993), 281-335. Douglas Anglin and Yih-Ing Hser, "Treatment of Drug Abuse," in Drugs and Crime, ed. Michael Tonry and James Q. Wilson, vol. 13 of Crime and Justice: A Review of Research, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1990). President's Commission on Model State Drug Laws, Final Report (Washington, DC: U.S. Government Printing Office, 1993). National Institute of Justice, Drug Use Forecasting (DUF)-1993 Annual Report (Washington, DC: National Institute of Justice, 1994). Paul Gendreau, Francis T. Cullen, and James Bonta, 1994, "Intensive Rehabilitation Supervision: The Next Generation in Community Corrections?" Federal Probation 58:72-78. 20. R. A. Ball, C. R. Huff, and J. R. Lilly, House Arrest and Correctional Policy (Newbury Park, CA: Sage, 1988). 21. Thomas G. Blomberg, William Bales, and Karen Reed, 1993, "Intermediate Punishment: Redistributing or Extending Social Control?" Crime, Law, and Social Change 19:187-201. Marc Renzema, "Home Confinement Programs: Development, Implementation, and Impact," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 41, 46. Terry L. Baumer and Robert I. Mendelsohn, "Electronically Monitored Home Confinement: Does It Work?" in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 54. Robert Lilly, 1993, "Electronic Monitoring in the U.S.: An Update," Overcrowded Times 4(5):4, 15. 22. S. C. Baird and D. Wagner, 1990, "Measuring Diversion: The Florida Community Control Program," Crime and Delinquency 36:112-25. 23. Florida Legislature, An Alternative to Florida's Current Sentencing Guidelines-A Report to the Legislature and the Sentencing Guidelines Commission (Tallahassee, FL: Florida Legislature, Economic and Demographic Research Division, Joint Legislative Management Committee, 1991). 24. Dennis J. Palumbo, Mary Clifford, and Zoann K. Snyder-Joy, "From Net-Widening to Intermediate Sanctions: The Transformation of Alternatives to Incarceration From Benevolence to Malevolence," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 25. Terry L. Baumer, M. G. Maxfield, and R. I. Mendelsohn, 1993, "A Comparative Analysis of Three Electronically Monitored Home Detention Programs," Justice Quarterly 10:121-42. 26. Terry L. Baumer and Robert I. Mendelsohn, "Electronically Monitored Home Confinement: Does It Work?" in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). Marc Renzema, "Home Confinement Programs: Development, Implementation, and Impact," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 49. 27. Terry L. Baumer and Robert I. Mendelsohn, "Electronically Monitored Home Confinement: Does It Work?" in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 66. 28. George Mair, 1993, "Day Centres in England and Wales," Overcrowded Times 4(2):5-7. 29. George Mair, Probation Day Centres (London: H.M. Stationery Office, 1988). 30. Dale Parent, Day Reporting Centers for Criminal Offenders: A Descriptive Analysis of Existing Programs (Washington, DC: National Institute of Justice, 1990). Dale Parent, Jim Byrne, Vered Tsarfaty, Laura Valade, and Julie Esselman, Day Reporting Centers, National Institute of Justice Issues and Practices Report (Washington, DC: U.S. Department of Justice, 1995). John J. Larivee, 1991, "Day Reporting in Massachusetts: Supervision, Sanction, and Treatment," Overcrowded Times 2(1):7-8. Jack McDevitt and Robyn Miliano, "Day Reporting Centers: An Innovative Concept in Intermediate Sanctions," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 31. Dale Parent, Jim Byrne, Vered Tsarfaty, Laura Valade, and Julie Esselman, Day Reporting Centers, National Institute of Justice Issues and Practices Report (Washington, DC: U.S. Department of Justice, 1995), xi-xii. 32. Dale Parent, 1991, "Day Reporting Centers: An Emerging Intermediate Sanction," Overcrowded Times 2(1):6, 8. Jack McDevitt and Robyn Miliano, "Day Reporting Centers: An Innovative Concept in Intermediate Sanctions," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 33. Dale Parent, Jim Byrne, Vered Tsarfaty, Laura Valade, and Julie Esselman, Day Reporting Centers, National Institute of Justice Issues and Practices Report (Washington, DC: U.S. Department of Justice, 1995), xii. 34. A number of small in-house evaluations are cited in John J. Larivee, 1991, "Day Reporting in Massachusetts: Supervision, Sanction, and Treatment," Overcrowded Times 2(1):7-8. Jack McDevitt and Robyn Miliano, "Day Reporting Centers: An Innovative Concept in Intermediate Sanctions," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 35. See, for example, John Doble, Stephen Immerwahr, and Amy Robinson, Punishing Criminals: The People of Delaware Consider the Options (New York: Edna McConnell Clark Foundation, 1991). 36. Warren Young, Community Service Orders (London: Heinemann, 1979). Gill McIvor, Sentenced to Serve: The Operation and Impact of Community Service by Offenders (Aldershot, England: Avebury, 1992). Peter J. P. Tak, 1995, "Netherlands Successfully Implements Community Service Orders," Overcrowded Times 6(2):4, 16-17. 37. Douglas McDonald, Punishment Without Walls: Community Service Sentences in New York City (New Brunswick, NJ: Rutgers University Press, 1986). 38. C. Lloyd, National Standards for Community Service Orders: The First Two Years of Operation (London: Home Office Research and Planning Unit, 1991). 39. Ken Pease, "Community Service Orders," in Crime and Justice: A Review of Research, vol. 6, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1985). 40. Ibid. Gill McIvor, Sentenced to Serve: The Operation and Impact of Community Service by Offenders (Aldershot, England: Avebury, 1992). Gill McIvor, 1993, "CSOs Succeed in Scotland," Overcrowded Times 4(3):1, 6-8. 41. Anton M. van Kalmthout and Peter J. P. Tak, Sanctions-Systems in the Member-States of the Council of Europe: Deprivation of Liberty, Community Service, and Other Substitutes (Boston: Kluwer, 1992). Peter J. P. Tak, 1994, "Sentencing and Punishment in the Netherlands," Overcrowded Times 5(5):5-8. Peter J. P. Tak, 1994, "Sentencing in the Netherlands," Acta Criminologica 7:7-17. 42. Douglas McDonald, Punishment Without Walls: Community Service Sentences in New York City (New Brunswick, NJ: Rutgers University Press, 1986). Douglas McDonald, "Punishing Labor: Unpaid Community Service as a Criminal Sentence," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 43. Sally Hillsman, Joyce Sichel, and Barry Mahoney, Fines in Sentencing: A Study of the Use of the Fine as a Criminal Sanction (Washington, DC: National Institute of Justice, 1984). George F. Cole, Barry Mahoney, Marlene Thornton, and Roger A. Hanson, The Practices and Attitudes of Trial Court Judges Regarding Fines as a Criminal Sanction (Washington, DC: National Institute of Justice, 1987). 44. George F. Cole, "Monetary Sanctions: The Problem of Compliance," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 45. Fahy G. Mullaney, Economic Sanctions in Community Corrections (Washington, DC: National Institute of Corrections, 1988). 46. George F. Cole, Barry Mahoney, Marlene Thornton, and Roger A. Hanson, The Practices and Attitudes of Trial Court Judges Regarding Fines as a Criminal Sanction (Washington, DC: National Institute of Justice, 1987). 47. Peter J. P. Tak, 1994, "Sentencing in the Netherlands," Acta Criminologica 7:7-17. Sally Hillsman and Judith A. Greene, "The Use of Fines as an Intermediate Sanction," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992), 125. Silvia G. Casale, Fines in Europe, Fines in Sentencing Working Paper no. 10 (New York: Vera Institute of Justice, 1981). Norval Morris and Michael Tonry, Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System (New York: Oxford University Press, 1990), chapter 4. 48. Gerhard Grebing, The Fine in Comparative Law: A Survey of 21 Countries, Occasional Paper no. 9 (Cambridge: University of Cambridge, Institute of Criminology, 1982). 49. Silvia G. Casale, Fines in Europe, Fines in Sentencing Working Paper no. 10 (New York: Vera Institute of Justice, 1981). Norval Morris and Michael Tonry, Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System (New York: Oxford University Press, 1990), 144. 50. Thomas Weigend, 1993, "In Germany, Fines Often Imposed in Lieu of Prosecution," Overcrowded Times 4(1):1, 15-16. 51. Peter J. P. Tak, 1994, "Sentencing and Punishment in the Netherlands," Overcrowded Times 5(5):5-8. 52. Sally Hillsman and Judith A. Greene, "The Use of Fines as an Intermediate Sanction," in Smart Sentencing: The Emergence of Intermediate Sanctions, ed. James M. Byrne, Arthur J. Lurigio, and Joan Petersilia (Newbury Park, CA: Sage, 1992). 53. Douglas McDonald, Judith Greene, and Charles Worzella, Day Fines in American Courts: The Staten Island and Milwaukee Experiments: Issues and Practices (Washington, DC: National Institute of Justice, 1992). Susan Turner, 1992, "Day-Fine Projects Launched in Four Jurisdictions," Overcrowded Times 3(6):5-6. 54. Susan Turner and Joan Petersilia, Day Fines in Four U.S. Jurisdictions, RAND No. 1153-NIJ (Santa Monica, CA: RAND, 1996). ------------------------------ Chapter 3 Sentencing Guidelines More than one-third of States had some form of sentencing guidelines in place by mid-1996. To someone new to sentencing policy that descriptive statement is neither striking nor interesting. However, people who have several decades' perspective know that it encapsulates developments that are radical in the context of what are typically glacial legal system changes. Sentencing guidelines have endured because in some places they have achieved their proponents' major aims- comprehensive sentencing policies have been adopted; unwarranted disparities in general, and racial and gender disparities in particular, have been reduced; and sentencing policies have been effectively linked with corrections resources. Still, guidelines have not solved all sentencing problems. No guidelines system has as yet devised a way to prevent manipulation by prosecutors. Some of the early systems dealt only with felonies, which meant that judges retained complete and unreviewable sentencing discretion over a large proportion of all convicted offenders. None of the early systems set standards for punishments other than imprisonment. Guidelines created presumptions as to whether confinement was appropriate and, if so, for how long, but if confinement was not ordered, judges retained full discretion over nonincarcerative punishments. None of the early systems created standards governing choices among nonincarcerative punishments and between nonincarcerative punishments and confinement. Legislation to establish the first two sentencing commissions, in Minnesota and Pennsylvania, was enacted in 1978. By 1987, presumptive guidelines created by sentencing commissions were in place in Minnesota, Washington, and (after an initial legislative rejection) Pennsylvania. Voluntary guidelines in Florida, on which work began in the late 1970's, were converted into presumptive guidelines.[1] The pace of sentencing commission activity increased after the mid-1980's. The Federal legislation was passed in 1984, commissioners were appointed in 1985, and the guidelines took effect in 1987. Oregon's guidelines took effect in 1989. After 1990, guidelines created by sentencing commissions took effect in Arkansas, Louisiana, North Carolina, Tennessee, Kansas, Missouri, and Ohio; Louisiana's were later rescinded. New guidelines systems will likely be adopted when sentencing commissions, active in 1997, in Massachusetts, Maryland, Michigan, South Carolina, Oklahoma, and Montana complete their work. Effects on Sentencing Practices Guidelines developed by commissions have changed sentencing practices and patterns, reduced disparities, ameliorated racial and gender differences, and helped States control their prison populations. Racial and gender differences have been reduced but, when current offenses and criminal histories are taken into account, women and whites continue to receive mitigated sentences more often than men and blacks, and, conversely, men and blacks more often receive aggravated sentences. Finally, although some jurisdictions have managed to regulate their prison populations, others have not yet tried. Changing Sentencing Decisions Data are available from four States and the Federal system on judges' compliance with guidelines. In a large majority of cases, judges conform the sentences they announce to applicable sentencing ranges. However, this assertion requires two important caveats. First, guideline developers have often insisted that guidelines should be disregarded when a case's special features warrant different treatment; they are, after all, guidelines, not mandatory penalties, and a judge's reasons for imposing some other sentence can be stated for the record and later examined by higher courts. From this perspective, a guidelines system that elicited 100 percent compliance would be undesirable because it would appear that judges were not distinguishing among cases as they should. Second, if judges freely allow plea bargaining, compliance may be more apparent than real. Guidelines make sentencing predictable. Imagine a guidelines grid as a dart board. For example, figure 2 shows the initial 1989 Oregon grid. Each cell specifies a range of presumptively appropriate sentences; to "fix" the game, counsel need only be sure that the dart hits the right cell. This can be done by means of charge dismissals. If counsel negotiate a 15-month prison sentence, and the applicable sentencing range for offense X is 15 to 16 months, the defendant need only plead guilty to X, and the prosecutor will dismiss any other charges to ensure that the agreed-upon sentence is given. Considerable evidence suggests that counsel do bargain around guidelines. Comparisons of plea bargaining in Minnesota before and under guidelines showed a marked shift away from sentence bargaining and toward charge bargaining. Frase, on the basis of a quantitative analysis of Minnesota sentencing patterns through 1989, concluded, "It appears likely that whatever plea-trial disparities there were before the guidelines went into effect continued to exist in the early post-guidelines years, and still exist today; plea bargaining, and its accompanying charge and sentence disparities, is 'alive and well' in Minnesota." Similar evidence is available from Washington and Pennsylvania.[2] Recalling, therefore, that compliance rates may be misleading, figure 3 shows compliance and departure rates for selected recent years in Minnesota, Oregon, Pennsylvania, Washington, and the Federal system. Departure rates are low in every jurisdiction. Judges sworn to enforce the law are presumably more comfortable not departing from guidelines that the legislature has adopted or ratified. Reducing Disparities Every sentencing commission claims that its guidelines have reduced disparities compared with sentencing patterns before guidelines took effect. Research on disparities, however, faces a number of formidable problems. First, promulgation of guidelines may affect plea bargaining and before- and-after comparisons may be confounded if charging and bargaining practices change with the guidelines. Second, as was true during most of the 1980's, if public and officials' attitudes toward offenders have become more punitive over time, sentences are likely to have become harsher with or without guidelines, and that rising tide complicates disparity analyses. Most of the analyses discussed below are efforts to compare sentencing disparities in the first year, or in the first few years, under a guidelines system with sentencing trends in some earlier period. As time goes by, it becomes increasingly difficult, and soon impossible, to reach conclusions about disparities. Political and policy environments change over time, and these changes also alter sentencing patterns, regardless of guidelines. Disparity reduction in the States. There are plausible grounds for believing that the State guidelines in their early years reduced disparities. In Minnesota, where more evaluations have been conducted than anywhere else, an evaluation of the first 3 years' experience concluded, "Disparity in sentencing decreased under the sentencing guidelines. This reduction in disparity is indicated by increased sentence uniformity and proportionality."[3] Outside evaluators agreed: Minnesota "was largely successful in reducing pre-guideline disparities in those decisions that fall within the scope of the guidelines."[4] Frase, drawing on data for 1981 to 1989, concluded that "the Minnesota guidelines have achieved, and continue to achieve," most of their goals, including disparity reduction.[5] No independent evaluations have been published concerning the Oregon, Pennsylvania, Washington, and Delaware guidelines, but each commission, relying on its regularly collected monitoring data, has concluded that disparities declined. The Washington commission, relying on 1985 data, concluded that "the Sentencing Reform Act has clearly increased consistency in the imprisonment decision."[6] Looking back over the first 6 years' experience with guidelines, the Washington commission reported that "the high degree of compliance with sentencing guidelines has reduced variability in sentencing among counties and among judges."[7] In the early years of the Pennsylvania guidelines, its commission concluded for 1983, "It appears that Pennsylvania's guidelines are accomplishing their intended goal of reducing unwarranted sentencing disparity,"[8] and for 1984, "sentences became more uniform throughout the state." Kramer and Lubitz agreed.[9] Finally, reporting on the first 15 months' guidelines experience in Oregon, its commission concluded that "the guidelines have increased uniformity in sentencing considerably. Dispositional variability for offenders with identical crime seriousness and criminal history scores has been reduced by 45 percent over the variability under the pre-guidelines system." Later analyses concur.[10] In Delaware, no evaluation has been published of the effects of its voluntary guidelines on disparities. However, a number of publications by Delaware officials, based on data from the State's Statistical Analysis Center,[11] list "consistency and certainty" among the guidelines' goals and present data showing that the guidelines have succeeded in increasing use of incarceration for violent offenders and use of intermediate punishments for nonviolent offenders. At the very least, this arguably supports an inference of greater consistency in Delaware sentencing. Most likely, sentencing guidelines reduced disparities in most of these jurisdictions compared with what they would have been without guidelines. Because presumptive guidelines set standards for sentences where none existed, conformity rates would have to be low in order for the guidelines to have no effect on sentencing decisions. Even when plea bargaining is taken into account, the bargaining takes place in the shadow of the guidelines, and it is likely that the bargained sentences are more consistent than they otherwise would have been. Disparity reduction in the Federal system. The evidence on Federal sentencing disparities is mixed, and the best conclusion at present is that we do not know whether disparities have increased or decreased.[12] Although the U.S. commission, on the basis of an evaluation of the first 4 years' experience with Federal guidelines, claims "the data . . . show significant reductions in disparity,"[13] there is reason to doubt that conclusion. The U.S. General Accounting Office (GAO), which at congressional behest reviewed the commission's study and conducted its own statistical analyses, was more doubtful.[14] The commission and GAO approached disparities in two ways-by asking participants whether they believed disparities had been reduced and by conducting sophisticated quantitative analyses. Figure 4 shows answers to questions about reduced disparities from GAO interviews in 4 sites, commission interviews in 12 sites, and a commission mail survey. Most judges and defense counsel did not believe disparities were reduced; however, significant and slight majorities of prosecutors and probation officers disagreed. An earlier analysis of the guidelines by the Federal Courts Study Committee reported that many judges and the committee itself believed that disparities had increased.[15] The quantitative analyses were also inconclusive. The GAO's conclusion, based both on examination of the U.S. commission's statistical analyses and reanalysis of the commission's data, is that "limitations and inconsistencies in the data available for pre-guidelines and guideline offenders made it impossible to determine how effective the sentencing guidelines have been in reducing overall sentencing disparity."[16] Sentencing Patterns Most commissions have adopted guidelines intended to change existing sentencing patterns. Minnesota and Washington sought to increase use of imprisonment for violent offenses and to decrease it for property offenses; the States assert that their monitoring data show these objectives were achieved in the guidelines' early years.[17] Oregon had the same goals and found that the proportion of offenders convicted of felonies against persons who received State prison sentences increased from 34 percent before guidelines to 48 percent under guidelines, whereas the proportion of property felons sentenced to State prisons declined from 19 to 9 percent; imprisonment for sex abuse felonies tripled from 13 to 42 percent. More recent data show a similar pattern.[18] Pennsylvania sought to increase sentencing severity and appears to have succeeded; for most serious crimes, the proportion of convicted offenders incarcerated and their average minimum sentences before parole eligibility increased after the guidelines took effect.[19] Delaware policymakers also sought to increase prison use for violent offenders and appear to have succeeded. Although published monitoring data are less detailed than elsewhere, figure 5 shows that the proportion of violent offenders among Delaware prisons increased after the guidelines took effect and the proportion of nonviolent offenders decreased.[20] David Boerner has shown how changes in guideline severity adopted in Washington were quickly followed by increases in average sentence severity. Figure 6 shows sentencing patterns from 1988 to 1992 for second-degree burglaries committed before and after the effective date of guideline amendments that divided second-degree burglary into residential and nonresidential types and increased penalties for each. Average sentences for residential burglary increased substantially, whereas sentences for grandfathered cases (both nonresidential and residential) increased only slightly. Boerner gives many such examples, all of which tend to demonstrate that changes in guidelines for sentence severity were quickly followed by increases in the severity of sentences for affected offenses.[21] Sentencing patterns have changed substantially in North Carolina since guidelines took effect in late 1994. The guidelines were intended to increase the likelihood of imprisonment for people convicted of violent crimes. In 1995, the first full calendar year under guidelines, 81 percent of violent offenders received prison sentences, up from 67 percent in 1993. Further, the North Carolina commission sought to reduce use of prison sentences for nonviolent crimes and achieved this: the percentage of nonviolent offenders sentenced to prison fell from 42 percent in 1993 to 23 percent in 1995. Finally, the commission sought to increase the length of sentences for violent crimes; this, too, was accomplished: in 1993 violent offenders received an average of 56 months, whereas in 1995, they received 87 months.[22] The U.S. Sentencing Commission sought to increase the proportion of offenders sentenced to imprisonment and to increase the average length of prison sentences; it succeeded in both objectives. The commission's self-evaluation showed that, overall, the percentage of convicted Federal offenders sentenced to probation declined from 52 percent in late 1984 to 35 percent in June 1990. However, reduction in the use of probation was even greater than the commission reports. The commission's numbers overstate the use of probation by including split sentences that included a period of incarceration. According to the commission's 1993 annual report, 22.7 percent of cases resulted in "probation" sentences, but only 14.8 percent resulted in " 'straight' probation" (that is, no confinement condition).[23] The severity of Federal prison sentences likewise increased. The commission reports that the mean sentence "expected to be served" for all offenders increased from 24 months in July 1984 to 46 months in June 1990. Prison sentences for drug offenses increased by 248 percent from 1984 to 1990 and from an average of 60 months for robbery in 1984 to 78 months in 1990.[24] Racial and Gender Differences Every sentencing commission has included among its goals, the reduction or elimination of racial and gender discrimination in sentencing, and most claim to have succeeded to a considerable extent. The Minnesota commission's 3-year evaluation concluded that racial differences in sentencing declined under guidelines; nonetheless, minority defendants were likelier than whites to be imprisoned when the presumptive sentence did not prescribe State imprisonment; minority defendants received longer sentences than similarly categorized whites; and men received longer prison sentences than similarly categorized women. Miethe and Moore, using the same data but more sophisticated statistical techniques, agreed. Frase, using the commission's monitoring data through 1989, also agreed.[25] Washington's experience revealed similar patterns. The initial evaluation found an overall decline in racial differences in sentencing but reported that a "substantial racial and gender disparity was found in the use of sentencing alternatives;" whites were almost twice as likely as blacks to benefit from special mitigating provisions for first-time and some sex offenders.[26] In a 10-year review of Washington sentencing reforms, although concluding that racial and gender differences had diminished, the commission in 1992 acknowledged "significant gender and ethnic differences in the application of options" to incarceration.[27] The Oregon racial data from the first 15 months of guidelines experience also show declines in racial and gender disparities. Whites were slightly less likely than minority defendants to receive "aggravated departures," slightly more likely to receive "mitigated departures," and much more likely to benefit from an "optional probation" alternatives program. As whites are to minority defendants in Oregon sentencing, so women are to men: women were less likely to receive upward departures, more likely to receive downward departures, and more likely to be sentenced to optional probation.[28] The research design of the U.S. Sentencing Commission's self-evaluation and the GAO reanalysis precluded any overall conclusions about racial and gender disparities.[29] Tying Policy to Resources The Oregon, Washington, and Minnesota commissions operated under enabling legislation which directed them to give substantial consideration to the impact of their guidelines on correctional resources, which was generally interpreted to refer to prison beds and capacity. All three States managed to hold prison populations within capacity for extended periods. Figure 7 shows prison population increases in Minnesota, Washington, and Oregon compared with California and the Federal System from December 31, 1997 to June 30, 1994. The Minnesota and Washington rates increased much more slowly than elsewhere in the years after guideline adoption. In both States, the legislature in the late eighties increased penalties sharply for many crimes;[30] both prison populations thereafter rose rapidly. Oregon's prison population decreased after guidelines took effect (though the effects of a 1994 voter referendum that greatly increased sentences for many crimes will likely result in faster prison growth). Endnotes 1. Deborah M. Carrow, Judith Feins, Beverly N. W. Lee, and Lois Olinger, Guidelines Without Force: An Evaluation of the Multi-Jurisdictional Sentencing Guidelines Field Test, Report to the National Institute of Justice (Cambridge, MA: Abt Associates, 1985). 2. Kay A. Knapp, The Impact of the Minnesota Sentencing Guidelines: Three Year Evaluation (St. Paul, MN: Minnesota Sentencing Guidelines Commission, 1984), chapter 6. Richard Frase, 1993, "Implementing Commission-based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota," Cornell Journal of Law and Public Policy 2:279-337. Washington State Sentencing Guidelines Commission, Sentencing Practices Under the Sentencing Reform Act (Olympia, WA: Washington State Sentencing Guidelines Commission, 1985). Michael Tonry, Sentencing Matters (New York: Oxford University Press, 1996), table 2-2 and table 2-3. Pennsylvania Commission on Sentencing, 1983 Annual Report: Sentencing in Pennsylvania (State College, PA: Pennsylvania Commission on Sentencing, 1984), table 1. Pennsylvania Commission on Sentencing, 1986-1987 Annual Report (State College, PA: Pennsylvania Commission on Sentencing, 1987), table 4. 3. Kay A. Knapp, The Impact of the Minnesota Sentencing Guidelines: Three Year Evaluation (St. Paul, MN: Minnesota Sentencing Guidelines Commission, 1984), v-vi. 4. Terance D. Miethe and Charles A. Moore, 1985, "Socioeconomic Disparities Under Determinate Sentencing Systems: A Comparison of Preguideline and Postguideline Practices in Minnesota," Criminology 23: 337-363. 5. Richard Frase, 1993, "Implementing Commission-based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota," Cornell Journal of Law and Public Policy 2:279-337. 6. Washington State Sentencing Guidelines Commission, Preliminary Evaluation of Washington State's Sentencing Reform Act (Olympia, WA: Washington State Sentencing Guidelines Commission, 1986), 7. 7. Washington State Sentencing Guidelines Commission, A Decade of Sentencing Reform: Washington and Its Guidelines, 1981-1991 (Olympia, WA: Washington State Sentencing Guidelines Commission, 1992), 12. 8. Pennsylvania Commission on Sentencing, 1983 Annual Report: Sentencing in Pennsylvania (State College, PA: Pennsylvania Commission on Sentencing, 1984), i. 9. Pennsylvania Commission on Sentencing, 1984 Annual Report: Sentencing in Pennsylvania (State College, PA: Pennsylvania Commission on Sentencing, 1985), i. John H. Kramer and Robin L. Lubitz, 1985, "Pennsylvania's Sentencing Reform: The Impact of Commission-Established Guidelines," Crime and Delinquency 31: 481-500. 10. Kathryn Ashford and Craig Mosbaek, First Year Report on Implementation of Sentencing Guidelines: November 1989 to January 1991 (Portland, OR: Oregon Criminal Justice Council, 1991), viii. Kathleen Bogan and David Factor, 1995, "Oregon Guidelines, 1989-1994," Overcrowded Times 6(2):1, 13-15. 11. For example, see Richard S. Gebelein, 1996, "SENTAC Changing Delaware Sentencing," Overcrowded Times 7(4):1, 9-11. 12. William Rhodes, 1992, "Sentence Disparity, Use of Incarceration, and Plea Bargaining: The Post- Guideline View From the Commission," Federal Sentencing Reporter 5:153-55. David Weisburd, 1992, "Sentencing Disparity and the Guidelines: Taking a Closer Look," Federal Sentencing Reporter 5:149-52. Michael Tonry, 1993, "The Failure of the U.S. Sentencing Commission's Guidelines, " Crime and Delinquency 39:131-49. 13. U.S. Sentencing Commission, The Federal Sentencing Guidelines: A Report on the Operation of the Guidelines System and Short-Term Impacts on Disparity in Sentencing, Use of Incarceration, and Prosecutorial Discretion and Plea Bargaining (Washington, DC: U.S. Sentencing Commission, 1991), 419. 14. U.S. General Accounting Office, Sentencing Guidelines: Central Questions Remain Unanswered (Washington, DC: U.S. General Accounting Office, 1992). 15. Federal Courts Study Committee, Report (Washington, DC: Administrative Office of the U.S. Courts, 1990). 16. U.S. General Accounting Office, Sentencing Guidelines: Central Questions Remain Unanswered (Washington, DC: U.S. General Accounting Office, 1992), 10. 17. Kay A. Knapp, The Impact of the Minnesota Sentencing Guidelines: Three Year Evaluation (St. Paul, MN: Minnesota Sentencing Guidelines Commission, 1984), 31. Washington State Sentencing Guidelines Commission, Preliminary Evaluation of Washington State's Sentencing Reform Act (Olympia, WA: Washington State Sentencing Guidelines Commission, 1986), figure 1. These analyses are examined in some detail in Michael Tonry, "Structuring Sentencing," in Crime and Justice: A Review of Research, vol. 10, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1988), 306-309. 18. Kathryn Ashford and Craig Mosbaek, First Year Report on Implementation of Sentencing Guidelines: November 1989 to January 1991 (Portland, OR: Oregon Criminal Justice Council, 1991), viii, 11. Kathleen Bogan and David Factor, 1995, "Oregon Guidelines, 1989-1994," Overcrowded Times 6(2):1, 13-15 (table 1). 19. Pennsylvania Commission on Sentencing, 1986-1987 Annual Report (State College, PA: Pennsylvania Commission on Sentencing, 1987), table 19. 20. Richard S. Gebelein, 1996, "SENTAC Changing Delaware Sentencing," Overcrowded Times 7(4):1, 9-11. 21. David Boerner, 1993, "The Legislature's Role in Guidelines Sentencing in 'The Other Washington,'" Wake Forest Law Review 28:381-420. 22. Robin L. Lubitz, 1996, "Sentencing Changes in North Carolina," Overcrowded Times 7(3):1, 12-15. 23. U.S. Sentencing Commission, Annual Report-1993 (Washington, DC: U.S. Sentencing Commission, 1994), 67. 24. U.S. Sentencing Commission, The Federal Sentencing Guidelines: A Report on the Operation of the Guidelines System and Short-Term Impacts on Disparity in Sentencing, Use of Incarceration, and Prosecutorial Discretion and Plea Bargaining (Washington, DC: U.S. Sentencing Commission, 1991), 378. 25. Kay A. Knapp, The Impact of the Minnesota Sentencing Guidelines: Three Year Evaluation (St. Paul, MN: Minnesota Sentencing Guidelines Commission, 1984), 61. Terance D. Miethe and Charles A. Moore, 1985, "Socioeconomic Disparities Under Determinate Sentencing Systems: A Comparison of Preguideline and Postguideline Practices in Minnesota," Criminology 23: 337-363. Richard Frase, 1993, "Implementing Commission-Based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota," Cornell Journal of Law and Public Policy 2:279-337. 26. Washington State Sentencing Guidelines Commission, Preliminary Statistical Summary of 1986 Sentencing Data (Olympia, WA: Washington State Sentencing Guidelines Commission, 1987), 59. 27. Washington State Sentencing Guidelines Commission, A Decade of Sentencing Reform: Washington and Its Guidelines, 1981-1991 (Olympia, WA: Washington State Sentencing Guidelines Commission, 1992), 12. 28. Kathryn Ashford and Craig Mosbaek, First Year Report on Implementation of Sentencing Guidelines: November 1989 to January 1991 (Portland, OR: Oregon Criminal Justice Council, 1991), 47-52. 29. U.S. Sentencing Commission, The Federal Sentencing Guidelines: A Report on the Operation of the Guidelines System and Short-Term Impacts on Disparity in Sentencing, Use of Incarceration, and Prosecutorial Discretion 0and Plea Bargaining (Washington, DC: U.S. Sentencing Commission, 1991), 302, 324. 30. Richard Frase, 1993, "Implementing Commission- Based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota," Cornell Journal of Law and Public Policy 2: 279-337. Roxanne Lieb, 1993, "Washington Prison Population Growth Out of Control," Overcrowded Times 4(1):1, 13-14, 20. ------------------------------ Chapter 4 Incorporating Intermediate Sanctions in Sentencing Guidelines Integrating intermediate sanctions effectively requires establishing both a graduated array of punishments between prison and probation and a system for appropriately distributing offenders among them. Studies on creating and operating cost- effective intermediate sanctions (see chapter 2) and on creating and operating systems of presumptive sentencing guidelines that effectively structure judicial decisions about confinement (see chapter 3) already exist. However, little is known about combining the two approaches. Obstacles Intermediate sanctions have not been overlooked by sentencing commissions or by drafters of guidelines-enabling legislation. Section 9(5)(2) of the statute creating Minnesota's commission authorized establishment of nonincarceration guidelines: "The sentencing guidelines promulgated by the commission may also establish appropriate sentences for prisoners for whom imprisonment is not proper. Any [such] guidelines . . . shall make specific reference to noninstitutional sanctions including but not limited to the following: payment of fines, day fines, restitution, community work orders, work release programs in local facilities, community-based residential and nonresidential programs, incarceration in a local correctional facility, and probation and the conditions thereof." The Minnesota commission's guidelines created presumptions as to who among convicted felons should be sent to State prison (roughly 20 percent) and for how long, but set no presumptions for sentences for nonimprisonment sanctions for felons or for sentences of any kind for misdemeanants. For a variety of reasons, guidelines for use of nonincarcerative punishments run into special problems. These include a shortage (or, in some places, the absence) of credible, well-managed intermediate sanctions and instinctive judicial resistance to proposals for nonincarcerative guidelines. Both problems are soluble. Effective intermediate sanctions can be established. Many judges have been persuaded that presumptive imprisonment guidelines improve the quality of sentencing generally, and there should be no reason why they cannot be persuaded of the merits of nonincarcerative guidelines. The challenge is to do both things simultaneously and that has proven difficult. Efforts to create comprehensive nonincarceration guidelines are bound to fail if adequate intermediate sanctions do not exist. Likewise, from a policy perspective, intermediate sanctions are often bound to be misused in a jurisdiction that lacks some means to regulate judges' decisions. The need for simultaneity has not gone unnoticed by policymakers. North Carolina, as noted below, has moved as far as any State toward structured use of intermediate sanctions. In the statutory background were both enabling legislation to create (1) the Sentencing Policy and Advisory Commission and to adopt guidelines, and (2) the State-County Criminal Justice Partnership Act, which encourages and provides financial incentives for creating new county-level programs. Pennsylvania has also moved both to include intermediate sanctions in its guidelines and to foster and fund new community- based programs. Still, there remains another difficult problem: many judges believe that, in principle, guidelines are incompatible with the mildly-to-moderately serious crimes for which intermediate sanctions are most appropriate. While fairly simple systems for proportioning prison time to crime severity may work for the most serious crimes, more considerations-appropriate treatment conditions, effects on the offender's family and employment, the judge's reasons for imposing a particular sentence, the aggregate weight of multiple work, restriction on liberty, and monetary conditions-are often seen as relevant for less serious crimes and cannot easily be encapsulated in a guidelines grid. This too is a soluble problem, but the widespread perception that community sentencing is too complicated and inherently too individualized to be subjected to general rules is likely to prove a formidable obstacle. Efforts at Integrating Intermediate Sanctions Into Sentencing Guidelines In many States, commissions are at work on proposals to integrate intermediate and noncustodial penalties into guidelines and to devise systems ofinterchangeability between prison and nonprison sanctions. Three sets of interrelated issues must be considered. First, should guidelines permit judges to choose between incarcerative and nonincarcerative sanctions, and if so, to what extent? Second, how are choices among different kinds of nonincarcerative sanctions to be made? Third, how authoritative ought guidelines to be about intermediate sanctions? These questions are discussed in the following sections. Because little research or policy discussion has focused on the latter two issues, most of the discussion concerns the first. Three devices have been used to authorize judges to choose between prison and nonprison sentences for cases that fall within a single guideline cell. The first creates cells in guidelines grids that expressly authorize judges to choose between sentencing options. The second is to establish interchangeability policies that allow judges to substitute equivalently burdensome punishments for imprisonment. The third is to create "categorical exceptions" that allow judges to disregard otherwise applicable guidelines for qualifying offenders. These usually involve boot camps, first offenders, or sex offenders. Delaware's unique system of guidelines, which sets five "sanctioning levels," are also discussed. Because they are voluntary guidelines, the interchangeability question does not arise. However, because they establish a continuum of sanctions of graded severity, they warrant mention. Interchangeability Between Prison and Nonprison Sentences Every guidelines system allows for interchangeability between prison and nonprison sentences, although the extent ofinterchangeability varies widely. "Just deserts"arguments have been made asserting that such interchangeability should never or only seldom be permitted because sanctions vary fundamentally in their character.[1] If punishment is largely about attribution of blameworthiness, the argument goes, punishment should be closely proportioned to the seriousness of the crime. Punishments are qualitatively different, and permitting substitutions among them obscures differences in offenders' blameworthiness. Thus, the argument concludes, guidelines incorporating different kinds of punishments should permit little or no overlap in their use. Figure 8 shows what such a system might look like. The most serious crimes are at the top of the pyramid, and for them only full-time incarceration would be authorized. In the next lower tier, partial incarceration such as day or night confinement, house arrest, work release, or day-reporting would be permitted. The third tier might include intensive forms of supervision, the fourth substantial fines, the fifth standard probation, and the sixth minor fines and community service. Within each tier, choices could be made only between sanctions that were equivalently burdensome, and imposition of punishments from differenttiers on comparable offenders ordinarily would be forbidden. Thus, for particular defendants, judges would seldom be permitted to choose between incarcerative and nonincarcerative sentences. To be realistic, figure 8 would need to be developed in more detail. How that might be done is shown in figure 9. Within offense severity levels, for example, sublevels could specify ranges of allowable sentence durations or amounts for different offenses. Similarly, consideration of prior criminal history could be done in various ways. Figure 9 does this by indicating criminal record categories atop each cell. At the grid's top, where offense severity is the primary consideration, the relative weight of prior records is small. At the grid's base, where offense severity is less, the weight of prior records and discrimination among them are greater. No jurisdiction has adopted a system like those set out in figures 8 and 9. Plausible arguments can be made that their premise-that blameworthiness measured only in terms of current and past crimes is the only valid calibrator of sentences-is oversimplified. In any event, every existing guidelines system does permit some interchangeability between incarcerative and nonincarcerative punishments. Residual interchangeability. Minnesota's guidelines, for example, permit interchangeability in three ways. (Figure 10 shows Minnesota's grid as it was in 1985.) First, for any case falling into a cell above the bold black line, judges have broad discretion to choose among a jail term of up to 12 months, any combination of nonincarcerative punishments, and split-sentences combining jail time with other penalties. This is no small power since 80 to 85 percent of felony defendants fall within cells above the bold line. Moreover, the guidelines do not cover misdemeanors, so judges have comparable discretion over them. Second, because Minnesota's guidelines are presumptive, judges have authority in every case not governed by a statutory mandatory minimum-if they document their reasons-to depart from recommended prison sentences and to impose a nonincarcerative sentence or a split-sentence in its place. Judges do this in about one-third of the cases for which imprisonment is the presumptive sentence, [2] just as they impose prison sentences in a smaller percentage of presumptive nonprison cases. The bold black line is the product of difficult policy choices made by the Minnesota Sentencing Commission, and, inevitably, there are many cases that fall in cells on either side of it that elicit judicial ambivalence. Third, judges and lawyers can negotiate sentences different from those stipulated by the guidelines. Sometimes this involves substitution of a nonincarcerative penalty for a lengthy presumptive prison sentence in a case where there are no valid grounds for a departure (for example, because the State supreme court has expressly held such considerations insufficient). Some people may consider departures of this sort an inappropriate circumvention of guidelines, but both experience and research instruct that they are not uncommon.[3] Although "illicit departures" are always possible, this special type of interchangeability is not discussed again in the rest of this report. Most of the early presumptive guidelines systems gave judges comparable discretion over interchangeability decisions. In Oregon, as in Minnesota, guidelines cover only felonies, and 18 to 20 percent of convicted felons are sentenced to State prison.[4] Pennsylvania's guidelines cover misdemeanors, but as figure 11 (the August 1991 version of Pennsylvania's guidelines grid) shows, both incarcerative and nonincarcerative punishments were authorized for most misdemeanors and the less serious felonies, meaning that Pennsylvania judges had about the same authority to choose between incarceration and nonincarceration as did Oregon and Minnesota judges. Limited interchangeability. The Federal sentencing guidelines provide for only very limited use of intermediate sanctions or interchangeability; probation and prison are the only alternatives. Fines are not authorized as sole penalties for individuals; nor are intermediate sanctions such as community service, house arrest, or intensive supervision probation-these may be ordered only as conditions of probation. Figure 12 shows the Federal grid as it existed on November 1, 1993. It applied to all Federal felonies and misdemeanors. Confinement was authorized for every offender. Only for the first 8 (of 43) offense levels (Zone A), where sentencing ranges started at zero, did judges sometimes have complete discretion to choose between prison and probation. In 1993, of 34,642 cases on which the commission received complete guideline application information, only 13.7 percent fell within those 8 levels.[5] In levels 9 and 10 (Zone B), judges could sometimes substitute partial, community, or home confinement on a day-for-day basis for total incarceration, for a period not less than the minimum specified period. In levels 11 and 12 (Zone C), some substitution was permitted, but at least half of the guideline minimum had to be served in total confinement. (The preceding describes rules for offenders in the lowest criminal history cateory; as the lines defining Zones A, B, and C show, for offenders with ampler criminal histories,judges had less discretion.) Judges were allowed to depart from the guidelines (though the permitted grounds for departures were much narrower than in most State systems). Even taking departures into account, in 1994 only 14.4 percent of sentenced offenders received a probation sentence without a confinement condition; another 7.8 percent received probation with a confinement condition.[6] Bounded interchangeability. In most jurisdictions, the vast majority of convicted felons and misdemeanants are not sentenced to State prison. By the late 1980's, it was widely recognized that achievement of sentencing reform goals required that nonincarcerative penalties be brought within the scope of guidelines. This was equally evident whether the goals were idealistic (for example, reducing sentencing disparity, or imposing the least restrictive appropriate alternative) or managerial (for example, improving predictability and with it resource planning). The first recommendation that received attention was to replace the Minnesota/Washington "in or out" approach-in which guidelines cells either specified a range of authorized prison sentences or accorded the judge complete authority to choose between confinement and nonconfine-ment sentences-with a larger number of bounded choices. District of Columbia. The prototype was developed by the District of Columbia Superior Court Sentencing Commission. Figure 13 (on previous page) shows the proposed grid for unarmed offenses. It is divided into four zones. For offenses falling in cells marked with an "a," the sentence is to be served in the community (including probation, restitution, fines, community service). In cells marked with a "b," community sentences are presumptively appropriate, but incarceration may be ordered if the judge states for the record the reason for selecting some other sentence. In cells marked with a "c," both incarcerative and community sentences are presumptively appropriate, and the judge may impose either without being required to provide special justification. In the remaining cells, the presumption is for imposition of a prison sentence from within a narrow range of authorized sentence lengths; a community sentence would be considered a departure and would require that thejudge explain his decision. Pennsylvania. In 1994, Pennsylvania implemented revised guidelines that follow the District of Columbia's approach. Figure 14 shows the Pennsylvania guidelines for felonies and misdemeanors occurring on or after August 12, 1994. They create four zones of discretion which are identified by shading. 1. Cells in Level 1 provide for "restorative sanctions" (RS) such as standard probation, community service, and restitution. 2. Cells in Level 2, although they vary in detail, in general authorize judges to choose among restorative sanctions, "restrictive intermediate punishments" (RIPs), and short jail terms. RIPs involve full or partial confinement (for example, inpatient drug treatment, day-reporting centers, halfway houses) or intensive community penalties (for example, house arrest or intensive supervision probation with electronic monitoring). If confinement is required, policy statements recommend a treatment component. If only RS or RIPs are authorized, policy statements recommend RS. Level 2 encompasses many nonviolent crimes and some less serious violent crimes. 3. Cells in Level 3 provide for total or partial confinement or for RIPs. The guideline ranges for confinement set outer limits on RIP sentence length. Judges are free to choose among the different kinds of punishments. Policy statements encourage judges to consider restitution to the victim or rehabilitation of the offender as primary goals and point out that partial confinement coupled with work release and restitution or inpatient drug treatment are authorized means to those goals. 4. Cells in Level 4, which primarily apply to offenders convicted of major violent or drug offenses, often with prior violent crime records, provide for presumptive minimum prison terms to be served before parole eligibility. Compared with the Federal guidelines, Pennsylvania continues to delegate substantial discretion to the sentencing judge over the choice of sentence and allows much greater scope for nonconfinement sentences. Figure 15 shows how Pennsylvania offenders sentenced in 1992 would have been distributed among cells in the 1994 guidelines grid had it then existed. The 4 Level 1 cells, which authorize only RS and preclude any confinement, govern sentencing of 6,879 offenders or 17 percent of the total. The 16 Level 2 cells, all of which authorize RS or RIPs, and some of which also authorize confinement of 3 or 6 months, govern sentencing of 17,261 offenders or 43 percent of the total. Of these, 8,944 (22 percent) fall into cells in which only RS or RIPs are authorized. Only 5,512 offenders (14 percent) fall within Level 4 cells in which total confinement is the only presumptively appropriate sentence. Compared with the Federal guidelines, Pennsylvania's mechanically simpler guidelines represent a more complex philosophy of sentencing. Confinement is not the only punishment available for most offenders. Judges have substantial discretion to choose among different kinds of punishments. Even within a single level, judges may individualize sentences depending on how they weigh restorative, rehabilitative, and retributive considerations. North Carolina. North Carolina is the first State to attempt from the outset to include in its guidelines standards for felonies and misdemeanors and for incarcerative and nonincarcerative punishments. Pennsylvania incorporated these levels of punishment eventually, but 13 years after its initial guidelines took effect. The North Carolina guidelines took effect October 1, 1994. North Carolina's guidelines resemble Pennsylvania's but are more distinct than they may first appear. Figure 16 shows the grid for felony sentencing. As in Pennsylvania (see figure 11), three ranges of presumptive lengths of prison sentences are shown- standard, mitigated, and aggravated. Also as in Pennsylvania, interchangeability is provided by use of zones of discretion. In other ways the guidelines are substantially different. Pennsylvania's guidelines set minimum parole eligibility dates, whereas North Carolina abolished parole release and "good time." North Carolina's guidelines thus prescribe time to be served. More importantly, North Carolina's guidelines are much more restrictive of judicial discretion. A Pennsylvania judge who departs from the guidelines need only "provide a contemporaneous written statement of the reason or reasons. " There is no general evidentiary test that must be met, and appellate courts tend to use a deferential "abuse of discretion" standard inconsidering sentencing appeals. In North Carolina, if the guidelines specify a prison sentence, judges must set a term from within the authorized range unless, for less serious cases, the court finds "that extraordinary mitigating factors of a kind greater than the normal case exist and that they substantially outweigh any factors in aggravation." In addition, the court must also find that imposition of a prison sentence would be a "manifest injustice." Even then, the possibility of an intermediate punishment is forbidden for all drug traffickers, offenders convicted of murder or first-degree rape, and offenders with any significant prior record. North Carolina recognizes three types of sentences: active punishments (immediate total confinement), intermediate punishments (split-sentences, residential programs, electronic house arrest, and intensive supervision probation), and community punishments (supervised or unsupervised probation, community service, outpatient treatment programs, fines). Figure 16 has two principal bands-active punishments (A cells) or either intermediate or active punishments (I/A cells). In addition, two cells authorize only intermediate punishments (I), two authorize community or intermediate punishments (C/I), and one authorizes only community punishments (C). At first glance, it may appear that North Carolina's guidelines are more restrictive of the use of community punishments than are Pennsylvania's, just as North Carolina's prison guidelines are more restrictive of judicial discretion concerning prison sentences than are Pennsylvania's. However, this may be misleading. Pennsylvania's grid includes felonies and misdemeanors. North Carolina's applies only to felonies; a second grid (figure 17) sets guidelines rules for misdemeanors. It authorizes community punishments for all misdemeanors and authorizes intermediate and active punishments for some. Precisely how the two States'policies compare in relation to the restrictions they impose on judicial discretion can be determined only by analysis of data showing precisely which crimes appear in each cell of each grid and how many offenders fall into each cell. The choice between separate and combined grids for felonies and misdemeanors raises at least two significant considerations. First, it could be argued that misdemeanors are typically less serious crimes and that the offense itself should be the principal sentencing consideration. North Carolina has only three criminal history categories for misdemeanors, but six for felonies, and authorizes community penalties for all misdemeanors. (Pennsylvania's seven criminal history categories [see figure 14] might be seen as overkill.) Second, however, Pennsylvania's approach permits policymakers to look behind statutory offense classes and to distinguish among misdemeanors depending on the behaviors they involve. Thus while most Pennsylvania misdemeanors (there are three statutory classes) are included in the lowest 3 of Pennsylvania's 13 offense-gravity levels, some involving firearms, drugs, and offenses against children were placed in Levels 4 and 5. Misdemeanor manslaughters involving driving-under-the-influence (DUI) convictions were placed in Levels 7 and 8, and providing weapons to an inmate was placed in Level 9. This difference raises a broader concern that most guidelines drafters have faced: (1) whether to base policy solely on statutory offense classes, or (2) whether to devise a separate offense severity scale and, when necessary, allocate behaviors falling within the same statutory offense class to different levels. The first approach is more deferential to legislative decisions. The second approach offers several advantages: many criminal codes define crimes in broad general terms and establish only a small number of offense classes; criminal codes are seldom revised; and legislatures in most jurisdictions must approve guidelines (or at least not reject them) before they take effect. Practical application. There is no literature that shows how the different interchangeability approaches described in this section work in practice. The Pennsylvania and North Carolina guidelines are too new to have generated an evaluation literature, the D.C. Superior Court guidelines were never implemented, and the absence of policy concerning intermediate sanctions in most States has meant that the scant sentencing reform evaluation literature has little to say on the subject. Substitution of Penalties Two other related approaches for setting policies governing substitution of incarcerative and nonincarcerative punishments have been tried. The first develops a generic common currency, typically called punishment units or custody units, into which all punishments can be exchanged. This approach was discussed extensively during the development phase of the U.S. Sentencing Commission's work,[7] but only in Oregon has it been partially implemented in presumptive guidelines. Louisiana included a punishment unit approach in its voluntary guidelines, but they were repealed in 1995. The second approach sets specific exchange rates between different kinds of penalties. The details of these exchanges vary widely. Oregon's November 1993 guidelines provided that 16 hours of community service could be substituted for 1 day's confinement. New York City's community service program was designed to require 70 hours of community service as a substitute for 6 months in jail.[8] Punishment units. The idea is to create generic punishment units into which all sanctions can be converted. For example, a hypothetical system might provide for the following conversion values: 1 year's confinement: 100 units 1 year's partial confinement: 50 units 1 year's house arrest: 50 units 1 year's standard probation: 20 units 25 days' community service: 50 units 30 days' intensive supervision: 5 units 90 days' income (day fines): 100 units 30 days' electronic monitoring: 5 units The preceding is by no means a complete list; such things as drug testing, treatment conditions, and restitution might or might not be added. The values can be divided or multiplied to obtain values for other periods (for example, 75 days' confinement equals 20 units). If, for example, guidelines set "120 punishment units" as the presumptive sentence for a particular offender, a judge could impose any combination of sanctions representing 120 units. One year's confinement (100 units) plus 60 subsequent days' intensive supervision (10 units) on electronic monitoring (10 units) would be appropriate, but so would a 90-unit day fine (100 units) plus 1 year's standard probation (20 units), and so would 25 days' community service (50 units) and 6 months' intensive supervision (30 units), followed by 2 years' standard probation (40 units). Oregon's guidelines, since their initial promulgation, have incorporated sanction units (originally termed custody units) in relation to nonprison sentences; figure 18 shows the 1993 version. (See figure 2 for the 1989 version.) Cells containing only two numbers (the upper and lower limits of the authorized sentence range) govern prison and jail sentences. Cells containing three numbers are subject to the sanction unit system. The top number in each sanction-unit cell represents the applicable sanction unit maximum and the bottom number represents the maximum county jail term (in days) that can be imposed. The maximum number of sanction units increases from 90 to 120 to 180 as crime seriousness or criminal history increase. Sanction units not used as part of a jail term remain available for use to punish violations of probation conditions. Oregon's sanctions-unit scheme bears no relation to the hypothetical scheme previously described. No values are attached to intensive supervision, fixed fines and day fines, restitution, outpatient drug or sex offender treatment, or electronic monitoring. All of the sanctions affected except community service, are forms of custody, and for custody a day equals a day equals a day. No sanctions-unit scheme is required to express that equivalence. The policy that 16 hours of community service equals 1 day of confinement could be expressed in a simple one sentence statement. Oregon's sanctions-unit scheme does provide a system for limiting the scope of back-up penalties attached to condition violations, but this too could be defined simply by stating the maximum number of days such penalties could involve. Two reasons probably explain why Oregon's sanctions-units scheme is so modest. Both are among the reasons why the Pennsylvania commission, which briefly considered a punishment-units system in 1993, did not adopt one. First, as is discussed at greater length below, many believe that confinement is the basic form of punishment and that to be comparable any other sanction must be equally burdensome. Thus, many people would be uneasy that the hypothetical scheme treats 1 year's imprisonment, 50 days' community service, or a fine equal to 90 days'income as equivalent. Second, if conditions like house arrest, drug testing, electronic monitoring, restitution, community service, and fines are given unit values, the resulting combinations of numbers may seem arbitrary. For example, many people would consider a sentence of 2 years' incarceration more severe than one comprising all the elements listed previously (house arrest, drug testing, and so on), even if each was designated as 200 sanction units. After trying to work out the details of such a scheme, Pennsylvania decided it was unworkable. Exchange rates. Another approach to substitution of penalties is simply to specify equivalent custodial and noncustodial penalties and authorize judges to impose them as alternatives. Washington's commission did this and later proposed a more extensive system,[9 ]which the legislature did not adopt. Partial confinement and community service were initially authorized as substitutes for presumptive prison terms on the basis of 1 day's partial confinement or 3 days' community service for 1 day of confinement. The partial confinement/confinement exchange is probably workable (for short sentences; house arrest, assuming it is considered partial confinement, is seldom imposed for more than a few months), but the community-service exchange rate is not. Like the punishment-unit proposals, the equivalency approaches to date have been unable to overcome the psychological and political pressures to make "equivalent" punishments as objectively burdensome as prison, which limits their use to the most minor offenses and offenders. Under Washington's initial policy (3 days' community service equals 1 day's confinement), community service could be used in place of 3 to 10 days' confinement, if existing successful programs were used as models. The difficulty is that, in order to be credible, community service programs must be enforced, and experience in the United States and elsewhere instructs that they must be short. For these reasons, the best-known U.S. program in New York set 70 hours as a standard.[10] The national policies in England and Wales, Scotland, and the Netherlands set 240 hours as the upper limit; these programs were designed for offenders who otherwise would receive prison sentences of up to 6 months.[11] It is easy to criticize the Oregon commission for not carrying its innovation further and the Washington commission for lack of imagination, but that would be unfair. Working out exchange rates is very difficult, if not impossible, if policymakers begin from the premise that current and past crimes are the only valid determinants of sentences and that alternate sentences must be as burdensome as the prison sentences they displace. If officials try to make prison and community sanctions equivalent, the range for substitution between prison and community penalties is likely to be very narrow. A system like New York's community service program (70 hours' work in place of 6 months' jail) can be justified-the idea was to give repetitive property offenders some meaningful enforced penalty rather than to impose an expensive jail term that would likely have no deterrent effects-but it requires acceptance of the view that interchageable punishments need not be equally burdensome, a view which no sentencing commission has yet been prepared to accept. Categorical Exceptions Categorical exception policies, focusing not on the sanction but on the offender, are permissive. They authorize but do not direct judges to disregard otherwise applicable sentencing ranges if offenders meet specified criteria. One example is a Federal guidelines provision (Rule 5.K.1) that empowers judges to depart from guidelines if the prosecution files a motion proposing such a departure on the rationale that the defendant has provided "substantial assistance [to the government] in the investigation or prosecution of another person." Once the motion is made, the judge is free from guidelines' presumptions about appropriate sentences. Empirically speaking, this is an enormously significant escape hatch from the Federal guidelines, because it principally benefits offenders convicted of serious multiparty offenses and because it affects large numbers of cases. In 1994, 19.5 percent of all Federal sentences were downward substantial-assistance departures. So were nearly one-third (31.7 percent) of sentences imposed on Federal drug offenders.[12] The Federal categorical exception concerning substantial assistance, however, has no special relevance to intermediate sanctions. Only Washington State has developed extensive categorical exception policies. Under the First- Time Offender Waiver, judges may disregard otherwise applicable guidelines when sentencing qualifying offenders, and "the court is given broad discretion in setting the sentence."[13] Available alternatives include up to 90 days' jail or 2 years' probation and financial penalties, compulsory treatment, and community service. To be eligible, the offense must be a first conviction for a nonviolent, nonsexual offense (some drug offenders are also ineligible). In 1993, 2,139 (of 7,224 eligible) offenders (27 percent) were sentenced under the first-offender exception.[14] Washington's Special Sex Offender Sentencing Alternative authorizes judges to suspend prison sentences for most first-time sex offenders.[15] To qualify, the offender must agree to two examinations by certified sex-offender treatment specialists and to preparation of a treatment plan. Offenders whose otherwise applicable presumptive sentence does not exceed 8 years are eligible. Following a determination that the offender is amenable to treatment, the judge may suspend the presumptive sentence and impose a community sentence that includes sex-offender treatment, up to 90 days in jail, community supervision, various financial obligations, and community service. In 1993, of 940 eligible offenders, 400 (43 percent) received special sex-offender departures.[16] No other State has attained as much experience with use of categorical exceptions to sentencing guidelines as Washington. Washington also has a "work ethic [boot] camp" program that permits substitution of 4 to 6 months' boot camp for 22 to 36 months in prison. The idea, however, has potentially broad application to guidelines systems. Delaware's Voluntary Continuum of Sanctions Delaware is a special case. In some ways its approach does not fit into this discussion. The emphasis in this report is on presumptive guidelines that attempt to structure sentencing discretion. Delaware's guidelines are voluntary, and judges are free to ignore or follow them. The guidelines lack legal authority, and no one may appeal if a judge ignores them. However, Delaware was the first State to attempt explicitly to incorporate nonprison sanctions into its sentencing policies, and in the mid- to late- 1980's received substantial professional and media attention for this reason. Moreover, though there have been no independent evaluations of Delaware's voluntary guidelines, the chairman of Delaware's Sentencing Accountability Commission (SENTAC) has published articles that present data suggesting that the guidelines have produced greater consistency and predictability in Delaware sentencing.[17] Delaware's guidelines are not embodied in a grid. Instead, Delaware Supreme Court rules provide standards for sentences for typical instances of specific offenses. Sentences are increased or decreased to take into account aggravating or mitigating circumstances that SENTAC has identified. Judges are required to state, on the record,reasons for sentences that deviate from the standards. The adequacy or persuasiveness of these reasons, however, cannot be appealed to higher courts. SENTAC drafted the sentencing standards and also devised a five-level continuum of punishments that judges incorporate in their sentences. In decreasing order of severity, they are: Level V (imprisonment), Level IV (house arrest or residential treatment programs), Level III (intensive supervision), Level II (standard probation), and Level I (unsupervised probation). Judges can use the sanction levels in three ways. First, sentences are sometimes expressed in terms of X months at Level V, followed by Y months at Level III, and Z months at Level II. Second, judges use the levels as a means to provide measured responses to condition violations. Judges need not choose between ignoring a violation or sending the offender to jail or prison. For example, a Level II offender who violates conditions can be sanctioned by a control upgrade to Level III or Level IV. Third, an offender who is doing well can be rewarded by a downgrade. For example, a Level III offender who is performing conscientiously may have his or her control status reduced to Level II. Because little has been written about Delaware's guidelines, not much more can be said here about them. The crucial and, in the absence of an evaluation, as yet unanswerable question is how the guidelines are used by Delaware judges and whether they achieve better or worse consistency in sentencing than in States using presumptive guidelines. Interchangeability Among Nonincarcerative Punishments No jurisdiction has devoted significant attention to alternate ways of structuring or guiding judicial discretion over choices among different nonincarcerative punishments. The North Carolina and Pennsylvania zones of discretion, respectively, distinguish among "community" or "restorative" sanctions-like standard probation, community service, and fines-and more restrictive sanctions like house arrest and intensive supervision. Both States' guidelines contain a few cells in which only community or restorative sanctions are authorized. Within any zone of discretion, however, judges receive little guidance for their decisions among authorized nonincarcerative sanctions. Pennsylvania commentary urges judges to take rehabilitative considerations into account when fashioning nonprison sentences, and North Carolina commentary suggests, and implicitly recommends, "normal" durations for various nonprison sanctions. Each of the methods for integrating intermediate sanctions into sentencing guidelines discussed previously could be adapted to govern such choices. As figures 8 and 9 illustrated, for example, many more zones of discretion could be established that would relate particular kinds of nonincarcerative sanctions to differences in offense severity. Figure 19 shows a 10-category punishment classification, which Delaware considered and rejected in the early 1980's, that could have been used in that way. Delaware's current five punishment levels provide a simpler approach. Or, combining the exchange rate and categorical exception approaches, exchange rates could be developed for many other kinds of sanctions, and policy statements could specify the kinds of offenses or offenders to which particular sanctions apply. For example, rules might provide that property offenders should ordinarily receive financial penalties or community service, that drug-dependent offenders should ordinarily receive intensive supervision coupled with drug treatment conditions, and that all moderately serious violent offenders should ordinarily receive partial or intermittent confinement with restitution or treatment conditions as appropriate. To date, however, no jurisdiction has made any such effort. Except for the few cells in the North Carolina and Pennsylvania grids that preclude both restrictive intermediate punishments and confinement, and limits placed on duration of community confinement sentences (for example, house arrest, partial confinement, or day-reporting centers), once systems authorize judges to impose a nonconfinement sentence, the judges have wide, unguided discretion to choose among those that are available. Authority The question here concerns the nature and weight of the legal presumptions that govern choices between incarcerative and nonincarcerative punishments, and among nonin-carcerative punishments. Judges typically have wide, unregulated discretion concerning both choices. However, in Pennsylvania and North Carolina, both of which have a zones-of-discretion approach, a sentence to a generic sanction that is more severe than is authorized is considered a departure that requires reasons. Thus, for the four cells in Pennsylvania's guidelines grid and for the one cell in North Carolina's felony guidelines grid that specify only "restorative" or "community" punishments, intermediate or incarcerative sentences are presumptively unauthorized. There are, however, few such cells in either system. The large numbers of cells permitting less intrusive community penalties also permit more intrusive ones. Distinctions between voluntary and presumptive guidelines, and among the latter, distinctions between those that are restrictive and those that are flexible, are important in relation to imprisonment sanctions. They are nearly irrelevant in relation to nonimprisonment sanctions. Within the usually broad range of sanctions permitted in any cell, judges in every system have complete discretion to choose among them. This is true concerning choices between prison and nonprison penalties, and among nonprison penalties. For example, for cases falling into the intermediate punishments zone of the grid in North Carolina, judges may impose any combination of the authorized punishments, for any duration up to 5 years, and in addition, may impose any combination of the lesser punishments included within the "community punishments" category. No reasons need be given for such choices, and no appeal is available. The scope of legal authority allowed sentencers in zone-of-discretion and penalty-units systems is potentially different. In Oregon, for example, the penalty-units system for non-State-prison sentences limits the defendant's maximum vulnerability to punishment, even in relation to back-up sanctions for breaches of technical conditions. Because the Pennsylvania and North Carolina systems do not limit judges' choices among nonincarcerative sentences, there are few limits on offenders'vulnerability. In cell H-1 of North Carolina's grid, for example (see figure 16), a judge could impose 12 months of unsupervised probation for one offender, and a 5-year term of probation including 6 months in jail (as part of a split-sentence), residential drug treatment, intensive supervision with electronic monitoring, a fine, restitution, and community service for another. A different way to make this point is to observe that reduction in disparities in prison sentences is a major goal of many guidelines systems, but that few efforts are typically made to reduce or avoid disparities in nonprison sentences. There are various ways that policymakers could try to reduce disparities among nonprison sentences. To date, few attempts have been made to do so. Endnotes 1. Andrew von Hirsch, Martin Wasik, and Judith Greene, 1989, "Punishments in the Community and the Principles of Desert," Rutgers Law Review 20:595-618. 2. Richard Frase, 1991, "Defendant Amenability to Treatment or Probation as a Basis for Departure Under the Minnesota and Federal Guidelines," Federal Sentencing Reporter 3:328-33. Richard Frase, 1991. "Implementing Commission-Based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota," Cornell Journal of Law and Public Policy 2:279-337. 3. Ilene H. Nagel and Stephen J. Schulhofer, 1992, "A Tale of Three Cities: An Empirical Study of Charging and Bargaining Practices Under the Federal Sentencing Guidelines," Southern California Law Review 66:501-66. 4. Craig Mosbaek, Fourth Year Report on the Implementation of Sentencing Guidelines-1993 Felony Sentences (Salem, OR: Oregon Criminal Justice Council, 1994). 5. U.S. Sentencing Commission, Annual Report-1994 (Washington, DC: U.S. Sentencing Commission, 1995), table 30. 6. Ibid., figure E. 7. Norval Morris and Michael Tonry, Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System, (New York: Oxford University Press, 1990), 73-75. 8. Douglas McDonald, Punishment without Walls: Community Service Sentences in New York City (New Brunswick, NJ: Rutgers University Press, 1986). 9. David Boerner, Sentencing in Washington: A Legal Analysis of the Sentencing Reform Act of 1981 (Seattle: Butterworths, 1985). Washington State Sentencing Guidelines Commission, A Decade of Sentencing Reform: Washington and Its Guidelines, 1981-1991 (Olympia, Washington State Sentencing Guidelines Commission, 1992). 10. Douglas McDonald, Punishment Without Walls: Community Service Sentences in New York City (New Brunswick, NJ: Rutgers University Press, 1986). 11. Ken Pease, "Community Service Orders," in Crime and Justice: A Review of Research, vol. 6, ed. Michael Tonry and Norval Morris (Chicago: University of Chicago Press, 1985). Gill McIvor, Sentenced to Serve: The Operation and Impact of Community Service by Offenders (Aldershot, England: Avebury, 1992). Gil McIvor, 1993, "CSOs Succeed in Scotland," Overcrowded Times 4(3):1, 6-8. Peter J. Tak, 1995, "Netherlands Successfully Implements Community Service Orders," Overcrowded Times 6(2):4, 16-17. 12. U.S. Sentencing Commission, Annual Report-1994. (Washington, DC: U.S. Sentencing Commission, 1995), Tables 33 and 53. 13. Washington State Sentencing Guidelines Commission, A Statistical Summary of Adult Felony Sentencing, Fiscal Year 1993 (Olympia: Washington State Sentencing Guidelines Commission, 1994), I-18. 14. Ibid., I-18-19. 15. Washington State Sentencing Guidelines Commission, Implementation Manual-FY 1994 (Olympia: Washington State Sentencing Guidelines Commission, 1994), I-19-21. 16. Washington State Sentencing Guidelines Commission, A Statistical Summary of Adult Felony Sentencing, Fiscal Year 1993 (Olympia: Washington State Sentencing Guidelines Commission, 1994), I-18. 17. Richard S. Gebelein, 1996, "SENTAC Changing Delaware Sentencing," Overcrowded Times 7(4):1, 9-11. ------------------------------ Chapter 5 Problems and Prospects The task of incorporating intermediate sanctions into sentencing guidelines is, in the late 1990's, at about the same stage sentencing guidelines were in the early 1980's. The need to devise means to structure judicial discretion was widely recognized, and a few States (notably Minnesota, Pennsylvania, and Washington) had adopted policies aimed at doing so. Today, the need to incorporate intermediate sanctions into sentencing guidelines is widely recognized, and a few States (notably North Carolina and Pennsylvania) have attempted to do so. Now, as then, the pioneering States have laid important foundations on which they and others can build. Problems and alternative policy choices have been illuminated, but much remains to be done. Intermediate sanctions, if used for the offenders for whom they are designed, can play an important part in a just and cost-effective system of sentencing. Chapter two shows that well-run and targeted intermediate sanctions can provide credible mid-level punishments at less cost than imprisonment with no worse recidivism rates for new crimes. Chapter three shows that sentencing guidelines have reduced sentencing disparities, improved consistency, and enabled jurisdictions that wish to do so to tie their sentencing policies to existing and planned corrections resources. Chapter four discusses the methods used to date to build intermediate sanctions into sentencing guidelines. Some of those methods are promising and warrant further development. Others appear to be at dead ends. Still other possible ways to structure judicial discretion concerning intermediate sanctions deserve consideration. This chapter explains these assertions and suggests possible next steps that might be considered by jurisdictions wishing to incorporate intermediate sanctions into sentencing guidelines. Such a system would presumably strive to (1) achieve consistency in sentencing; (2) avoid racial, gender, and other unwarranted disparities; and (3) generate flows and types of offenders who could be accommodated in existing and planned corrections programs, both institutional and community-based. The first two of these goals are shared by every existing guidelines system, though the degree to which they have been realized varies. Some systems, like those in Minnesota, Washington, and Oregon, have achieved sig-nificant success in relation to confinement but little or no success in relation to intermediate sanctions. The third goal, linking policies to resources, is sought in relation to prison beds by those States that have adopted "resource constraint" policies (see "Tying Policy to Resources" under "Effects on Sentencing Practices" in chapter 3). Some States, including Minnesota, Washington, Oregon, and Kansas, have been markedly successful for extended periods. Movement toward realization of these three goals as they relate to intermediate sanctions is the subject of this report. If improved consistency and reduced disparities are to be achieved, and if policy is to be tied to resources, sentencing must be made proportional and predictable. Based on those criteria, progress toward incorporation of intermediate sanctions into guidelines has been slight. Even in North Carolina and Pennsylvania, the States in which the greatest advances have been made, no rules govern choices among intermediate sanctions or, in the portions of their guidelines grids in which both confinement and intermediate sanctions are options, between them. In the long term, mechanisms must be developed that will set policies governing these choices. Much of the following discussion of how that can be done is speculative and exploratory, as it extrapolates from, rather than documents, relevant experience. Building on the Past This chapter describes the four approaches taken thus far to incorporate intermediate sanctions in sentencing guidelines-zones of discretion, punishment units, exchange rates, and categorical exceptions. Zones of Discretion Zones of discretion, adopted in North Carolina and Pennsylvania, offer the broadest promise. By defining various offense-offender combinations (guidelines cells) for which only confinement, an intermediate sanction, or a community penalty is presumptively appropriate, zones of discretion make some level of proportionality and predictability in the use of various sanctions more likely. However, because these zones set no presumptions governing choices among intermediate sanctions (where that is the only presumptively applicable penalty) or between intermediate sanctions and confinement (where both are authorized), such sentencing decisions must be made in a policy vacuum. That vacuum can be filled, however. North Carolina and Pennsylvania have taken small steps to provide guidance concerning choices among intermediate sanctions. North Carolina's Training and Reference Manual, although it neither creates dispositional presumptions nor makes recommendations, provides valuable information on typical durations of intermediate sanctions. For example, "The current average length of electronic monitoring is 90 days or less," and "the current average length of intensive probation is from 6 to 9 months." The rationales presumably are that such information will help judges decide what duration to impose and that judges will be inclined to follow such conventions.[1] Pennsylvania, likewise, provides information to judges that may be intended to influence their decisions. Pennsylvania's Sentencing Guidelines Implementation Manual reminds judges that, in selecting among confinement and restrictive intermediate sanctions in Level 3 (prison or a restrictive intermediate sanction), they "may choose to place the primary focus of the sentence on treatment of the offender by placing the offender in an inpatient treatment facility." In slightly less neutral language, the Manual suggests that judges "should consider" sentences to boot camp or drug or alcohol treatment for qualifying offenders in Level 2.[2] Punishment Units In the late 1990's, the punishment-units approach does not appear to have broad relevance. The contemporary belief that confinement is the standard punishment and that any "equivalent" punishment must be comparably intrusive and burdensome greatly limits the general potential of the punishment-units approach. Oregon is the pioneering punishment-units jurisdiction and, as chapter 4 demonstrates, progress has been slight. In this system, 1 day in total confinement, inpatient treatment, partial confinement, or house arrest equals 1 punishment unit, as does 16 hours of community service. The punishment-unit idea is not necessary to express an equivalence between 1 day's total confinement and 1 day of confinement elsewhere. No policies were set for valuing fines, house arrest, restitution, or intensive supervision. In practice, the 2-to-1 conversion from community service to confinement limits use of community service to only the most trifling crimes. Fifteen days of confinement has the same value as 240 hours of community service. By contrast, in England, Scotland, and the Netherlands, 240 hours of community service is designed as a substitute for prison sentences of up to 6 months. Oregon uses punishment units in a second way, which may have broader relevance. Sanction units not used as part of a jail term remain available for use to punish violations of probation conditions. In effect, punishment units can operate as aggregate limits on "back-up" sanctions that are imposed for breach of conditions of the initial penalty. In many courts, judges sentence offenders who have breached conditions of a community penalty to jail or State prison as a penalty. When the breach is of a technical condition, such as prohibitions on alcohol or drug use or violation of curfews, imprisonment will appear disproportionately severe to many observers and, from a cost-benefit perspective, disproportionately costly. Use of punishment units to limit the scope of back-up sanctions is a way to control both excesses. However, that notion has no inherent link with punishment units. Any guidelines system could set presumptions for back-up sanctions that are proportional to the seriousness of the original crime. Exchange Rates Exchange rates, in practice, are a simpler version of punishment units and, in the late 1990's, appear no more likely than punishment units to be broadly useful. Rather than establishing some generic currency into which all sanctions can be converted and then exchanged, exchange rates directly identify equivalent punishments. In Washington's initial guidelines, for example, 1 day's confinement was made exchangeable for 1 day's partial confinement or 3 days' community service. As in Oregon's punishment-units scheme, the idea of exchange rates is unnecessary to equate one length of confinement to another, and equating 3 days' community service to 1 day's confinement greatly limits the scope of community service. Also, as in Oregon's scheme, Washington's exchange rates do not take account of such common penalties as fines, restitution, or intensive supervision. However, unlike Oregon's punishment-units scheme, which potentially could play an important role in relation to back-up sanctions, no important residual functions for exchange rates have yet been identified. Exchange rates are limited in their potential uses for the same reasons punishment units are: Prevailing views require that imprisonment be considered the "normal" punishment and that alternatives to imprisonment be comparably burdensome and intrusive. As a result, exchange rates are unlikely to play a significant role in sentencing guidelines. Categorical Exceptions Categorical exceptions, both permissive and presumptive, have a role to play in incorporating intermediate sanctions into sentencing guidelines. Permissive exceptions, like Washington's Special Sexual Offender Sentencing Alternative, authorize but do not direct the judge to set aside the normally presumptive range of sentences for a specific category of offenders. In effect, they operate as trumps that the judge may decide whether and when to use. As chapter 4 shows, Washington judges often assert their authority over permissive exceptions in order to craft individualized sentences for sexual offenders and first offenders. Presumptive exceptions, as the words imply, indicate to the judge that defined categories should ordinarily be handled in a particular way. The Federal Sentencing Reform Act of 1984, for example, in Section 994(j), provided that the Federal guidelines shall "reflect the general appropriateness of imposing a sentence other than imprisonment in cases in which the defendant is a first offender who has not been convicted of a crime of violence or an otherwise serious offense." The U.S. Sentencing Commission could have reiterated that precise language in its guidelines, but it did not. Had it done so, Federal judges would have operated under a presumption that some sentence other than imprisonment was appropriate for most first offenders. Both permissive and presumptive exceptions can potentially be useful in incorporating intermediate sanctions into sentencing guidelines. Looking to the Future To this point, this report and this chapter have tried to describe documented experience and to comment on it. This final section looks to the future and offers more speculative suggestions that policymakers might consider as they continue their efforts to build intermediate sanctions into sentencing guidelines. Zones of discretion and categorical exceptions have roles to play in the effort to combine sentencing guidelines and intermediate sanctions into a single comprehensive system of structured sentencing discretion. Use of zones of discretion has permitted policymakers to specify categories of offenses and offenders for which only particular kinds of sanctions are presumptively appropriate (for example, only imprisonment or only intermediate sanctions or only less intrusive community penalties). Little guidance has as yet been provided to judges as they choose between imprisonment and other sanctions or among intermediate sanctions. Categorical exceptions are the most promising tools available for providing that guidance. Before categorical exceptions are discussed, however, introduction of two simple jurisprudential concepts may be helpful. These are the distinction between purposes of sentencing and the concept of parsimony. Jurisprudential Principles That Can Frame the Design of Guidelines for Intermediate Sanctions Purposes at sentencing. In recent years, there has been widespread belief that abstract sentencing purposes have either near-absolute or virtually no relevance to sentencing policy; this is a mistake. Proponents of "just deserts" theories have urged that ideals of proportionality be the primary criteria for setting sentencing policy.3 Because this would leave little role for rehabilitative, deterrent, incapacitative, and other purposes that many policymakers and practitioners believe are relevant, no jurisdiction has adopted such a single-purpose scheme. Although indeterminate sentencing was nowhere a single-purpose system, rehabilitative considerations were especially influential. There are, however, few contemporary proponents of primarily rehabilitative systems. Most modern sentencing systems purport to be multipurpose, but it has proven difficult to give operational meaning to that idea. Although there have long been vigorous debates over the merits of retribution, rehabilitation, prevention, general and specific deterrence, and incapacitation as penal goals, consensus is seldom reached that one is more important than the others. This lack of agreement occurs in part because the various purposes are relevant to different cases in different ways. For example, among three offenders in a convenience store robbery (without firearms), one may have been involved in 10 prior robberies; for this individual, incapacitation may seem the most important sentencing purpose and confinement the mechanism. The second may be a drug-dependent first offender; in this case, rehabilitation may be the most important purpose and outpatient drug treatment the mechanism. The third may be a non- drug-dependent first offender, employed and with a family; for this individual, retribution and deterrence may be the primary purposes, and a combination of a substantial fine and house arrest during nonworking hours the mechanisms. The three sentences described in the preceding paragraph are difficult to reconcile with any single punishment purpose, which is why policymakers frequently adopt all purposes. The difficulty with this is that informing a judge that all purposes are relevant provides no guidance whatsoever in sentencing particular cases. If the choice is between a single punishment purpose or multiple purposes, lack of guidance may appear unavoidable. These problems disappear when purposes of sentencing are distinguished from purposes at sentencing. Traditional debates concern purposes of sentencing, that is, the overall purposes of the sentencing process or system. Although vigorous arguments that retribution or rehabilitation are the only or primary purposes of sentencing can and have been made, policymakers have typically been unwilling to reject in principle the legitimacy of other traditional purposes. Purposes at sentencing are those that are relevant to the disposition of individual cases. They vary with the circumstances of the offense and the offender, as the example of the three convenience store robbers shows. The idea of purposes at sentencing is especially relevant to nonincarcerative sanctions. When guidelines dealt only with who went to jail or prison and for how long, the lack of guidance resulting from multiple punishment purposes created few problems. However, when guidelines encompass intermediate sanctions, the idea of purposes at sentencing-when combined with the idea of categorical exceptions-supplies a tool for providing guidance to judges when choosing among different sanctions. Guidelines could easily require that judges choosing between confinement and intermediate sanctions, or among intermediate sanctions, be guided by a series of presumptions about purposes relevant to an individual case at sentencing. Any sentence inconsistent with the presumption would be considered a departure and would require a provision of reasons that could be reviewed on appeal. The principle of parsimony. The principle of parsimony, or the concept of the least restrictive appropriate alternative, is equally relevant to intermediate sanctions. In terms both of humane treatment of offenders and economy in public expenditure, law reform bodies (including the American Law Institute [in the Model Penal Code], the National Commission on Reform of Federal Criminal Laws, and the American Bar Association [in all three editions of its standards for sentencing]) have consistently urged adoption of least restrictive alternative policies. In various forms, the least-restrictive-alternative concept has a long history. The utilitarian philosopher Jeremy Bentham, for example, asserting that all avoidable human suffering is undesirable, urged the adoption of a "principle of parsimony" by which punishment could be justified only to the extent that the offender's suffering was more than offset by diminution of suffering by others.[4] Contemporary writer Norval Morris proposed an influential theory of "limiting retributivism" in which retribution (or "just deserts") sets upper limits on deserved punishments and lower limits for especially serious crimes; within those limits, parsimony calls for the least restrictive alternative, unless articulable rationales justify harsher treatment for particular offenders.[5] Placed in the context of intermediate sanctions and sentencing guidelines, concern for parsimony would yield a least-restrictive-alternative presumption that intermediate sanctions are to be preferred to confinement, and that among intermediate sanctions the least restrictive and intrusive among those authorized are to be preferred. Suggestions for Incorporating Intermediate Sanctions in Sentencing Guidelines Efforts to incorporate intermediate sanctions into sentencing guidelines are at an early stage. Although current efforts are modest and limited, foundations are available on which comprehensive sentencing policies can be built, supported by past accomplishments of sentencing guidelines systems and ultimately providing guidance for judges in all their sentencing decisions. The suggestions offered below are speculative. Whether they appear promising or warrant serious consideration are judgments for sentencing policymakers. Nonetheless, a system of sentencing guidelines that contains the following provisions would seem to offer substantial promise of bringing greater fairness, consistency, and predictability to the use of intermediate sanctions. The guidelines grid should contain four to six zones of discretion. The polar zones are one in which the crimes are so serious that any punishment less harsh than imprisonment would unduly depreciate the seriousness of the crime, and a second in which the crimes are so venial that any punishment harsher than standard probation, a minor fine, or restitution would be unjust. At least two other zones should be created: one authorizing restrictive sanctions like inpatient drug or other treatment and partial confinement, and another authorizing less restrictive sanctions like day fines, intensive supervision, house arrest, and community service. At its upper and lower margins, each zone would overlap with the next, thereby giving judges authority, without "departing," to choose among sanction types. The preceding proposal is but a sketch. A sentencing commission staff document explaining all the possible choices in such a grid, and the considerations for and against each, would be quite lengthy. Four general observations might, however, be made about this first suggestion. First, it potentially applies to all guidelines systems, even those, like Florida's and Delaware's, that do not use a grid. The absence of grids in such jurisdictions is entirely cosmetic. To help overcome negative judicial stereotypes about guidelines and "sentencing by mathematics," Delaware's Sentencing Accountability Commission promulgated its guidelines in narrative form: the normal sentence for offense X should be Y. This is much less efficient than a grid because it requires many pages of text, but with a few days' work an analyst could start from the statements and collapse their content into a grid. Likewise, the contents of North Carolina's grid, including its intermediate sanctions elements, could be expressed in a lengthy narrative manual. Grids, though an efficient way to organize and display a vast amount of information, are not essential. Second, in order to maintain norms of proportionality, guideline cells in each zone could specify maximum durations or amounts for sanctions authorized in each cell, and these could vary with offense seriousness or extent of criminal history. The cells could also specify maximum aggregate penalties, including back-up sanctions. Third, grids containing more than four zones could be particularly useful in setting back-up sanctions when offenders breach conditions of their sentences. Often, judges confronted by an offender breaching conditions of a nonincarcerative penalty believe their only choices are, in effect, to ignore the breach or to send the offender to jail or prison. Under a system with six zones of discretion, however, depending on the seriousness of the breach, a judge could punish condition breaches by a Zone 2 offender by imposing sentences authorized by Zones 3 through 6. Policy statements could provide guidance to judges on the details of revocation and resentencing to a higher zone's sanctions. Delaware's SENTAC guidelines operate like this by use of Delaware's five sanction levels (see chapter 4). (Another method to bring greater consistency and predictability to use of back-up sanctions, using categorical exceptions, is discussed later.) A second option might be to craft another set of guidelines to deal with revocations both at the time of probation and at parole. Fourth, although the preceding discussion mentions confinement only in reference to the top zone, in practice, both North Carolina and Pennsylvania authorize confinement as an alternative to other sanctions in a large majority of cells. In some ways, this undermines both the abstract notion of proportionality in a continuum of sanctions and the mechanism of zones of discretion. In many jurisdictions, however, the availability of confinement as an authorized penalty for most crimes may be politically necessary. This could be achieved by permitting judges to depart from guidelines in which confinement is not presumptively applicable, if they give reasons for why they are doing so. Even if departure authority is not enough, because policymakers want the availability of confinement to be obvious from the guidelines grid, many concerns about proportionality and predictability can be addressed by means of categorical exceptions and presumptions. For example, cells could authorize both confinement and nonconfinement sanctions, but subject to a least-restrictive-appropriate- punishment presumption requiring that judges provide reasons for imposing confinement; these could be made appealable, depending on how strong policymakers want the presumption to be. Guidelines should include dispositive presumptions. A significant limitation of the zones-of-discretion approach adopted by North Carolina and Pennsylvania is that judges are given little guidance in choosing among sanctions (including confinement) authorized in various zones. Many cells in Pennsylvania's Level 2, for example, allow judges to select among restorative (least severe), intermediate, and short confinement options. In Level 3, judges choose among any intermediate sanction and prison or jail terms. Some policy guidance could be given by means of presumptions. One possibility, mentioned previously, would be to adopt a least-restrictive- alternative presumption and to establish policies that order sanctions in terms of their level of restrictiveness. One possible ordering, from least to most restrictive, might be unsupervised probation, probation, small fines, community service, large fines, intensive supervision, house arrest, partial or intermittent confinement (day- reporting centers, halfway houses, night or weekend jail confinement), and total confinement. Judges would thereby be directed to impose the least restrictive sanction authorized in the applicable cell, or to explain why another sanction was chosen and to explain why each less restrictive option was deemed inappropriate. A second, related possibility would be to adopt a series of offender- or offense-specific dispositive presumptions. All relate to choices among sanctions authorized in the applicable cell. The following are illustrative possibilities: 1. Nonviolent property offenders who are not drug-dependent should ordinarily be sentenced to standard probation, community service, or fines (separately or in combination with other sanctions), if these sentences are authorized in the applicable cell. 2. Drug-dependent property, drug, and minor violent (for example, robberies not involving firearms or injuries) offenders should ordinarily be required to participate in drug treatment (outpatient or residential, depending on their drug-use history) and, to the extent feasible, should also be sentenced as if they were not drug- dependent. 3. Persons convicted of crimes involving gratuitous infliction of violence (that is, beyond that otherwise inherent in their crimes) should ordinarily be sentenced to confinement. 4. Offenders who are primary care or income providers to their families should ordinarily be sentenced to a community penalty that will permit them to continue these roles; any confinement required should be partial or intermittent. A sentencing commission might adopt a dozen such dispositive presumptions. Their exact provisions would vary with the jurisdiction, and their cumulative effect would be to provide guidance to judges when choosing among authorized sanctions. The dispositive presumptions would interact with the least-restrictive-alternative presumption. For example, for a drug-dependent person convicted of robbery not involving guns or injuries, the drug- dependency presumption would override the least- restrictive-alternative presumption and might, depending on the circumstances, justify intensive supervision with outpatient or inpatient drug treatment. Three additional issues warrant mention. First, a cynic might argue that a series of presumptions like those suggested previously would be mere boilerplate that would either be ignored by judges or invoked disingenuously by rote. If that is true, dispositive presumptions would add nothing useful, but they also would do no harm. More importantly, however, such a view is far too cynical to serve as the basis for policymaking. Judges are sworn to uphold the law and, in criminal law particularly, are accustomed to working with evidentiary and probative presumptions. Most conscientious judges would take such presumptions seriously, especially if they comported with widely shared views about meaningful differences between offenders' circumstances. Even if only some judges took the presumptions seriously, the overall effect would be to make sentencing more consistent and predictable. Second, an observer might suggest that if greater consistency in the use of intermediate sanctions would be a good thing, a series of dispositive presumptions would still leave too much discretion in the hands of judges. From that perspective, guidelines systems would need to become much more detailed and set clear rules tying offenders and particular guideline cells to particular sanctions. The difficulty with this is that, as the Federal guidelines experience shows, judges deeply dislike and actively resist guidelines they believe are too rigid and detailed. Even if it were feasible to devise highly detailed guidelines for intermediate sanctions, they would likely become even more detailed than the Federal guidelines (which primarily involve confinement) and would ultimately provoke similarly negative reactions from judges and others. Third, such dispositive presumptions would authorize the imposition of different kinds of punishments on like-situated offenders, which would violate "just deserts" concerns that sentencing should be tied only or primarily to the severity of the crime. To people who are persuaded by the distinction between purposes of and purposes at sentencing, such an observation will be unimportant; the distinction is premised on the assumption that many or most judges believe that both the offense and the offender's personal circumstances are relevant sentencing considerations. Guidelines should authorize judges to declare and be guided by the relevant purposes at sentencing in every case. This concept provides a rationale for use of dispositive presumptions. Whether there are 3, 6, or 10 zones of discretion, for all but the most and least serious offenses judges will often be able to choose among generically different penalties. Whether a particular penalty is appropriate often depends on the offender's characteristics. For crimes of comparable severity that fall in the same offense-severity level of a guidelines grid but are different in their character, noncustodial penalties may be variously appropriate. For example: 1. For a drug-dependent shoplifter or burglar or drug dealer, prevention of future crimes and rehabilitation may be the most important purposes at sentencing; compulsory drug treatment (residential or outpatient backed up by intensive supervision, depending on the offender's drug problem and prior treatment experience) might be the optimal primary sentence with restitution or community service as an adjunct. 2. For a bank teller who has embezzled funds, retribution and general deterrence may be predominant purposes at sentencing, and restitution and community service or a fine the optimal sentence. 3. For the perpetrator of a commercial fraud, retribution and general deterrence may be the predominant purposes, and restitution, stigmatizing community service, and a very substantial fine the optimal sentence. 4. For an employed blue-collar head of family who has committed a serious assault while intoxicated, retribution and deterrence may be the predominant purposes, and a substantial fine and nighttime and weekend confinement (thereby permitting him to continue to work and support his family) the optimal sentence. 5. For a third-time street mugger, deterrence and incapacitation may be the predominant purposes, and a period of confinement followed by intensive supervision the optimal sentence. Current guidelines systems provide no guidance to judges in discriminating among different offenders who fall into the same guidelines categories. A purposes-at-sentencing approach would provide a framework within which judges could work, and the attendant statements of governing purposes and their relation to the sentence imposed would enable observers to understand the judge's reasoning. There is a reasonable chance that greater consistency in sentences would result. Guidelines should establish policies and presumptions concerning categorical exceptions. Some types of offenders have distinctive characteristics or present distinctive challenges that may justify having every case decided on its individual merits. Policies governing such offenders are typically permissive rather than presumptive; they authorize but do not direct judges to treat defined categories of offenders as eligible for exceptional treatment. First offenders are one example. Washington's guidelines contain a first-offender exception that allows judges, for first offenders convicted of many kinds of crimes, to disregard the applicable guidelines and impose some other, usually less intrusive or burdensome, sentence. Sometimes this exception may be applied because the offense seems out of character and unlikely to be repeated and the offender is a fundamentally law-abiding person. In other cases, it might be used because the offense occurred under circumstances of unusual stress or emotionality, or for a combination of those reasons and because the defendant's family would suffer unduly were he or she incarcerated. Whatever the precise reason, first offenders often provoke compassion from judges and prosecutors; a permissive exception would allow judges and prosecutors openly to treat a case as special rather than, as often happens, do so surreptitiously. Intrafamilial sex offenders are another example. Because such offenses often involve psychopathology; because a prison sentence would break up the family, possibly leaving the victim feeling guilt-ridden for having reported the crime; and because such conditions are sometimes successfully treated, judges are often more interested in treatment and family preservation than in deterrence and retribution. Yet, guidelines often set lengthy presumptive prison sentences for sex offenses. Creating a permissive exception, as Minnesota has done by case law and Washington by statute, allows judges openly to impose what seem to them to be just and appropriate sentences. There is some overlap between permissive exceptions and both the purposes-at-sentencing notion and the creation of dispositive presumptions. The purposes- at-sentencing notion, however, deals with judges' discretion as bounded by applicable guidelines cells and zones. Permissive exceptions are broader and apply throughout the guidelines system. Permissive exceptions and dispositional presumptions differ in their literal meanings. Exceptions are permissive; they authorize but do not direct judges to treat cases exceptionally. Presumptions direct judges to treat cases in a particular way; judges who choose to do otherwise must offer convincing reasons for their departures. Together the suggestions offered here for incorporating intermediate sanctions into sentencing guidelines may appear to constitute a system of bewildering complexity, but that is not the case. Each suggestion is simple. Because the suggestions move beyond current practice and are discussed in close succession, they may appear more complicated than they are. Singly or together, they constitute modest incremental steps toward creating comprehensive sentencing systems that incorporate confinement and nonconfinement sanctions and that attempt to achieve reasonable consistency in sentencing while allowing judges to take account of meaningful differences between cases. Endnotes 1. North Carolina Sentencing and Policy Advisory Commission, Structured Sentencing for Felonies-Train-ing and Reference Manual (Raleigh, NC: North Carolina Sentencing and Policy Advisory Commission, 1994), 29-30. 2. Pennsylvania Commission on Sentencing, Sentencing Guidelines Implementation Manual, 4th ed. (Harrisburg, PA: Pennsylvania Commission on Sentencing, 1994), 6-7. 3. For example, see Andrew von Hirsch, Doing Justice: The Choice of Punishments (New York: Hill & Wang, 1976). 4. Jeremy Bentham, Introduction to Principles of Morals and Legislation (Oxford: Oxford University Press, 1988; originally published 1789). 5. Norval Morris, The Future of Imprisonment (Chicago: University of Chicago Press, 1974).