National Assessment of Structured Sentencing. MENU TITLE: Structured Sentencing Monograph Series: BJA Published: February 1996 136 pages 288,437 bytes National Assessment of Structured Sentencing February 1996 NCJ 153853 This document was prepared by the National Council on Crime and Delinquency, supported by grant number 92-SA-CX-0003 awarded by the Bureau of Justice Assistance, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed 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. Bureau of Justice Assistance 633 Indiana Avenue NW., Washington, DC 20531 U.S. Department of Justice Response Center 800-421-6770 Bureau of Justice Assistance Clearinghouse 800-688-4252 Bureau of Justice Assistance Internet Address http://www.ojp.usdoj.gov/BJA The Bureau of Justice Assistance is a component of the Office of Justice Programs, which also includes the Bureau of Justice Statistics, the National Institute of Justice, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. ------------------------------ Foreword With the incidence of violent crime continuing to rise and public demands for harsher and more certain punishment increasing, many States are examining sentencing of offenders with an eye to instituting reforms. Indeed, considerable reform of criminal justice sentencing has already taken place over the past two decades. Although these reforms have come primarily as a response to criticism of rehabilitation attempts, they have sought to accomplish widely differing goals, including reducing disparity that results from discretionary sentencing, increasing sentencing fairness, establishing truth in sentencing, and balancing sentencing policy with limited correctional resources. To meet these goals, sentencing reforms have taken a variety of forms, and many States have adopted a combination of sentencing schemes that can include determinate or indeterminate sentencing, mandatory minimum sentences for certain offenses, presumptive sentencing guidelines, and voluntary or advisory sentencing guidelines. The result is a patchwork of structured sentencing models. This monograph presents the findings of the first national assessment of sentencing reforms, which was funded by the Bureau of Justice Assistance and conducted by the National Council on Crime and Delinquency, the Pennsylvania Commission on Crime and Delinquency, and the Pennsylvania Commission on Sentencing. The publication offers lessons learned in the diverse efforts to structure sentencing over the past two decades. These lessons are offered in the context of a historical perspective of sentencing practices used in the United States, with a discussion of the issues that led to the structured sentencing movement. They are based on a national survey of existing sentencing practices in the 50 States and the District of Columbia. The Federal Government and 16 States have implemented presumptive or voluntary/advisory sentencing guidelines. Each of these States has established guidelines for different purposes, and most of them were asked to meet multiple goals, including punishment, deterrence, incapacitation, and rehabilitation. These purposes and goals influence nearly every aspect of the guideline development process. While sentencing guideline systems should be evaluated against their established goals, several factors affect their successful implementation: the political environment, legislative mandates, adequate funding, and the development of coherent sentencing policy. However, despite the lengthy and complex process of writing guidelines and the challenges of implementation, the existing sentencing commissions and their guidelines have succeeded in changing historic sentencing patterns. Structured sentencing also can deter potential offenders, incapacitate dangerous offenders, and ensure that offenders who are not dangerous receive fair and appropriate punishment. As prison populations continue to grow, interest in structured sentencing reforms also will increase. It is our hope that this assessment of structured sentencing reforms will be a valuable resource to those States considering the formation of sentencing commissions and the development of sentencing guidelines. ------------------------------ Acknowledgments The Bureau of Justice Assistance wishes to thank James Austin, Ph.D.; Charles Jones, J.D.; John Kramer, Ph.D.; and Phil Renninger of the National Council on Crime and Delinquency for their efforts in conducting this evaluation and compiling this report. In addition, BJA acknowledges the Advisory Board members: the Honorable Helen Corrothers, formerly of the U.S. Sentencing Commission; Donna Hunzeker of the National Conference of State Legislatures; Phyllis Newton of the U.S. Sentencing Commission; Judge Thomas W. Ross of the North Carolina Sentencing and Policy Advisory Commission; and Sandra Shane-DuBow of the Wisconsin Sentencing Commission. ------------------------------ Table of Contents Executive Summary Chapter 1. Introduction Structured Sentencing Definitions Focus of This Report Chapter 2. Historical Trends and Issues in Structured Sentencing Historical Developments Leading to Structured Sentencing The History and Nature of Structured Sentencing Reforms Chapter 3. An Overview of Current Sentencing Practices in the United States Mandatory Minimum Incarceration Sentences in the United States Currency of State Criminal Codes Use of Parole and Good Time Sentencing Commissions in the United States Chapter 4. Sentencing Commission Structures and Their Mandates Purposes of Reform Sentencing Commission Structure, Mandates, and Organization Political and Legal Context Summary Chapter 5. Writing Sentencing Guidelines Offense Seriousness Criminal History Other Factors Developing Sentencing Recommendations Summary Chapter 6. The Impact of Sentencing Guidelines Methodological Issues in Assessing Impact Impact on Sentencing Disparity Impact on Plea Bargaining Impact on Incarceration Rates Impact on Prison Crowding Projected Prison Populations for Guidelines States Impact on Public Safety and Crime Rates Summary Chapter 7. Summary Major Findings Policy Implications Recommendations Sources for Further Information Bibliography Appendix A: Information on States Used for Impact Analysis ------------------------------ Executive Summary Criminal justice sentencing has undergone considerable reform during the past two decades. Because of excessive sentencing disparity and criticisms of the rehabilitative ideal, many States have replaced indeterminate sentencing with structured sentencing schemes such as determinate sentencing, mandatory minimum penalties, and sentencing guidelines. Nevertheless, most States retain indeterminate sentencing structures while indicating interest in adopting such reforms. The criminal justice community has yet to develop a clear consensus on the goals of structured sentencing. The following lists the most frequently cited goals: o Increase sentencing fairness. o Reduce unwarranted disparity, either in the decision to imprison (dispositional disparity) and/or sentence length (durational disparity). o Establish truth in sentencing. o Establish a balance of sentencing policy with limited correctional resources. Structured sentencing reforms can be used to deter potential offenders and incapacitate dangerous offenders. Sentencing reforms can also be used to reduce the likelihood and length of imprisonment for the so-called non-dangerous offender. This study is a review of lessons learned over the past two decades in the diverse attempts to structure sentencing. It was funded by the Bureau of Justice Assistance (BJA) and jointly conducted by the National Council on Crime and Delinquency (NCCD), the Pennsylvania Commission on Crime and Delinquency (PCCD), and the Pennsylvania Commission on Sentencing (PCS). Structured Sentencing Definitions This examination of structured sentencing revealed at the outset that consensus is lacking regarding the meaning of sentencing guidelines, voluntary sentencing guidelines, presumptive guidelines, and advisory guidelines. Therefore, the first task in preparing this report was to establish the following definitions: o Determinate: Sentences of incarceration in which an offender is given a fixed term that may be reduced by good time or earned time. o Indeterminate: Sentences in which an administrative agency, generally a parole board, has the authority to release an offender and determine whether an offender's parole will be revoked for violations of the conditions of release. o Mandatory minimum: A minimum sentence that is specified by statute and that may be applied for all convictions of a particular crime or a crime with special circumstances (e.g., robbery with a firearm or selling drugs to a minor within 1,000 feet of a school). o Presumptive sentencing guidelines: Sentencing that meets the following conditions: (1) the appropriate sentence for an offender in a specific case is presumed to fall within a range of sentences authorized by sentencing guidelines that are adopted by a legislatively created sentencing body, usually a sentencing commission; (2) sentencing judges are expected to sentence within the range or provide written justification for departure; (3) the guidelines provide for some review, usually appellate, of the departure. Presumptive guidelines may employ determinate or indeterminate sentencing structures. o Voluntary/advisory sentencing guidelines: Recommended sentencing policies that are not required by law. Usually based on past sentencing practices, they serve as a guide to judges. The legislature has not mandated their use. Voluntary/advisory guidelines may employ determinate or indeterminate sentencing structures. Major Findings One major objective of this study was to conduct a national survey of sentencing practices to classify each State's primary sentencing systems. The results of this survey are summarized below. Current Sentencing Practices o An unprecedented number of structured sentencing reforms have taken place over the past two decades. To date, 16 States and the Federal Government have implemented, or are about to implement, presumptive or voluntary/advisory sentencing guidelines. Another five States have adopted determinate sentencing systems. o All States employ some version of mandatory minimum sentencing laws, which target habitual offenders and the crimes of possessing a deadly weapon, driving under the influence of alcohol, and possessing and distributing drugs. o Most States continue to allow inmates to earn good-time credits either to reduce a sentence or to advance a parole eligibility date. o Most States, including those that have adopted determinate and sentencing guideline models, have retained some form of discretionary release and postrelease supervision. o Most States do not employ sentencing guidelines. Furthermore, five States have tried and failed to adopt sentencing guidelines. Sentencing Guideline Models: Purposes and Goals o Virtually all guideline commissions were asked to meet the multiple goals of punishment (just deserts), deterrence, incapacitation, and rehabilitation. o Few jurisdictions explicitly state the goal of eliminating disparity with respect to race, gender, or social or economic status. o Only a few guideline commissions were required to consider the impact of guidelines on the need for future correctional resources (i.e., number of new prison beds). Structure of Sentencing Guidelines and Sentencing Commissions o The structure of sentencing guidelines varies dramatically for both assessing sentencing disposition criteria and determining sentence length. o The most common format for guidelines is the two-variable matrix, which relies on offense severity and prior criminal history. o Sentencing commission membership varies considerably in number and type of members selected. o In general, membership includes judges, prosecutors, defense attorneys, and private citizens. Corrections officials, legislators, and law enforcement officials are often included on these commissions. Implementation Issues o The process for implementing guidelines is neither quick nor inexpensive. Implementation often requires 2 years. o The annual budgets for guideline commissions and support staff range from $250,000 to $500,000. Considerable support is required to monitor and analyze compliance with guidelines and to project the impact of modifying guidelines. o Before guideline implementation, detailed data on current sentencing patterns should be collected and analyzed. A jurisdiction will also need to develop a simulation model for estimating the impact of the proposed guidelines on prison, parole, probation, and jail populations. Impact of Sentencing Guidelines o States that have implemented presumptive guidelines have reported high compliance rates and have changed historical sentencing trends. o Repeat offenders and offenders convicted of violent crimes are much more likely to be imprisoned and serve longer prison terms under sentencing guidelines. Conversely, first-time offenders charged with property crimes are less likely to be imprisoned and generally serve shorter prison terms. o Partly because of guidelines, persons convicted of drug crimes (possession and sale) are now much more likely to be imprisoned and serve lengthy prison terms. This trend has directly increased the rate of imprisonment for African-American and Hispanic-Latino/Latina offenders. o Guidelines have helped reduce sentencing disparity; however, disparity reductions have eroded somewhat over time. o There is limited evidence that guideline sentencing structures designed to account for correctional resources have had slower rates of growth in incarceration and helped to control prison crowding. However, prison crowding remains a problem for most States. o Guideline States, like nonguideline States, are likely to suffer from overcrowding over the next decade unless they embark on a substantial prison construction program or reduce prison terms for violent and drug offenders. Policy Implications Reducing Disparity While Maintaining Discretion The decisions that a jurisdiction must make about its criminal justice laws will have a major impact on the quality of justice and cost to its citizens. If laws or commissions do not make these decisions, then they are left to individual judges who must (1) determine what they perceive as just sentences and (2) determine the best use of State and local correctional resources. The major question remains: Can the development of structured sentencing, and presumptive sentencing guidelines in particular, overcome well-established organizational values that may facilitate and protect inequitable sentencing practices? One continuing concern is the ability to individualize sentencing and consider a wide range of sentencing purposes while maintaining an equitable sentencing system. Most States have attempted to control the individualization of judicial practices while maintaining that sentences should consider the full range of traditional sentencing purposes. For example, Pennsylvania requires judges to consider offenders' rehabilitative potential and community protection as well as the guidelines. In Washington, the enabling legislation mandates that sentencing guidelines incorporate the goals of retribution, incapacitation, rehabilitation, and frugal use of correctional resources. Both of these States may be the most explicit examples of the effort to structure sentencing while maintaining adequate court discretion to fashion sentences to individual offenders. Recent innovations in Oregon and Louisiana have expanded the range of sentencing options and provided for the reduction of disparity. This new approach illustrates that sentences can be very different yet equal; this approach is an important move forward in the evolution of sentencing guidelines. Future commissions may develop guidelines that will increase fairness, equity, and proportionality while still providing for a full range of sentencing options. Displacement of Discretion One of the key issues facing those attempting to control sentencing discretion is displacement of discretion from the courts to the prosecutors. The concern is that guidelines have merely shifted discretion from parole boards, prison officials, and judges to prosecutors. Little evidence exists to document how much this has occurred. Clearly, more research is needed in this area of sentencing reform. Prison Crowding To date, structured sentencing reforms have not had any appreciable effect on the problem of prison crowding. Until the legislative process and sentencing commissions can tune out pressures to get tough on crime, there is little reason to believe that structured sentencing models will solve the prison-crowding problem. Furthermore, as State prisons remain crowded, they will continue to employ discretionary early release programs. Depending on how such programs are structured, attempts to reduce disparity may be lost in the determination of how much time similarly situated offenders will serve. Recommendations (The recommendations in this section are those of the Advisory Board.) The purpose of this project was not to advocate one form of sentencing policy, whether it be presumptive guidelines, voluntary/advisory guidelines, determinate sentencing, or indeterminate sentencing. Disparity, incarceration rates, and prison crowding can be reduced by several sentencing reform measures. The question is how best to achieve these goals. The most promising model is sentencing guidelines developed by sentencing commissions. Such systems have enhanced the State's ability to set sentencing goals and, in some instances, limits on what a State is willing to invest in prisons. In view of both the strong trend toward sentencing commissions and the continued growth of prison populations, it is likely interest in sentencing guidelines developed by commissions will continue. However, it should be noted that this report also provides examples of guidelines failing to improve the quality of justice. Guidelines work depending on how they are designed and implemented (and for what purposes). Guidelines should not be expected to solve all the problems surrounding the dispensing of justice and reducing crime. What follows are the Advisory Board's recommendations that States might consider in developing structured sentencing policies. o Commission representation participation. Broad participation in the guideline development process cultivates commitment to the final guidelines and improves their quality. o Resources. Adequate financial support is necessary for a commission to study a State's past sentencing practices; study what other States have accomplished; prepare reports and proposals; meet regularly and hire consultants when necessary; provide ongoing feedback to the court and other governmental agencies; and monitor and evaluate the impact of the sentencing guidelines. o Ongoing monitoring and support. Commissions that have successfully implemented guidelines have received increases in their budgets and staffs in the postimplementation years, probably because of continually increasing responsibilities imposed on the commissions by the legislature and the cost of maintaining and evaluating guidelines. o Appellate review. Without an enforcement mechanism, guidelines are merely voluntary/advisory and so may have little impact on sentencing practices. The common procedure is to provide for an appellate review initiated by either the defense or the prosecution. o Mandatory minimum sentences. The increased use of mandatory minimum penalties is preventing the achievement of the dual goals of reducing disparity and controlling population growth in institutions. States should resist such provisions if they affect large proportions of its sentenced population. o Controlling the use of departures. To help reduce unwarranted disparity, States should specify as much as possible the type of departures from guidelines that are acceptable. Moreover, sentencing commissions should carefully monitor the frequency and direction of departures, at least according to crime category, race, and gender. o Role of the Federal Government. States adopting guidelines need assistance. The Federal Government can play an important role by providing a modest amount of assistance in the following areas: o Establishing and maintaining a national clearinghouse on structured sentencing. o Convening annual meetings for State sentencing commissions to share information and research findings. o Establishing a funding program to help States develop technologies necessary to support the development or modification of existing sentencing structures. Assistance is required, especially in the following areas: o Conducting comprehensive studies of existing sentencing practices that can be used to guide States in the formulation of new sentencing structures. o Developing new methods or providing training to States in existing analytic methods for measuring disparity. o Designing information systems that can be used to monitor compliance with sentencing criteria. o Developing and improving criminal court and correctional population simulation technology (jails, prisons, probation, and parole). o Developing and improving methods of designing and pilot-testing sentencing guideline methods. Finally, more research is needed to assess whether guidelines and other forms of structured sentencing are reducing sentencing disparity. As indicated above, a number of States have implemented, or are about to implement, sentencing guidelines. Independent process evaluations and impact evaluations of these new structured sentencing reforms would be valuable to the field. A better understanding is also needed concerning the effect of reforms external to the guidelines on sentencing disparity, especially along racial lines. Topics to be addressed include the impact of mandatory minimum provisions for certain drug crimes on disparity and the effect of guidelines on shifting discretion from the courts to the front end of the system (arrest, charging, and plea bargaining). Such studies will help clarify how best to correct undesirable and unequal sentencing practices. ------------------------------ Chapter 1 Introduction The past two decades have brought considerable reforms in criminal justice sentencing. Criticisms of excessive disparity and of the rehabilitative ideal have prompted many States to replace indeterminate sentencing with structured sentencing schemes such as determinate sentencing, mandatory minimum penalties, and sentencing guidelines. Nevertheless, most States retain indeterminate sentencing structures while indicating interest in adopting these reforms. The criminal justice community has yet to develop a clear consensus on the basic definitions of structured sentencing and its goals. Indeed, States that have adopted various forms of structured sentencing have done so with multiple goals and expectations. The most frequently cited goals are the following: increasing sentencing fairness; reducing unwarranted disparity, either in the decision to imprison (dispositional disparity) and/or in sentence length (durational disparity); establishing truth in sentencing; and establishing a balance of sentencing policy with limited correctional resources. Structured sentencing reforms can be used to increase the certainty of punishment and mandate the period of imprisonment so that potential offenders are deterred and dangerous offenders are incapacitated. At the same time, sentencing reforms can be used to reduce the likelihood and length of imprisonment for the so-called nondangerous offender. Some State officials hope that these reforms can help them avoid a severe prison-crowding situation by regulating prison population growth according to available correctional resources. However, these purposes are not universally accepted, and the means used to implement them vary considerably among jurisdictions. This study, which was funded by the Bureau of Justice Assistance (BJA) and jointly conducted by the National Council on Crime and Delinquency (NCCD), the Pennsylvania Commission on Crime and Delinquency (PCCD), and the Pennsylvania Commission on Sentencing (PCS), is a review of lessons learned over the past two decades in the diverse attempts to structure sentencing. Structured Sentencing Definitions This examination of the various forms of structured sentencing revealed at the outset the lack of consensus regarding the meaning of basic, commonly used terms such as determinate sentencing, sentencing guidelines, voluntary sentencing guidelines, presumptive guidelines, and advisory guidelines. Therefore, the first task in preparing this report was to develop a series of definitions of terms used throughout the document in the hope of bringing some consistency to the field as well as helping the reader follow this discussion. The following definitions apply throughout the report: o Determinate: Sentences of incarceration in which an offender is given a fixed term that may be reduced by good time or earned time. There are usually explicit standards specifying the amount of punishment and a set release date with no review by an administrative agency (parole board). Postincarceration supervision (parole) may be a part of the sentence. o Indeterminate: Sentences in which an administrative agency, generally a parole board, has the authority to release an offender and determine whether an offender's parole will be revoked for violations of the conditions of release. In one form of indeterminate sentencing, the judge specifies only the maximum sentence length; the associated minimum duration is automatically implied but is not within the judge's discretion. In the more traditional form of indeterminate sentencing, the judge specifies a maximum and minimum duration that is set by statute. The sentencing judge has discretion on the minimum and maximum sentence. o Mandatory minimum: A minimum sentence that is specified by statute and that may be applied for all convictions of a particular crime or a particular crime with special circumstances (e.g., robbery with a firearm or selling drugs to a minor within 1,000 feet of a school). o Presumptive sentencing guidelines: Sentencing that meets the following conditions: (1) the appropriate sentence for an offender in a specific case is presumed to fall within a range of sentences authorized by sentencing guidelines that are adopted by a legislatively created sentencing body, usually a sentencing commission; (2) sentencing judges are expected to sentence within the range or provide written justification for departure; (3) the guidelines provide for some review, usually appellate, of the departure. Presumptive guidelines may employ determinate or indeterminate sentencing structures. o Voluntary/advisory sentencing guidelines: Recommended sentencing policies that are not required by law. Usually based on past sentencing practices, they serve as a guide to judges. The legislature has not mandated their use. Voluntary/advisory guidelines may employ determinate or indeterminate sentencing structures. Focus of This Report In the course of completing the study, researchers visited a number of States to collect various types of information regarding the structure of a State's sentencing system and to document the process of implementation. The researchers also reviewed and synthesized major studies of current structured sentencing systems that assess their impact on disparity, the use of incarceration, and prison crowding. The remainder of this publication provides a broad overview of sentencing reforms and a detailed examination of sentencing guidelines. Chapter 2 provides the reader with a summary of the historical trends in sentencing reforms since 1970. The chapter begins with a discussion of those factors or issues underpinning the structured sentencing movement, such as the disillusionment with indeterminate sentencing and rehabilitation. The chapter then reviews three major forms of sentencing reforms (voluntary/advisory guidelines, determinate sentencing, and presumptive guidelines), providing examples of States that have attempted each type of reform. Chapter 3 describes the results of a national survey of the 50 States and the District of Columbia. This survey represents the first major national assessment of the various sentencing schemes that now exist throughout the Nation. Included in the presentation is a detailed matrix that lists the key sentencing law attributes for each State, including the use of parole and mandatory sentencing provisions. In Chapter 4, the focus shifts to sentencing guidelines and the use of sentencing commissions. The chapter describes the foundation on which a sentencing commission and its guidelines are built: the political and legal context, the State's purposes or goals for reform, the specific legislative structure of the commission, and the legislative mandates included in the enabling legislation. Chapter 5 describes the complex decisionmaking and technical aspects of writing sentencing guidelines. The chapter also provides detailed information about how various States have made decisions in writing sentencing guidelines. Chapter 6 is an analysis of the relative effects of sentencing reforms on a number of key areas, including sentencing disparity, incarceration rates, prison crowding, and future prison population growth. The chapter summarizes the findings of studies on disparity, using an analysis of national criminal justice data to compare trends in incarceration, prison crowding, and crime rates among some of the early guideline States and comparable nonguideline States. The last chapter summarizes the major findings of this study and its policy implications. Specific recommendations are offered for further advancement of sentencing reforms with help from the Federal Government, which can provide additional technical assistance to the States in improving criminal justice sentencing. ------------------------------ Chapter 2 Historical Trends and Issues in Structured Sentencing Controlling discretion has presented a quandary for the criminal justice system throughout this century and particularly during the past 20 years. In his comprehensive review of discretion within the criminal justice system, Samuel Walker makes the following statements: . . . [T]he criminal justice "system" is nothing more than the sum total of a series of discretionary decisions by innumerable officials . . . [the real problem] . . . is not discretion, but its misuse.1 Police misuse of discretion received much attention during the 1960's. During the 1970's, concern over discretion exercised by the courts, parole, and corrections grew as crime rates increased. The past 20 years have produced many accusations but few studies documenting the misuse of discretion by judges, parole boards, and corrections officials, resulting in unwarranted sentencing disparity. Such unwarranted disparity results in undue leniency, excessive harshness, and uncertainty of sanctions.2 The solution to the problem of unwarranted sentencing disparity has been to structure sentencing discretion. However, a variety of mechanisms has been used to structure discretion, resulting in a patchwork of structured sentencing models that includes determinate sentencing, voluntary/advisory guidelines, and presumptive guidelines. What follows is a historical overview of these various reform efforts, which were intended to remove disparity in sentencing as well as improve the credibility of the criminal justice system. Historical Developments Leading to Structured Sentencing The Indeterminate Period At the end of the 19th century, sentencing reform in the United States involved replacing the flat sentence with indeterminate sentences. The criminal justice system was coping with a rapidly increasing number of individuals imprisoned as a result of an expanding population of immigrants and transients, increased efficiency of the police and courts, the fixed sentence, and other factors. Overcrowded prisons and the mere warehousing of inmates resulted.3 At first, piecemeal reforms such as the use of pardons, good time, and probation provided needed flexibility for grappling with an expanding and diversifying prison population. These reforms gradually led to an increase in indeterminate sentences given by 19th-century judges. The movement toward indeterminate sentences was spurred by the Declaration of Principles of the First Congress of the National Prison Association, which met in Cincinnati in 1870. The declaration urged that "preemptory sentences ought to be replaced by those of indeterminate length. Sentences limited only by satisfactory proof of reformation should be substituted for those measured by the mere lapse of time."4 Under indeterminate sentences, offenders received a minimum and maximum sentence, and the parole board determined the time of actual release. Alan Dershowitz later characterized this shift as one from a judicial to an administrative model of sentencing.5 The parole board's determination of when the sentence was served, in turn, depended on its judgment of whether the prisoner had been reformed or cured or had simply served enough time. With indeterminate sentencing, discretion was distributed not only among the prosecutor, defense counsel, and judge but also, for those imprisoned, among prison officials and the parole board. The latter two agencies had considerable influence over an offender's length of stay. Prison officials had discretion over the amount of good time an inmate could earn, which would dictate the prisoner's parole eligibility and/or discharge date. Parole boards controlled the actual release decision for most inmates. The result was a system of sentencing in which there was little understanding or predictability as to who would be imprisoned and for how long. Attacks on Indeterminacy Under indeterminate sentencing, the sentence was individualized so that the punishment fit the criminal rather than the crime. However, this approach prompted numerous accusations of disparity in sentencing and, over time, protests from inmate groups, penologists, and other critics of the penal system. One of the most influential reports critical of indeterminate sentences was the 1971 American Friends Service Committee's work entitled Struggle for Justice: A Report on Crime and Punishment in America, written by "scholars working in the field of criminal justice" and "those who have been on the receiving end of the justice system."6 The report used anecdotal evidence, results of prison riot studies, and personal testimony from inmates to critique the function of punishment and reject the rehabilitation model. Specifically, the report questioned the assumption that crime is a product of individual pathology and that rehabilitation can be achieved within a prison system designed to punish and not treat inmates. It also criticized the assumption that penologists have the knowledge to impose treatment or to accurately predict recidivism to justify discretion in determining when an inmate should be released. Drawing on Struggle for Justice: A Report on Crime and Punishment in America were two other important works also critical of indeterminate sentencing: the report of the Twentieth Century Fund's Task Force on Criminal Sentencing entitled Fair and Certain Punishment (particularly the background paper by Alan Dershowitz)7 and the report of the Committee for the Study of Incarceration, which was published as Andrew von Hirsch's Doing Justice: The Choice of Punishments.8 Both Dershowitz and von Hirsch argued for a shift from indeterminate to more determinate sentencing and for what has been termed the "presumptive sentence." They asserted that these concepts, if implemented, would lead to greater predictability in determining proper sentence disposition (prison versus probation) and length of imprisonment. However, there were important differences in the scope of each author's suggested reforms. Dershowitz's presumptive sentence would require legislatures to set specific penalties for crimes on the basis of the comparative seriousness of the offense, with more serious crimes appropriately receiving a harsher presumptive sentence. The judge would be obliged to impose the statutorily set penalty, which could be raised or lowered through the application of aggravating or mitigating factors by the court. Dershowitz's proposals called for the retention of parole, but with limitations. While Dershowitz called for a shift in emphasis from administrative sentencing to a greater legislative and judicial role, von Hirsch argued for a more radical departure by rejecting the value of general deterrence, rehabilitation, and incapacitation as punishment justifications. He concluded that States should base their sentencing schemes on a "just deserts" or "commensurate deserts" paradigm. Sentencing would be narrowly based on only the seriousness of the offense and the culpability of an offender rather than on an offender's need for "treatment" or on the naive hope that sentencing would somehow affect crime rates. Such a sentencing system would focus not on individualization of punishment but on structuring the system of punishment. The approach would sharply reduce judicial sentencing discretion by establishing punishment categories within which penalties for comparable serious offenses could be grouped on a scale. By using only two factors-- offense severity and, to a limited degree, prior criminal history--a presumptive sentence would be determined. Judicial discretion would thus be limited to setting a presumptive term within a specified sentencing range or justifying departures by applying allowable aggravating or mitigating factors. Departures would be monitored and somewhat circumscribed by allowing judicial review of sentencing, a process that was previously almost nonexistent. Von Hirsch initially insisted that discretionary release by parole boards be abolished. Another influential voice in the early discussions of structured sentencing was that of Judge Marvin Frankel in his book Criminal Sentences: Law Without Order. In 1972, when Frankel's book was published, no models for sentencing guidelines existed. Although he was a judge, Frankel made such a strong case for controlling discretion that the concept of judicial discretion was shattered. Moreover, Frankel foresaw the need to develop a regulatory approach, articulating the principle on which his proposal rested as follows: We boast that ours is a "government of laws, not men." We do not mean by the quoted principle that men make no difference in the administration of law. Among the basic things that we do mean is that all of us, governors and governed alike, are or ought to be bound by laws of general and equal application. We mean, too, that in a just legal order, the laws should be knowable and intelligible so that, to the fullest extent possible, a person meaning to obey the law may know his obligations and predict within decent limits the legal consequences of his conduct.9 The problem according to Frankel was that " . . . the sweeping power of a single judge to determine the sentence, as a matter of largely unreviewable `discretion' is a--perhaps `the'--central evil in the system."10 One part of the solution was the development of appellate review of sentencing. At the time of Frankel's writing, there was "in practical effect no appeal from the trial judge's sentence."11 Like the other authors previously cited, Frankel attacked the indeterminate sentence, primarily on the grounds of cruelty and injustice. However, he suggested that indeterminate sentences may be appropriate for dangerous offenders, drug users, some sex offenders, and juvenile offenders. Frankel recommended that the legislature provide direction as to the purposes and justifications of criminal sanctions. He proposed the establishment of a Commission on Sentencing, which would be a permanent agency formed to study sentencing, corrections, and parole; formulate rules and laws based on its studies; and enact rules subject to checks by Congress, State legislatures, and the courts. However, Frankel did not expect the proposed sentencing commission to escalate the use of imprisonment, which he regarded as already too severe. He hoped that a sentencing commission could act as a buffer, shielding the legislature from political pressures to respond to ever increasing demands for more punitive sanctions and to help control prison population growth. Early Guideline Developments During the same period, researchers who were developing parole release guidelines for the U.S. Board of Parole began to experiment with applying this methodology to sentencing decisions.12 In the 1960's, parole guidelines had been developed by using a two-dimensional matrix relating the seriousness of the instant offense and the probability of recidivism (or the salient factor score) to an expected time served before release on parole. A small range was provided, within which parole hearing examiners had to set the length of incarceration. Departures were permitted, but written reasons for such departures were required. Because the judiciary supported and strongly commended parole guidelines, the concept of guidelines was believed to have value and to be adaptable to sentencing. A model for parole guidelines was applied experimentally to sentencing for the first time between 1974 and 1976 in four jurisdictions. Two sites (Denver and the State of Vermont) were active participants, and two sites (Essex County, New Jersey, and Polk County, Iowa) were observers. This feasibility study was described as an action research project in which researchers studied existing sentencing processes. The judiciary was involved in all phases of the project, from the gathering of all information relevant to sentencing discretion, to the analysis of factors that accounted for the greatest variation in the judges' sentencing decisions. The researchers did not attempt to develop prescriptive notions of what would be a "right" sentence. Instead, they were trying to understand what underlying factors influenced the sentencing decision. The analysis indicated that although there was considerable disparity among judges' sentencing decisions, two factors explained the largest variations: seriousness of the current offense and the extent of the offender's prior record. More importantly, judges who participated in the research expressed support for the concept of guidelines to better help them reach sentencing decisions. The History and Nature of Structured Sentencing Reforms Since the late 1970's, nearly all State sentencing practices have fallen into four broad categories: (1) voluntary/advisory sentencing guidelines, (2) determinate sentencing, (3) presumptive sentencing guidelines, and (4) traditional indeterminate sentencing. Along with mandatory sentencing, the first three models represent the most common forms of structured sentencing, although there are many variations of each. The remainder of this chapter presents a historical overview of structured sentencing efforts to date. The overview identifies the significant issues in sentencing reform that are addressed by this project and describes the key attributes of structured sentencing systems that some jurisdictions have considered or incorporated in developing their own structured sentencing systems. These examples show that structured sentencing is not a single or unified concept. Instead, it embodies a variety of approaches to identifying the particular goals of sentencing in a specific jurisdiction and to achieving those goals by structuring the discretion of sentencers (judges, parole boards, prosecutors, and prison officials) at various points in the criminal justice system. Voluntary/Advisory Sentencing Guidelines Among the earliest sentencing innovations in the United States was the experiment with voluntary/advisory guidelines, called "voluntary" because judges were not required to comply with them. They emerged in part to counter determinate sentencing reforms as adopted in California, Illinois, and Indiana in 1977. The earliest voluntary/advisory guidelines projects originated as a result of a multiyear project beginning in 1974 and funded by the National Institute of Justice. The purpose of the project was to study the feasibility of empirically based guidelines for sentencing.13 The project provided support to researchers who had been instrumental in developing parole guidelines so that the guidelines could be field tested in Denver; Newark, New Jersey; Chicago; Phoenix; and Philadelphia between 1975 and 1980. In addition, experiments with voluntary/advisory guidelines in Florida and Maryland later became institutionalized. In Florida, statewide voluntary/advisory guidelines were revised by a sentencing commission and converted into presumptive guidelines in 1983. In the same year, Maryland formally adopted statewide voluntary/advisory guidelines. The States of Massachusetts, Michigan, Rhode Island, Utah, and Wisconsin also experimented with voluntary/advisory guideline systems at varying levels.14 Several evaluations of these early efforts yielded discouraging results.15 Sparks et al. found that the early Massachusetts guidelines were poorly conceived, poorly developed, and not effectively implemented.16 In its comprehensive assessment of the development and impact of voluntary/advisory guidelines, the National Center for State Courts found difficulties in their implementation in Denver, Chicago, Newark, and Phoenix.17 Specifically, the researchers found major methodological and analytical defects in the development of the Denver guidelines. Furthermore, the Denver guidelines were found to have exerted little influence on judicial decisions on whether to incarcerate, and compliance rates for lengths of prison terms were disappointingly low. The researchers drew the following conclusions: The various measures employed . . . converge on a single conclusion: sentencing guidelines have had no detectable, objectively manifested impact on the exercise of judicial sentencing discretion.18 These disappointing results were often linked to implementation difficulties and the lack of enforcement. In some instances, insufficient time was allowed for the guidelines to take hold. More important, the guidelines were voluntary; judges could simply ignore them. Nonetheless, they were an important step in the sentencing reform process because they helped change sentencing patterns from indeterminate to determinate. By 1994 six States (Arkansas, Louisiana, Maryland, Michigan, Virginia, and Wisconsin) had adopted voluntary/advisory guidelines. Several of these States were the early experimenters with voluntary/advisory guidelines. However, a review of all the major studies conducted on voluntary/advisory guidelines reveals low compliance by judges and, hence, little impact on reducing disparity.19 Although all voluntary/advisory guideline systems are not alike, an example of voluntary/advisory guidelines can be found in Virginia. The Virginia guidelines were not legally mandated as presumptive sentencing guidelines are. Likewise, there was no enabling legislation that set up a sentencing commission to develop the guidelines. Rather, they were based on historical sentencing practices (descriptive versus prescriptive). The process of guidelines development in Virginia exemplifies descriptive guideline development. In 1982, then Governor Charles Robb appointed a task force to examine statewide sentencing practices and to recommend any necessary changes. Between 1983 and 1988, the task force, or Judicial Sentencing Guidelines Committee (JSGC), which is accountable to the Judicial Conference of Virginia (JCV), developed a set of voluntary/advisory guidelines. The guidelines were descriptive: that is, they were developed through statistical analyses of past sentencing practices and then tested in six pilot districts for 1 year, 1988- 1989. The JSGC determined the guidelines to be effective and favorably received by judges in the six pilot districts and, thus, voted to expand their use throughout the State. In 1990, the State legislature commended the JCV for its work and urged adoption of the guidelines. A permanent JSGC (composed of seven circuit court judges) was formed in 1990. The committee decided to base the statewide guidelines on sentencing practices between 1985 and 1989, agreeing that the guidelines should be revised at 1-year intervals in accord with analyses of the previous year's sentencing practices. The primary purpose of the guidelines was the establishment of rational and consistent sentencing standards that reduce unwarranted sentencing disparity. However, Virginia's guidelines were not intended to affect prison populations, change the severity of sentences, change the philosophy of sentencing, or restrict judicial discretion. Instead, they were intended to provide a historical picture, descriptive of past sentencing practices for a variety of offenses, to assist judges in decisionmaking. Virginia's guidelines do not rank offenses according to a severity scale and do not base recommendations on a matrix of offense severity and criminal history scores. The guidelines consist of separate scoring schemes, factors to be considered, and sentence recommendations for eight categories of offenses: homicide, assault, robbery, sexual assault, burglary, larceny, drugs, and fraud. The guidelines are bifurcated in structure to reflect two sets of decisions: (1) whether to imprison an individual and for how long and (2) whether those not imprisoned should go to a local jail for a certain period of time or receive probation. In developing the guidelines, the Virginia committee did not consider any factors influencing sentences that were deemed inappropriate (e.g., race, gender, or age). Therefore, Virginia's model is not purely descriptive because it specifically eliminates from the guidelines certain factors that historically influenced sentencing in Virginia. The Virginia court has significant power to decide who should be incarcerated because that option is always available. In all categories of the Virginia matrices, the court is given a range of sentences based on the past 5 years of sentencing practices. The guideline ranges are updated every year to reflect the most recent sentencing patterns by discarding the oldest data and replacing them with current data. Because Virginia's guidelines were implemented as recently as 1991, there has been no major research on the degree to which they have reduced disparity or how they are used in court decisionmaking. However, the 1989 report on the test guidelines in the six pilot districts indicated that the guidelines foster greater uniformity and less unwarranted disparity and that judges comply with guideline recommendations at a rate of 78 percent. It should be noted that acceptance of the overall policy was not the same as compliance with a guideline recommendation. Because departures (i.e., exceptions) may be permitted or encouraged by the general policy, a 78-percent rate of compliance with the suggested decision outcome may indicate a 100-percent acceptance of policy. Virginia's new Governor, George Allen, established a commission in 1994 to study the abolition of parole and the development of a presumptive sentencing system. The State legislature established a committee to make recommendations on the appropriate role of parole and presumptive sentencing guidelines. In addition, legislation that radically alters the voluntary guideline structure has been adopted. Parole was abolished and the existing sentencing policy modified to greatly lengthen the period of imprisonment. Virginia's recent reforms highlight the volatile nature of sentencing policy. Determinate Sentencing The term "determinate sentencing" is generally used to refer to the sentencing reforms of the late 1970's. In those reforms, the legislatures of California, Illinois, Indiana, and Maine abolished the parole release decision and replaced the indeterminate penalty structure with a fixed (flat) sentence that could be reduced by a significant good-time provision. The only State that has adopted a true determinate sentencing system since 1980 is Arizona, which enacted a "truth in sentencing law" on January 1, 1994. These five States have retained their determinate sentencing models, although no other States have adopted such a structured sentencing scheme. Determinate sentencing was spurred initially by two opposing political forces. Some people attacked the disparities of unfettered judicial discretion that resulted in sentencing biases against the socially disadvantaged.20 On the other hand, the advocates of "law and order" and "get tough" crime policies viewed unfettered judicial discretion as too lenient. Nelson argues "that this twofold drive helps determine what type of determinate sentencing system, if any, will develop in a given jurisdiction since the resultant sentencing will largely be a product of both drives."21 In three of the States (California, Illinois, and Indiana), the legislators provided presumptive ranges of confinement. But those in Illinois and Indiana were so wide that they provided the court with extensive discretion on sentence length. For many offenses, there was no presumptive lead as to whether the sentence should be for, or against, incarceration. Thus, courts were left with extensive discretion in deciding both whether to incarcerate and the length of incarceration. It is arguable that the discretion attacked in these reforms was mainly that of parole boards, and that the discretion lost by parole boards was largely shifted to the courts or to the prosecutors who control the charging function.22 The Illinois law provides a clear example of these points. The legislation, which was passed on July 1, 1977, created six major classes of offenses for which convicted felons could be sentenced to prison: Class M (murder); Class X (robbery, assault, rape, and kidnaping); Class 1 (attempted robbery, rape, and drug sale); Classes 2, 3, and 4, which represent property crimes (burglary, theft, and fraud); drug offenses (possession and sale); and simple robbery. Class X was the most significant sentencing category because it mandated judges to sentence offenders convicted of these crimes to prison for anywhere from 6 to 30 years, with possible enhancements of 30 to 60 years. Offenders sentenced for these offenses began serving longer terms under the new law, despite the fact that inmates also were being awarded increased amounts of statutory good-time credit (day-for-day statutory good time, as opposed to the previous one-third statutory good-time system). The parole board's authority to grant release was abolished, while postrelease supervision was retained. A determinate sentencing system may be based on a "just deserts" concept, a utilitarian or crime control model, or some combination of sentencing goals. A purely utilitarian (or crime control) model of punishment (e.g., concerned with recidivism, incapacitation, and deterrence) focuses little on the crime the offender committed but demands longer sentences for crimes in which consequences associated with incarceration are of most concern.23 On the other hand, a pure "just deserts" model emphasizes the criminal act itself to the exclusion of crime control goals.24 California provides a well-known example of the evolution of a determinate sentencing model. Although numerous subsequent amendments have changed the statute considerably, California initially adopted a "just deserts" sentencing model. The enabling legislation clearly stated that the "purpose of imprisonment is punishment."25 The legislature determined that offenses should be placed into four categories of crimes on the basis of general severity; then presumptive lengths of incarceration were established for each category. In addition, for each presumptive incarceration length, the legislature provided a range for aggravating and mitigating circumstances. Data show the initial ranges provided for the four classifications and the ranges for aggravation and mitigation under California's determinate sentencing law. These ranges apply only if the court selects incarceration. The California model and the State's experience in the first few years of its administration helped to reverse the direction of the development of structured sentencing reform models. The initial determinate sentencing law was projected to have no impact on prison population growth. However, Messinger and Johnson report that during the first year, there were 43 amendments to the legislation, all of which increased the severity of sentences and, thus, the demand for prison space.26 The large number of amendments to the original law, they interpreted, was the result of the legislature's attempts to control sentencing. In Messinger and Johnson's view, California set an example of how not to develop structured sentencing in which the results are strongly influenced by the politics of crime. In a 1983 evaluation of the reform's impact, Casper and Brereton concluded that the amended legislation, as intended, increased the likelihood of incarceration and resulted in a phenomenal growth in prison populations.27 Sentencing Commissions and Presumptive Sentencing Guidelines By the early 1980's, States began to experiment with approaches using sentencing guidelines developed by sentencing commissions. These models differed from determinate and voluntary/advisory guidelines experiments in three respects. First, the guidelines were not developed by the legislature but by a sentencing commission that often represented a diverse array of criminal justice and sometimes private citizen interests. Second, the guidelines were explicit and highly structured, relying on a quantitative scoring instrument. Third, the guidelines were not voluntary/advisory. Judges had to adhere to the sentencing system or provide a written rationale for departures. As in the move to determinate sentencing and voluntary/advisory guidelines, the driving forces stimulating presumptive sentencing guidelines were issues of fairness (including disparity, certainty, and proportionality) and prison crowding. These concerns provided the impetus for States to adopt guidelines, replace indeterminate sentencing with determinate sentencing, and abolish or curtail discretionary parole release. The first four States to adopt presumptive sentencing guideline systems were Minnesota (1980), Pennsylvania (1982), Washington (1983), and Florida (1983). The Minnesota model in particular, with its focus on controlling prison population growth, has often been cited as a successful example of controlling disparity and rising corrections costs through sentencing guidelines. The American Bar Association has endorsed sentencing commission-based guidelines through its Criminal Justice Standards Committee's Sentencing Alternatives and Procedures (adopted by the ABA House of Delegates). In making such an endorsement, the Standards Committee relied "heavily upon the system of presumptive/ordinary offender sentencing pioneered in a guidelines system in Minnesota."28 The Federal Government and 16 States have established commission-based sentencing guidelines. According to the definitions presented in Chapter 1, 10 of the 16 States can be classified as using presumptive sentencing guidelines. The remaining six have voluntary/advisory guideline models. With the exception of mandatory minimum sentencing provisions, sentencing guidelines authored by legislatively created sentencing commissions are now the most popular form of structured sentencing. The next chapter provides a current listing of those States that have adopted guidelines, as well as other State sentencing structures such as determinate sentencing or indeterminate sentencing. Chapters 4 and 5 present detailed descriptions of presumptive sentencing guidelines and the commissions that have written them. ------------------------------ Notes 1. Walker, S. Taming the System: The Control of Discretion in Criminal Justice, 1950-1990. Oxford: Oxford University Press. 1993, p. 4. 2. Sentencing involves the totality of the sentences imposed and the sentences served as carried out by administrative agencies, parole boards, and prison officials. 3. Shane-DuBow, S.A., A.P. Brown, and E. Olsen. Sentencing Reform in the U.S.: History, Content and Effect. Washington, DC: U.S. Department of Justice. 1985. 4. Note 3, p. 5. 5. Dershowitz, A. "Criminal Sentencing in the United States: A Historical and Conceptual Overview." Annals of the American Academy of Political and Social Science 423(1976), pp. 117- 132. 6. American Friends Service Committee. Struggle for Justice: A Report on Crime and Punishment in America. New York: Wang. 1971. 7. Twentieth Century Fund Task Force on Criminal Sentencing. Fair and Certain Punishment. New York: McGraw-Hill. 1976. 8. von Hirsch, A. Doing Justice: The Choice of Punishments. New York: Hill & Wang. 1976. 9. Frankel, M.E. Criminal Sentences: Law Without Order. New York: Hill & Wang. 1972, p. 1. 10. Ibid., p. 69. 11. Ibid., p. 76. 12. Wilkins, L.T., J.M. Kress, D.M. Gottfredson, J Calpin, and A.M. Gelman. Sentencing Guidelines: Structuring Judicial Discretion--Report on Feasibility Study. Washington, DC: U.S. Department of Justice. 1978. 13. Tonry, M. "Structured Sentencing." In Crime and Justice: A Review of the Research, vol. 10, ed. M. Tonry and N. Morris. Chicago: University of Chicago Press. 1988. See also Note 12 and Kress, M. Prescription for Justice: The Theory and Practice of Sentencing Guidelines. Cambridge: Ballinger. 1980. 14. Ibid. See also Note 3. 15. Cohen, J., and J. Helland. "Methodology for Evaluating the Impact of Sentencing Guidelines." Unpublished paper. Pittsburgh: Urban Systems Institute, School of Urban and Public Affairs, Carnegie-Mellon University. 1982. See also Carrow, D.M., J. Feins, B.N.W. Lee, and L. Olinger. Guidelines Without Force: An Evaluation of the Multi-jurisdictional Sentencing Guidelines Field Test. Cambridge: Abt. 1985. See also Sparks, R.F., B.A. Stecher, J. Albanese, and P.L. Shelly. Stumbling Toward Justice: Some Overlooked Research and Policy Questions About Statewide Sentencing Guidelines. Report to the National Institute of Justice, U.S. Department of Justice. Washington, DC: U.S. Government Printing Office. 1982. See also Rich, W.D., L.P. Sutton, T.R. Clear, and M.J. Saks. Sentencing by Mathematics: An Evaluation of the Early Attempts to Develop Sentencing Guidelines. Williamsburg, VA: National Center for State Courts. 1982. 16. Sparks, R.F., B.A. Stecher, J. Albanese, and P.L. Shelly. Stumbling Toward Justice: Some Overlooked Research and Policy Questions About Statewide Sentencing Guidelines. Report to the National Institute of Justice, U.S. Department of Justice. Washington, DC: U.S. Government Printing Office. 1982. 17. Rich, W.D., L.P. Sutton, T.R. Clear, and M.J. Saks. Sentencing by Mathematics: An Evaluation of the Early Attempts to Develop Sentencing Guidelines. Williamsburg, VA: National Center for State Courts. 1982. 18. Ibid., p. xxiv. 19. Cohen, J., and M.H. Tonry. "Sentencing Reforms and Their Impacts." In Research on Sentencing: The Search for Reform, ed. A. Blumstein, J. Cohen, S.E. Martin, and M.H. Tonry, pp. 305-459. Washington, DC: National Academy Press. 1983. 20. Nelson, B. "The Minnesota Sentencing Guidelines: The Effects of Determinate Sentencing on Disparities in Sentencing Decisions." Law and Inequality 10(3)(1992), p. 217. See also Frankel, M.E. Struggle for Justice. New York: Hill & Wang. 1971. See also Messinger, S.L., and P.D. Johnson. California's Determinate Sentence Statute: History and Issues in Determinate Sentencing: Reform or Regression? Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. 1978. 21. Ibid. 22. Alschuler, A.W. "Departures and Plea Agreements Under the Sentencing Guidelines." Federal Rules Decision 117(1988), pp. 459-476. 23. Note 19. 24. Note 23. See also von Hirsch, A. "The Politics of Just Deserts." Canada Journal of Criminology 32(1990), p. 397. 25. California Penal Code, Section 1170(a)(1). 26. Messinger, S.L., and P.D. Johnson. California's Determinate Sentence Statute. History and Issues in Determinate Sentencing: Reform or Regression? Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. 1978. 27. Casper, J., and D. Brereton. "Evaluating Criminal Justice Reforms." Law and Society 18(1984), pp. 121-144. 28. American Bar Association. Standards for Criminal Justice--Sentencing Alternatives and Procedures. 3d ed. Boston: Little, Brown & Company. 1993. ------------------------------ Chapter 3 An Overview of Current Sentencing Practices in the United States This chapter describes the results of a national survey of sentencing practices in the United States conducted in late 1993 and early 1994. The survey represents the first major national assessment of existing sentencing practices. Information for this chapter was also culled from several other data sources besides the national survey. The Pennsylvania Commission on Sentencing maintains an extensive library containing documents on sentencing practices in a number of the States. The United States Sentencing Commission (USSC) recently conducted a national survey on mandatory sentencing for drug offenses. Followup telephone calls were made to some individuals who have the most detailed knowledge of their States' sentencing practices. Finally, information was supplemented by site visits to some States. It should be noted that this chapter summarizes some very complex State sentencing statutes. A primary objective of the national survey was to classify States according to sentencing structure. This classification required two steps: establishing operational definitions of sentencing practices (see Chapter 1) and surveying State respondents to determine how many of these sentencing practice types applied to their States. Data show how the responding States classified themselves according to the definitions outlined in Chapter 1 for determinate, indeterminate, and/or sentencing guideline structures. Including the District of Columbia, 29 States reported having indeterminate sentencing. Some form of determinate sentencing was indicated by 20 States, and 16 States have sentencing guidelines. All States reported having a variety of mandatory minimum incarceration sentencing. Data reflect the fact that State respondents could check all the sentencing practices that applied to their States. (Fourteen respondents checked multiple sentencing practices, excluding mandatory sentencing.) For example, some States have determinate sentencing for one class of offenses and indeterminate sentencing for another class. A respondent also may have checked both indeterminate sentencing and presumptive sentencing guidelines, indicating that the State uses guidelines within an indeterminate sentencing structure. Followup contacts with several respondents revealed (1) that they checked off presumptive sentencing guidelines but meant to check presumptive sentencing (guidance or presumptions that must be considered by the court in sentencing) or (2) that only some judicial districts within the State used informal guidelines. Data summarize the primary forms of sentencing practices in the United States on the basis of definitions presented in Chapter 1. States were classified according to their primary sentencing structure on the basis of all of the information received from the survey and followup contacts. For example, a State with sentencing guidelines and indeterminate sentencing was placed in the sentencing guidelines category; States with both indeterminate and determinate features--such as Alabama, Alaska, and Ohio--were classified as indeterminate if most inmates are sentenced under the indeterminate structure. Including the District of Columbia, 30 States were therefore classified as indeterminate sentencing types; 5 States were classified as determinate sentencing types; and 16 States (including North Carolina, whose guidelines became effective in October 1994) were classified as sentencing guidelines types. Mandatory Minimum Incarceration Sentences in the United States The national survey asked questions regarding mandatory minimum sentences: for example, if mandatory sentences existed, and if so, for which offenses. States were also asked to provide more detailed information on any mandatory minimum incarceration sentences for cases that involve controlled substances: drug type, amount of drug, and penalty for which the mandatory sentence applies. States also were asked to submit copies of statutes concerning all mandatory minimum sentences. Collecting data on mandatory minimum sentences was difficult. Originally, data were to be collected from States with mandatory minimum periods of incarceration for selected crimes; however, several States initially reported no mandatory sentences because (1) the sentences exist in statute but are not used, (2) an offender may receive the mandatory sentence but is still eligible for early release via good time and parole, or (3) the sentence is mandatory only because incarceration cannot be suspended. It was decided to include all these possibilities in the presentation of data. Data summarize the results of the survey on mandatory minimum sentencing. All States have some form of mandatory sentencing provisions; the most popular application of mandatory sentencing is for repeat or habitual offenders (41 States) and for crimes accompanied by the possession of a deadly weapon (41 States). Survey responses indicate that Hawaii, Kansas, Kentucky, Louisiana, Maine, Nebraska, New Hampshire, New Mexico, New York, Ohio, Oklahoma, Oregon, Tennessee, Texas, Utah, Vermont, Virginia, Washington, and Wyoming lack any drug mandatory minimums. Data indicate that mandatory minimum sentences are imposed for many offenses in Colorado, Illinois, Pennsylvania, Rhode Island, and Wisconsin. Summarizing all the information collected on mandatory minimum sentences for drug offenses is impossible because the types of drugs, amounts required for the mandatory sentences to apply, and penalties vary widely. Approximately 20 States have implemented mandatory minimum sentences for drug sales to minors and sales within a specified distance (usually 1,000 feet) of a school. One survey respondent stated that the mandatory minimum does not apply if the offender is shown to be a "drug-dependent person." Currency of State Criminal Codes States were asked to report when their criminal codes were most recently revised or rewritten to change the elements of offenses. Most States have not significantly revised their codes since the 1970's. This question proved difficult to answer for States in which revision of the criminal code is an ongoing process, with yearly changes taking place. Several large States reported that their codes had not been revised for some time: since 1872 for California, 1935 for Michigan, and 1967 for New York. Only nine States reported major revisions in the 1990's, and six States reported major revisions in the 1980's. States were also asked to indicate whether any commissions, task forces, or committees had recently been formed to study and/or make recommendations to change State sentencing practices. The following States are in the midst of studying changes: o Alaska established a sentencing commission in 1990, charging it with addressing sentencing reform and prison overcrowding. The commission released a report in December 1992 recommending changes in areas such as the use of alternative punishments, parole, and inmate classification. o Hawaii formed a committee to review any needed changes to the penal code and to report to the legislature. o Iowa established the Intermediate Crime Sanctions Task Force to recommend sentencing reforms. o The Kentucky Governor's Commission on Quality and Efficiency has recommended the implementation of sentencing guidelines. o Massachusetts is considering a proposal to adopt a determinate sentencing structure, has recently created a sentencing commission, and is considering sentencing guidelines. o Michigan established a sentencing commission to study the feasibility of developing sentencing guidelines. A specific goal of the commission is to further the concept of "truth in sentencing." o Missouri established a sentencing commission to study the State's sentencing practices and identify sentencing disparity. o Ohio established a sentencing commission in 1991 to study sentencing laws and correctional resources. A legislative bill was introduced to eliminate parole and good time and to create flat sentences (with a series of presumptions). o Oklahoma is considering a determinate "truth in sentencing" guideline structure. A commission has been established to develop recommendations for the legislature by 1995. Use of Parole and Good Time The survey asked questions about discretionary release from correctional facilities and the use of good time in these facilities. In all, 48 States reported having parole or some other form of postrelease supervision, and 48 States reported having good time for inmates at the local and/or State levels. Hawaii, Pennsylvania, and Utah reported not having good-time provisions for inmates. Only Arizona, Florida, and Maine lack parole supervision (except for inmates who still fall under the old sentencing structures, or for a small number of offenders sentenced to life with the possibility of parole). However, only Maine and the Federal system have apparently abolished parole boards and all forms of postrelease supervision for all offenders. In some States, such as California, Delaware, Illinois, and Oregon, parole in its purest sense has been eliminated, replaced by some form of postrelease supervision that is not called parole. Sentencing Commissions in the United States Information on States that have adopted sentencing guidelines and have sentencing commissions is summarized in data. It should be noted that use of the term "commission" is a generalization; entities in some States may have other names, such as "committee." In 10 of the 16 States having sentencing commissions and sentencing guidelines, the guidelines are presumptive; the other 6 States (Arkansas, Louisiana, Maryland, Michigan, Virginia, and Wisconsin) have voluntary/advisory guidelines. In addition to the 16 States listed, at least 5 other States (Kentucky, Massachusetts, Missouri, Ohio, and Oklahoma) have guidelines and/or a sentencing commission under study. Although it has voluntary/advisory guidelines, Michigan has established a sentencing commission to further study its sentencing practices and to propose a more structured sentencing scheme. In South Carolina, the status of guidelines remains in flux. Initial data indicated that the guideline commission would cease to exist in June 1994 and that guidelines would not be implemented. Although sentencing guidelines have been the dominant form of sentencing reform during the past two decades, it is also true that most States have not adopted that sentencing structure, and several others have tried but failed to implement guidelines. Some of the factors that impede or facilitate a State's ability to adopt sentencing guidelines are addressed in the following two chapters. ------------------------------ Chapter 4 Sentencing Commission Structures and Their Mandates The sentencing commission's sociopolitical environment and legislative mandate strongly influence its role, the quality of its guidelines, and its ability to implement--and obtain conformance to--its sentencing guidelines. For purposes of discussion in this chapter, a sentencing commission is a State agency that issues guidelines for the sentencing court. The purposes of the reform, the specific legislative structure of the commission and the mandates included in the enabling legislation, and the political and legal context are examined in the following pages. The collection of information for this report included the study of documents provided by all the State guideline systems and the United States Sentencing Commission (USSC). Visits were made to Delaware, Kansas, Louisiana, Maryland, North Carolina, South Carolina, Tennessee, Virginia, and Washington. In these States, key individuals involved in the development of the sentencing guidelines were interviewed. Finally, where guidelines have been implemented, available data on the impact of the guidelines were reviewed. However, because of the dynamic quality of sentencing--sentencing commissions and sentencing guidelines in particular--important changes occurred constantly and material had to be updated regularly. The classification of State sentencing systems was simplified in areas such as the degree of determinateness and the degree of presumptiveness for purposes of presentation. Therefore, some will disagree or argue that their State system is more complex. However, to communicate general trends to legislators and others who may know little about sentencing, it was necessary to simplify the discussion. Purposes of Reform States create sentencing commissions for many reasons. Researchers found that the most frequently cited reasons are to increase sentencing fairness, to reduce unwarranted disparity, to establish truth in sentencing, to reduce or control prison crowding, and to establish standards for appellate review of sentences. These purposes are not all or universally accepted, and the means used to implement them vary among jurisdictions. The purposes outlined by commissions influence nearly every aspect of the guideline development process. Therefore, each guideline system should be evaluated against its stated goals. This section describes the issues and the various approaches that many States and the Federal Government have taken in addressing their specific purposes. Fairness and Sentencing Disparity Fairness covers a number of issues that sentencing reforms have addressed over the past 20 years. Fairness refers to reducing unwarranted sentencing disparity, making sentences commensurate with the seriousness of offenses, and increasing certainty and predictability.1 For purposes of this discussion, certainty and predictability are considered issues related to what is commonly referred to as truth in sentencing, described in the next subsection. Although sentencing disparity is the primary fairness concern, it is also difficult to define. Generally, disparity is defined as different sentencing of "similarly situated" offenders and the similar sentencing of dissimilar offenders. However, the meaning of "similarly situated" can vary. For some it means that sentences should be proportionate to the severity of the offense.2 For others, it means that sentencing should focus on "culpability" or the seriousness of the offense considered with any prior record.3 Guidelines have generally provided detailed measures of offense severity and criminal history, relying on courts to consider other relevant sentencing factors. Proportionality is one of the most important principles in establishing a fair and equitable sentencing system. In setting sentences that are proportional, two dimensions are considered: ordinal proportionality, which refers to the setting of punishments relative to the offenses, and cardinal proportionality, which refers to the "absolute severity levels . . . chosen to anchor the penalty scale."4 Sentencing commissions generally begin drafting guidelines by placing offenses into levels of severity and then assigning penalties (prison, intermediate punishment, or probation and sentence lengths) to these severity levels. Commissions therefore establish sentences proportionate to each other (what von Hirsch refers to as ordinal proportionality5). For example, if the sentencing commission establishes a 2-year term of incarceration for assault with serious bodily injury and a 1-year term for assault with bodily injury, the assault with serious bodily injury is considered twice as severe as the assault with bodily injury. Cardinal proportionality, however, determines the overall penalty structure. For example, the USSC used the mandatory penalties established by Congress as its benchmark. Consequently, the overall scale of sentencing guidelines for Federal drug sentences are the drug mandatory sentences set by Congress. Similarly, the State of Washington's Sentencing Guidelines Commission (WSGC) chose its mandatory sentences for violent offenses to establish their most severe sentence recommendations. When establishing other penalties, WSGC worked down from these recommendations. Truth in Sentencing A recent theme of the structured sentencing reform movement is truth in sentencing. Truth in sentencing attacks sentencing models that allow offenders to serve only a portion of their imposed sentence. According to the U.S. Department of Justice, the average sentence imposed in 1990 was 65 months, while the average time served in prison was 22 months, or 34 percent of the sentence.6 Truth in sentencing is a legislative means of restoring public confidence in their criminal justice system by requiring offenders to serve their full (or almost full) sentences. Such legislation brings certainty and predictability to sentencing. Most truth in sentencing systems adopt several essential ingredients. First, they generally eliminate the indeterminate sentence, replacing it with a determinate sentencing format. Second, the sentence reflects the actual amount of time an offender will serve, with very limited amounts of good or earned time deducted from the sentence. Third, the parole release decision is eliminated, but postrelease supervision is usually retained. Such a sentencing system increases certainty and predictability for the duration of confinement. However, in most such systems, the decision about who will be incarcerated is less structured and less predictable. The discussion of the California and Illinois determinate sentencing systems in Chapter 2 exemplifies the uncertainty of the confinement decision under these determinate sentencing systems. When combined with statutorily enacted truth in sentencing provisions, sentencing guideline systems can reduce disparity and increase certainty about the decision of incarceration. Because sentencing commissions often concern themselves with both correctional resources and unwarranted sentencing disparity, they generally address the length of incarceration and who is incarcerated. Moreover, as sentencing guideline systems evolve over time, they are focusing on the vast array of intermediate punishments that States are developing. Thus, incarceration is often being exchanged for intermediate punishments such as house arrest, intensive supervision, and drug treatment. Prison Crowding Severe prison crowding stimulated the growth of sentencing commissions and sentencing guidelines in many States. The crisis of prison crowding strains State budgets and requires States to reassess their correctional resources. Several States with guidelines indicated that the problem of crowding became so severe that it drove the creation of a sentencing commission and capacity-linked sentencing guidelines. Who is incarcerated and for how long was an issue not only of fairness but also of economics. Many State correctional systems are being pressured to relieve prison crowding. In response, some States have implemented emergency release measures, accelerated parole release, and instituted other methods to control inmate populations. Moreover, States that have built additional space and avoided court intervention are facing skyrocketing correctional budgets. State legislatures must deal with many demands on extremely tight budgets. With correctional growth taking an increasingly large share of State budgets, legislators are looking for ways to manage the growth of correctional populations and thus the growth of the correctional budget. Consequently, some States have turned to commission-authored sentencing guidelines and the reform of sentencing structures to control who should go to prison and for how long. The guideline systems offer the ability to project future prison populations. Within the guidelines, resource-sensitive policy choices can be made and, if necessary, adjusted to manage expensive correctional capacity. The experiences of Minnesota, Oregon, and Washington are instructive in that these States have developed guidelines to manage who goes to prison and for how long. These States can therefore anticipate future needs and can focus prison resources on specific offenders, thereby conserving prison resources and ensuring space for violent offenders. Overcrowding is not divorced from the issue of fairness. Overcrowded prison populations can create dangerous and unlawful environments for the management and treatment of offenders. The principles of fairness and humanity dictate that offenders are provided a reasonable living environment. Thus, both guideline and nonguideline States have established capacity limits to reflect fairness as well as economic concerns. Sentencing Commission Structure, Mandates, and Organization Well-conceived enabling legislation is crucial for the success of a sentencing commission and its guidelines. First, the legislation identifies participants in the development of the sentencing guidelines (commission members). Second, it sets forth the legal mandates for the commission such as factors that must be considered in writing guidelines. Third, it establishes the implementation process and enforcement mechanism of the guidelines. Finally, it communicates the political agenda to the commission. In the following section, these and other aspects of a sentencing commission's structure and organization are reviewed. Commission Membership Data identify States that have established sentencing commissions, along with the membership and size of each commission. These data illustrate the range in commission size (from 7 on the United States Sentencing Commission (USSC) to 28 in North Carolina) and commission diversity. It should also be noted that except for USSC, all of the State sentencing commission members serve as part-time members. Commission membership and size reflect the interests of a State in creating a commission and the groups that are defined as significant in writing sentencing guidelines. For example, on Pennsylvania's 11-member sentencing commission, which is not mandated to manage or consider prison population size, corrections is not represented. Judicial and legislative appointments represent more than three-fourths of the commission's membership. When Pennsylvania's commission was created, its prisons were not suffering from overcrowding: there was apparently no need for a strong corrections presence. Because judges and legislators were seen as the major sentencing constituents, they played significant roles on the commission. Susan Martin juxtaposed Pennsylvania's judicially and legislatively focused commission with Minnesota's representation, which included private citizens and correctional representation.7 According to Martin, this difference in membership is a significant reason for Minnesota's sensitivity to correctional resources, compared with Pennsylvania's greater sensitivity to a political agenda. Kramer et al. assessed the impact of this difference in their study of the policy choices.8 They applied the Minnesota, Pennsylvania, and Washington guidelines to more than 8,000 felony cases and found that Minnesota and Washington, with their focus on correctional capacity, established guidelines that preserved resources for violent offenders in the State system. Pennsylvania, with no capacity constraint, had a more punitive guideline system overall. The severity of Pennsylvania's sentencing was based primarily on its more severe sentences for nonviolent offenders. The most obvious reasons for these differences were the legislative mandates for the Minnesota and Washington sentencing commissions to consider prison capacity and the representation of correctional interests on their State commissions. Pennsylvania had neither and consequently wrote guidelines increasing certainty and length of incarceration.9 The type and breadth of representation on State sentencing commissions influence the quality of the guidelines that are written and the ability of a commission to ensure implementation of the guidelines. For example, an interview with Robin Lubitz of the North Carolina Sentencing and Policy Advisory Commission revealed that its 28-member commission was a tremendous asset in developing support for both the concept of sentencing guidelines and the guidelines themselves: . . . although that size made it very difficult in the beginning, when it coalesced it was a potent force. So in retrospect if I were to do it again, I would probably have a Commission about the same size even though if you had asked me that a couple of years ago I would have said it was much too big to be effective.10 Oregon's guideline development process is another example of how broad-based participation fostered commitment to the guidelines among different criminal justice factions. Kathleen Bogan, former Executive Director of the Commission, indicated that: The strength of Oregon's approach was that all three branches of government, including judges and legislators, participated in the development of the guidelines at the council level, while at the same time avoiding any unconstitutional mix of the three branches in a promulgating body, thus broadening support for and understanding of the guidelines.11 The Oregon sentencing commission consists of judges, defense attorneys, prosecutors, legislators, and criminal justice professionals (including corrections officials). The Oregon experience illustrates the principle that when various groups contribute to the development of a guideline system, members of those groups will more likely take a proprietary and supportive interest in the final product. Sentencing Commission's Mandate and Purposes There are several important themes adopted by legislatures as principles under which a commission is to operate that emerge, as do several principles to be fulfilled by commission guidelines. This section presents some of these themes for States considering the creation of a commission or for States with commissions that may be considering revisions of their commissions' mandates. Purpose of Sentence. Judges have historically been responsible for passing sentences for diverse and conflicting purposes, including to rehabilitate, deter, incapacitate, restore, and punish. Many commissions are mandated to accomplish these purposes, whereas others prioritize them. For example, Arkansas, a State that recently implemented its guidelines, endorses all of these purposes of sentencing in its legislation, but leaves how to accomplish them to the guidelines and to the discretion of the judge implementing the guidelines. Clearly, many of these purposes of sentencing contradict each other: if one takes priority, the ability to accomplish another one is diminished. For example, when a sentence is intended to "punish an offender commensurate with the nature and extent of the harm caused by the offense, taking into account factors that may diminish or increase an offender's culpability,"12 how can this punishment be consistent with policies attempting to rehabilitate the offender? Commensurate punishments may sometimes be consistent with rehabilitative sentences; however, more often than not, the sentence imposed to meet a punishment goal will be inconsistent with a sentence passed to rehabilitate an offender. Minnesota's enabling legislation seems to recognize the potential dissonance among the goals of sentencing and sets forth a mandate stating that "[D]evelopment of a rational and consistent sentencing policy requires that the severity of sanctions increase in direct proportion to increases in the severity of criminal offenses and the severity of criminal histories of convicted felons."13 Although the reference to criminal history might have spurred the commission to develop an ineffective model, the commission used this mandate to adopt a modified just-desert model of sentencing. Pennsylvania adopted guideline modifications whose purposes vary depending on the severity of the offense and the severity of a defendant's prior record. Although Pennsylvania emphasizes punishment overall, the commission prioritizes other purposes that it has created in the guidelines. About offenders receiving sentences that call for relatively short terms of incarceration or that allow for nonconfinement options, the Pennsylvania commission states, "Many [of them] . . . suffer from drug and/or alcohol problems and the court should consider a treatment component to address the rehabilitative needs of such offenders."14 It should be noted that legislatures have failed in most circumstances to prioritize the purposes of sentencing; this failure has left the commission with the difficult responsibility of drafting sentencing guidelines that must address conflicting purposes. In most cases (except in Minnesota), commissions have therefore developed guidelines using measures of offense severity and criminal history, leaving to the courts the discretion to aggravate and mitigate the sentence as a means of considering rehabilitation and other sentencing purposes. Fairness and Certainty. According to data, many States direct their sentencing commissions to make sentences fair and certain. The concept of fairness and certainty is inherent in the reform of sentencing models such as the replacement of indeterminate sentencing with determinate sentencing. This reform usually involves the elimination of parole release, or the allowance of good time that varies from a 33-percent reduction in Minnesota to a 15-percent reduction in the Federal determinate sentencing guideline system. The guideline system constructed by the commission includes the ranges within the guidelines, the standard on appeal when the court departs from the guidelines, and the consideration of factors that may allow the court to consider what social scientists call extralegal factors, including education level and employment status. Minnesota's enabling legislation directed that "[S]entencing should be neutral with respect to race, gender, social, or economic status of convicted felons."15 The Federal system and Tennessee's enabling legislation contained similar mandates. Other States generally do not address these issues; the consideration of factors such as education and employment status, which may be indirectly racially linked, are therefore allowed unless the commission sets forth such a policy in the guidelines. Correctional Capacity Mandate Many State legislatures direct their sentencing commissions to consider prison capacity as they develop guidelines. This fact suggests the importance of prison crowding in the creation of sentencing commissions. In many instances, the enabling legislation requires the commission to estimate the fiscal impact of the guidelines on correctional resources. Minnesota, Oregon, and Washington have enabling legislation that requires guidelines to consider correctional resources. Minnesota's legislation set a precedent in 1978 by specifying that " . . . the commission shall take into substantial consideration correctional resources . . . " and Oregon's legislation states that "factors relevant to appropriate sentencing include . . . effective capacity of State and local corrections facilities. . . . "16 Washington was specifically mandated to make "frugal use of the state's resources." Its enabling legislation requires (1) the sentencing commission to study State correctional capacity, projecting the impact of guidelines on that capacity; and (2) the commission to prepare alternative sentence recommendations that do not exceed capacity (if guidelines are projected to exceed prison capacity). Furthermore, the legislation enables Washington's governor to declare a state of emergency regarding State prison crowding and to call the sentencing commission into emergency meetings to deal with the crowding situation. The issue of capacity restraint is important for limiting the severity of sanctions (number incarcerated and length). Susan Martin's research on the guideline process in Minnesota and Pennsylvania concludes that Minnesota's adoption of a capacity agenda imposed reasonable restrictions on the ability of the sentencing commission to rely on imprisonment as a sentencing option and a foil to critics who pushed for greater severity. Martin argued that the Minnesota Commission created a rationale for reasonable restraint on the capacity to punish.17 On the other hand, Pennsylvania, acting without a prison capacity limit on its guidelines, struggled to counter arguments for harsher penalties.18 Method of Adoption The method of adopting guidelines is important because it defines the final author of the guidelines and how strictly the guidelines are to be applied. There are three basic approaches to adopting guidelines. The first approach entails creating a commission mandated to draft guidelines that must be submitted in bill form to the legislature. For example, in Washington the legislature may adopt the guidelines as submitted, reject them, or amend them. The second approach is to delegate the commission to write the guidelines that become law even without further action on the part of the legislature. For example, in Pennsylvania 90 days after guidelines are submitted to the legislature, they become law unless a concurrent resolution to reject is passed. A third approach is the use of administrative rules or administrative order. Wisconsin's guidelines, enabled in bill form and legislated, are adopted through administrative rules, promulgated by the commission, and approved by the legislature. Delaware's guidelines, however, are issued by an administrative order from the Supreme Court. Each approach has advantages and disadvantages. Guidelines legislatively passed in bill form become law and obtain the standing of law. The major concern with such a process is that the legislature will either fail to act on the bill or will amend it during its passage. For example, New York's sentencing commission's guidelines were submitted in bill form but were never passed by the legislature; the commission's funding was eventually discontinued, causing its dissolution.19 Some blame such failures primarily on the commission's inability to obtain commitment to its guidelines.20 More recently, Tonry has recognized the political problems facing commissions.21 The political advantages of legislatively enacted guidelines result from the explicit support of the legislature and the governor as an outgrowth of direct legislative action. Legislative enactment dissolves any constitutional concerns such as those involving judicial participation in a legislative body (as in Pennsylvania) or delegation of authority by the Congress to a commission whose members are all appointed by the President (as in USSC).22 Furthermore, appellate courts reviewing sentencing under legislatively adopted guidelines may be more likely to review sentences on substantive (and not just procedural) issues, although this is not yet established.23 By contrast, under the "adopted unless rejected" approach such as exists in Pennsylvania, guidelines are enacted by the legislature's failure to reject them. Although such commissions are delegated much more power, the guidelines may have less standing under appellate review, and there may be constitutional concerns with legislative delegation to a commission whose members come from the judiciary and are appointed by the governor. Implementation through administrative rules or orders also has advantages and disadvantages. Delaware, a State whose guidelines are issued through administrative order by its Supreme Court, finds that this process allows substantial flexibility. Implementation of and amendments to any recommendations are possible within a relatively short time, and amendments are not as likely to be rejected or altered by the Supreme Court. However, sentences imposed under guidelines issued by administrative order generally are not appealable; the guidelines are therefore merely voluntary. The choice as to which model to follow may reflect the willingness of legislatures to delegate authority and/or to become more significantly involved in establishing sentences. North Carolina, whose guidelines are submitted in bill form, has a tradition of strong legislative control. Its legislature was averse to delegating broad-based power to the sentencing commission.24 However, legislative enactment risks the passage of amendments to the guidelines or failure to implement them. In New York, the legislature thwarted the guideline movement by failing to pass proposed guidelines.25 Such occurrences are rare but demonstrate the risks involved in requiring legislative enactment. The most common approach is to require the adoption of guidelines in bill form. It should be noted that Florida moved from a "failure to reject" approach to legislatively adopting its guidelines. The State of Washington, on the other hand, has researched changing its approach from legislative adoption to the failure to reject approach. Financial Support Strong legal and broad-based political support are important aspects of the ability of a sentencing commission to write guidelines. However, it is essential that adequate financial support be obtained for a commission to study past sentencing practices in the State; study what other States have accomplished; prepare reports and proposals for the commission to consider; pay for the commission to meet regularly and hire consultants when necessary; provide ongoing feedback to the court and other governmental agencies; monitor and evaluate the impact of guidelines; and conduct other basic and applied research on a large variety of sentencing issues. Cost varies considerably among States, depending on commission needs. The annual costs for States range between $250,000 and $650,000. If insufficient resources are provided for a commission's work, the risk of failure increases. The commission needs independence to hire adequate staff, to pay for travel, and to support other basic functions and activities. It is crucial that the commission's ongoing functions of monitoring guidelines, revising guidelines, and assessing the impact of legislation and other functions taken on as part of its sentencing function be fully considered and that adequate funding be provided. As a side note, a number of commissions that have successfully implemented guidelines have realized increases in budget and in staff size in the post-implementation years. These increases are the result of continually increasing responsibilities placed on the commissions by the legislature and the cost of monitoring and evaluating guidelines. Commission Support Staff Organization The size and composition of staffs supporting the work of sentencing commissions is relatively similar among States. USSC is the one exception to this rule. In 1991 USSC had 100 full-time staff and an appropriation close to $8.5 million. State sentencing commissions have implemented guidelines systems with 3 to 22 full-time staff members. Most States have had between five and seven full-time staff members during guideline development and implementation stages. Washington and Pennsylvania initially had five full-time and one half-time staff members, and Minnesota had seven. More recently, North Carolina has been operating with a staff of seven. Staff composition in the States has been heavily weighted toward social science researchers.26 Staffs are generally composed of an executive director, a policy director, a research director, an administrative assistant, and a secretary. Additionally, States have hired research associates (usually at least one with statistical experience) and data coders/input operators to supplement research efforts. Generally, commission staffs can be called upon to perform any of the following functions: designing and implementing a data collection system; providing information on the status of sentencing patterns; projecting the impact of any changes that guidelines might have on current trends; drafting guideline policies; monitoring appellate court decisions; conducting training and education seminars related to the guidelines; answering questions over the telephone on guideline application issues; and preparing meeting materials for the commission. Typically, at least one staff member maintains regular contact with legislators and legislative staff members. Commission Meetings The frequency with which commissions meet varies among States. In the initial stages of guidelines development, meetings are held much more frequently than after implementation. In fact, it is not uncommon in the early drafting stages for commission members to spend evenings, weekends, and multiple days per week in meetings. Some States have defined the minimum frequency for meetings in their enabling legislation. For example, in Arkansas and Pennsylvania, the enabling legislation dictates that the commission meet at least quarterly. After the initial stages, some States have regularly scheduled meetings. For example, the State of Washington has scheduled standing meetings for the full commission for the second Friday of the month; to cancel these meetings, they must be taken off the schedule. Adequate travel funds must be budgeted to support the frequent meetings and overnight stays required for a commission to meet its deadlines. Guideline Implementation The power of a commission to implement and monitor sentencing varies according to its enabling legislation. However, the enabling legislation should give a commission the authority to train judges, prosecuting attorneys, defense attorneys, probation officers, and others in the sentencing process. Strong enabling legislation also authorizes commissions to establish statistical monitoring systems. A statistical monitoring system allows a commission to assess the impact and quality of guideline implementation and to identify areas that need revision. For example, a commission may depend on the courts to accurately interpret the guidelines and to review justifications for departures from the guidelines. ABA standards state that the commission "should also be charged with responsibility to collect, evaluate and disseminate information regarding sentences imposed and carried out within the jurisdiction."27 Guideline Enforcement The presumptiveness of guidelines ultimately stems from the process by which sentences are reviewed. Without an enforcement mechanism, guidelines are merely voluntary and so may have little impact on changing sentencing practices. The common procedure is to provide for appellate review, with the review to be initiated by either the defense or the prosecution. The term "presumptive" is often used to describe guideline systems that rely on appellate review as an enforcement mechanism. Guidelines (such as those for Arkansas) that allow for no review are referred to as voluntary: The guidelines must be considered, but judges are free to depart without challenge. Rather than a dichotomy, this situation is more of a continuum from voluntary to very presumptive. The degree of presumptiveness is based on whether some form of appellate review exists and, if so, what standards the courts are held to on appellate review. Little research on or discussion of this issue has occurred, despite the hope that guidelines would stimulate a common law of sentencing. However, the works of McCloskey28 and Del Sole29 offer some information. The least presumptive (voluntary) form of guidelines would resemble the guidelines of Arkansas and Virginia, which do not allow appellate review of sentences. In States with some form of review, two factors affect the presumptiveness of guidelines. One of the most significant factors is the standard for departure that is established in the statutory mandate. Enabling legislation might specify "unreasonable" sentences as those that warrant a judge's departure from sentencing guidelines, as in Pennsylvania; or legislation might define a much higher standard, such as "clear and convincing" reasons for departure, as in Minnesota. The second factor is the interpretation that the appellate courts place on the terms. Obviously, the higher the standard, the greater the likelihood that the appellate courts will hold the lower court more accountable for departures. Similarly, the lower the standard, the greater the likelihood that the appellate court will allow greater latitude in departures from guidelines. During the 12 years that the Minnesota and Pennsylvania guidelines have been in effect, significant differences in how appellate courts review guideline cases have arisen. A comparison of the appellate review process in Minnesota and Pennsylvania concluded that the Minnesota appellate courts substantively reviewed sentences for the quality of the sentencing decision in reducing disparity. Pennsylvania's appellate courts, on the other hand, established a procedural appellate review approach that merely asks whether the trial court followed the proper procedures in applying the guidelines.30 In part, enforcement determines the presumptiveness of the guidelines. Voluntary guidelines and those without some explicit enforcement mechanisms, like the guidelines of Michigan and Virginia, are generally considered voluntary: judges are expected to consider them, but there is no appellate review if the judges do not conform to them. On the other hand, guidelines such as those of Pennsylvania, which have a relatively low standard under appellate review, are less presumptive than the Minnesota guidelines with the appellate standard of clear and convincing. Sentencing guidelines constrain judicial discretion to consider various factors and to determine the specific sentence. The jurisdictions having sentencing guidelines approach the standards to depart from the guidelines differently. Some have left that issue to the enabling legislation. Other commissions took the initiative and established their own standards. Minnesota's commission, for example, established the standard of clear and convincing evidence to depart from the guidelines. Similarly, Oregon's commission articulated a substantial and compelling standard for departures. North Carolina took a somewhat different approach. North Carolina's guidelines provide a standard range and a range of sentences for aggravating and mitigating circumstances. The guidelines provide for three types of penalties: active punishment, which is a sentence in the State prison system; intermediate punishment, including residential facilities, electronic house arrest, or intensive supervision; and community punishment, which is supervised or unsupervised probation that may include outpatient treatment, treatment alternatives to street crime, community service, restitution, or fines. If the guidelines call for an active prison sentence, the court may impose intermediate punishment, but only when it is found "that extraordinary mitigating factors of a kind significantly greater than the normal case exist and that they substantially outweigh any factors in aggravation."31 The effect of this approach in North Carolina is to draw a firmer line around the active prison sentence, increasing the likelihood that the guidelines will result in the incarceration of more serious violent offenders, as intended. Presumptiveness is important to a State's ability to project prison populations. Enforcement and presumptiveness of sentencing guidelines are important considerations for a legislature when creating a commission. The more presumptive the guidelines, the greater the confidence that the guidelines will control judicial discretion and achieve the goals of the legislation, including controlling prison populations. The more constraints imposed on judges by the guideline ranges, the more predictive are the State's prison populations. Moreover, the more presumptive the guidelines, the stronger the argument that mandatory penalties are not necessary. On the other hand, the more presumptive the guidelines, the higher the risk that the "invisible side" of criminal justice will become greater, with prosecutors bargaining on charges or prior records to avoid the commission's guideline range in favor of the local criminal justice culture's range.32 This issue is complex and lacks a clear-cut best policy. However, it is a policy issue that drafters of guideline-enabling legislation need to consider as they write the enabling legislation. Political and Legal Context Reform efforts build on existing political and legal structures. The more successful national efforts consider the political environment and grant their commissions the power and resources to fulfill their responsibilities. However, a State's statutory structure may not be conducive to building sentencing guidelines. With States consolidating their criminal statutes into broadly defined offenses over the past century, the imposition of sentencing guidelines exacerbates the problem of reducing disparity. Realizing this fact, Tennessee officials had their sentencing commission undertake the complex task of revising the criminal code. The criminal justice system brings together local and State agencies. The success of State guideline systems may well rest on the willingness of courts to implement the guidelines as intended.33 This process may seem simple: adopt guidelines, and they will be implemented. Foresight, planning, careful coordination to ensure the quality of sentencing guidelines, and consideration of those who must implement them are crucial to success. In fact, these factors should be evaluated and considered as early as possible in the drafting of the enabling legislation. Expectation of criminal justice reforms must not be too high. Eisenstein and colleagues conclude from their study of court communities in three States that "implementation of reform is not perfect."34 The court community resists change being imposed upon it. Guidelines are viewed as attempts to change the "going rates" for particular offenses. These guideline rates may not be consistent with the rates that have been developed in the local court. While local going rates are not always invariant, going rates provide rewards to the court community. Going rates fulfill a sense of justice, are efficient, and are effective. In fact, Eisenstein and colleagues conclude from their research that the "more radical a proposed change the less likely is its adoption."35 This conclusion is contradicted by the Spohn and Horney study of legal reforms of rape laws, which found that the more significant the reform, the more likely it is to affect court behavior.36 Research is contradictory, and each State must determine how best to approach the problem, depending on the relative autonomy of judges and the political context. Each State must, however, thoroughly consider the implementation of its guidelines if it is to meet its goals. Ongoing Monitoring of the Guidelines Resistance to reform parallels the resistance noted in the organizational literature, which suggests that participation in the development and implementation of guideline reforms by those who must implement the changes increases the likelihood of change. Sentencing reforms frequently demonstrate immediate changes, only to have practices quickly return to prereform levels.37 One reason for this phenomenon is that the initial attention and publicity surrounding a new law diminish, with the behavior of offender and system returning to prereform patterns. Another reason is that the changes are often mandated by a State agency. If the State agency fails to obtain the commitment to address the practical concerns of local officials who must implement the reform, then there is little chance of sustained effect. It should be emphasized that the participation of those who must implement the reforms is crucial; the implementation and monitoring are also necessary to sustain the reforms. Even with these systems in place, Minnesota's guideline system experienced backsliding (although not to preguideline levels) during the second and third years after implementation.38 Local Court Culture Although guideline systems are intended to structure discretion, local practitioners may undermine the implementation of the guidelines if they disagree with the concept of a guideline system. For example, trying to change the sentencing practices for offenses on the basis of the offense of conviction may result not in changes in sentencing, but in a shift in prosecutorial practices to "adjust" to the intended change, keeping the end result the same. An effective guideline system requires the enabling legislation and the guideline advocates to consider the local legal culture and how guidelines could fit into that culture. Defining the stakeholders and including them in the development of the guidelines facilitate acceptance. Their involvement is important in the early stages of development because the creators of the guidelines often take ownership of the project, serving as educators and marketers of the guidelines. According to a former executive director of the Washington State Guidelines Commission, serious thought should be given to maintaining the continued involvement of stakeholders and obtaining commitment from new stakeholders as positions turn over.39 Furthermore, if change requires financial investment by local government, resistance by county commissioners or local government officials should be anticipated. One way to obtain increased commitment from local courts is to include members of local criminal justice systems on the commission. Soliciting and carefully considering local input for the guideline writing process enhance the potential for guidelines to fulfill their goals. Broad-based support generated during the writing of the guidelines enhances a commission's ability to obtain acceptance and ensure smooth implementation. Political Acceptance The political acceptability of guidelines should also be considered when a sentencing commission is being created. The credibility of the commission and the acceptability of its guidelines are diminished when only minimal support exists for commission goals or when a commission is established solely to forestall other agendas.40 Little research has been done on the political context of sentencing guidelines.41 However, commissions have struggled to gain acceptance for guidelines where there is little commitment to values such as truth in sentencing, reducing discretion, conducting appellate review of sentences, or controlling prison populations. A commission, once created, can help generate support for itself; however, the experiences of New York and South Carolina--where commissions have been unsuccessful--indicate that this may not be adequate. A relatively new approach to politics and sentencing is the 1994 Oregon referendum on sentencing. In its 1994 election, Oregon passed referendums that significantly affect State sentencing guidelines. These referendums established mandatory minimum sentences for a wide range of offenses, including robbery, assault, murder, and rape. Staff at the Oregon Sentencing Guideline Commission have indicated that the mandatory sentence would apply to approximately 1,200 offenders. Previous sentencing practices under the guidelines sent 61 percent of the offenders to prison and passed alternative sentences on 39 percent. Mandatory sentencing provisions also increase the length of incarceration. For example, the average incarceration sentence for the type of robbery covered under the referendum is 57 months, but the new mandatory sentence will be 90 months. Furthermore, the ballot measure established that juveniles committing specific crimes will be remanded to the adult justice system. This measure will obviously have a significant impact on the courts as well as on the correctional system. Summary This chapter has reviewed the context and authority of the commission on sentencing guidelines. In general, successful commissions were created in supportive environments and were given strong legislative mandates. They were also provided with funding adequate to hire experienced staff and to meet regularly. These factors contributed to--although they did not ensure-- success. Legislatures have developed various approaches for designing and implementing their guidelines. The perceived problems of sentencing differ according to the perspective of the observer. Almost all States have adopted sentencing guidelines to impose some constraints on unfettered discretion; some States also wish to increase certainty and predictability. Other States aim to reduce prison overcrowding and increase truth in sentencing. Both State and local legal culture and the political environment are important to consider in the implementation of a State's guidelines. For example, sentencing guidelines may be politically overwhelming at one time, but as prison overcrowding reaches crisis levels or results in court-ordered constraints, their political acceptability may increase. At some point, the pressure of crowding and the financial cost of new constructions to accommodate rising prison populations may become overwhelming. One issue that should not be ignored is that the sentencing decision will be subjected to scrutiny, requiring a commitment to develop coherent sentencing policy. Research indicates that legislatures that have started this process have learned from it. However, with several States failing to implement sentencing guidelines, the pitfalls of the process become obvious, and a State may decide not to effect change. Nonetheless, in almost all jurisdictions, the sentencing commissions and their guidelines have flourished, despite diverse mandates and different approaches to writing sentencing guidelines. ------------------------------ Notes 1. Goodstein, L., J.H. Kramer, and L. Nuss. "Defining Determinancy: Components of the Sentencing Process Ensuring Equity and Release Certainty." Justice Quarterly 1(1)(1984), pp. 47- 73. 2. Singer, R.G. Just Deserts: Sentencing Based On Equality and Desert. Cambridge, MA: Ballinger. 1979. 3. von Hirsch, A. Doing Justice: The Choice of Punishments. New York: Hill & Wang. 1976. 4. von Hirsch, A. "Equality, Anisonomy, and Justice: A Review of Madness and the Criminal Law." Michigan Law Review 82(4)(1984), pp. 1093- 1112. 5. Note 4. 6. Bureau of Justice Statistics. National Corrections Reporting Program, 1990. Washington, DC: U.S. Department of Justice, Office of Justice Programs. 1993. 7. Martin, S. "The Politics of Sentencing Reform: Sentencing Guidelines in Pennsylvania and Minnesota." In Research and Sentencing: The Search for Reform, vol. 2, ed. A. Blumstein, J. Cohen, S. Martin, and M. Tonry. Washington, DC: National Academy Press. 1983. 8. Kramer, J.H., L. Lubitz, and A. Kempinen. "Sentencing Guidelines: A Quantitative Comparison of Sentencing Policy in Minnesota, Pennsylvania and Washington." Paper presented at the annual meeting of the American Society of Criminology, San Diego, November 1989. 9. However, it should also be noted that the Wisconsin commission has correctional representation but was not specifically charged to restrict prison populations. 10. Lubitz, R. "North Carolina Legislature Considers Sentencing Change." Overcrowded Times 4(2)(1993), pp. 1, 9-10. 11. Bogan, K.M. "Constructing Felony Sentencing Guidelines in an Already Crowded State: Oregon Breaks New Ground." Crime and Delinquency 36(4)(1990), pp. 467-487. 12. See entry for Arkansas in Table 4-2 (not available in this format). 13. See entry for Minnesota in Table 4-2 (not available in this format). 14. Pennsylvania Commission on Sentencing. Sentencing in Pennsylvania: Annual Report. State College, 1994. 15. Laws of Minnesota 1978, Ch. 723-S.F. No. 65-- Subdivision 5(2). 16. Note 11, p. 469. 17. Note 7. 18. Data in Chapter 6 comparing incarceration rates between Minnesota and Pennsylvania (since each has implemented sentencing guidelines) indicate that both States have significantly altered incarceration rates as intended, although other factors such as conviction numbers, decisions by others in the system, and the adoption of mandatory minimum sentencing provisions will influence the incarceration rate. 19. Griset, P. Determinate Sentencing: The Promise and the Reality of Retributive Justice. Albany: State University of New York Press. 1991. 20. Note 7. 21. Tonry, M. "Structured Sentencing." In Crime and Justice: A Review of the Research, vol. 10, ed. M. Tonry and N. Morris. Chicago: University of Chicago Press. 1988. 22. Misretta v. United States 488 U.S. 361 (1989) presents a Federal examination of the issue. 23. Del Sole, J. "Appellate Review in a Sentencing Guidelines Jurisdiction: The Pennsylvania Experience." Duquesne Law Review 31(1993), pp. 419-504. 24. Note 8. 25. Note 7. 26. Knapp, K. "Allocations of Discretion and Accountability Within Sentencing Structures." Colorado Law Review 64(1993), pp. 679-705. 27. American Bar Association. Standards for Criminal Justice--Sentencing Alternatives and Procedures. 3d ed. Boston: Little, Brown & Company. 1993, p. 5. 28. McCloskey, J. "The Effectiveness of Independent Sentencing Commission Guidelines: An Analysis of Appellate Court Decisions in Two Jurisdictions." Paper presented at the annual meeting of the American Society of Criminology, San Diego. 1985. 29. Note 23. 30. Note 23. See also Note 28. 31. North Carolina Sentencing and Policy Advisory Commission. Report of the 1993 Session for the General Assembly of North Carolina. Raleigh, 1993, p. 15. 32. Savelsberg, J. "Law That Does Not Fit Society: Sentencing Guidelines as a Neoclassical Reaction to the Dilemmas of Substantivized Law." American Journal of Sociology 97(s)(1992), pp. 1346-1381. 33. Local courts have been studied in detail. One of the overwhelming findings is the role played by informal norms in guiding local courts and the extent to which these norms vary across jurisdictions within a particular State. See also Notes 23 and 28. 34. Eisenstein, J., R.B. Flemming, and P.F. Nardulli. Contours of Justice: Communities and Their Courts. Boston: Little, Brown & Company. 1988, p. 296. See also Nardulli, P., R. Flemming, and J. Eisenstein. The Tenor of Justice: Criminal Courts and the Guilty Plea Process. Chicago: University of Illinois Press. 1988. 35. Note 34, p. 294. 36. Spohn, C., and J. Horney. Rape Law Reform: A Grassroots Revolution and Its Impact. New York: Plenum. 1992. 37. Ross, H.L. Deterring the Drunk Driver. Lexington, MA: Lexington Books. 1984. 38. Miethe, T.D., and C.A. Moore. Sentencing Guidelines: Their Effect in Minnesota. Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. April 1989. See also Stolzenberg, L., and S.J. D'Alessio. "Sentencing and Unwarranted Disparity: An Empirical Assessment of the Long-Term Impact of Sentencing Guidelines in Minnesota." Criminology 32(2)(1994), pp. 301-310. 39. Lieb, R. "Washington Prison Population Growth Out of Control." Overcrowded Times 4(1)(1993), pp. 1, 13-14, 20. 40. Note 7. 41. See Note 19 for exceptions to this. See also Kramer, J., and J. McCloskey. "The Role of Political Pressure in Developing Sentencing Guidelines for Drug Offenses." Paper presented at the 1988 annual meeting of the Academy of Criminal Justice Sciences, San Francisco. 1988. ------------------------------ Chapter 5 Writing Sentencing Guidelines A sentencing commission must immediately confront the difficult task of writing guidelines. Sentencing is a complex issue, involving the community, the courts, and the purpose(s) to be accomplished through the sentence. Many critics will question whether the sentence was appropriate, whether it can protect the public, or whether it can rehabilitate the offender. Despite the difficulty, many commissions have written, adopted, and implemented guidelines. This chapter describes this complex decisionmaking process. The writing of sentencing guidelines requires the sentencing commission to determine what factors should be considered for the guidelines; how the factors are to be measured in applying the guidelines so that they are reliably applied by the court; and which sentences can be applied on the basis of those factors. The descriptive approach looks at past sentencing practices, identifies the factors that judges have used in the past, and, after eliminating inappropriate factors such as race and gender, incorporates them into the guidelines. The in/out and sentence-length decisions are based primarily on the measurable factors deemed substantively important and statistically significant in past sentencing practices. In the prescriptive approach, the commission relies more on its own judgment in determining the factors for consideration in sentencing. In reality, sentencing commissions use data on past sentencing practices to guide their decisions even if they adopt the prescriptive approach. Past practices are generally used to determine the impact of the proposed changes and to test how the views of the commission members may compare with the actual sentencing practices of the judiciary. Although past practices may be used as a guidepost, they do not limit the commission's ability to change the guideline recommendation from previous sentencing practices. There are certain consistencies as well as key differences among the States that have written sentencing guidelines. States consistently identify the seriousness of the current offense and a defendant's prior record as the two major factors to be considered in sentencing. However, guideline systems vary considerably in incorporation of these factors. All States assess the seriousness of the conviction offense; however, Florida has separate guidelines for each general category, while Minnesota, Oregon, and Washington rank all offenses on a single scale of seriousness. The result is that Florida works with several different matrices, while Minnesota, Washington, and Oregon each work with one matrix. A commission's product reflects the process that it uses to create the guidelines. The more groups that are involved, the greater the likelihood of broad-based support. Offense Seriousness All States with sentencing guidelines employ offense seriousness as the primary factor to be considered in recommending sentences. However, some jurisdictions may measure an offense according to likelihood of relapse, as is often done in parole guidelines. If a sentencing guideline system adopted the prediction of recidivism as its major purpose, then ranking of the current offense would differ significantly from a measure targeting the most serious offenses. For example, murder is always the most serious offense when offenses are ranked by severity. However, when offenses are ranked by likelihood of relapse, murder would rank low because murderers have a relatively low risk of relapse. Theft offenders are more likely to relapse and therefore rank high on a risk measurement scale. Sentencing commissions have unanimously determined that offense rankings should rest on the seriousness of the offense and not on the risk of recidivism. One of the foundations for sentencing guidelines, then, is apparently focused on a retributive, or just deserts, model that determines penalties relative to the severity of the offense. It should be noted that some State statutes are very specific regarding bodily injury and use of weapons so the guidelines need no special enhancements to take these offenses into consideration. The table oversimplifies some guideline systems, such as that of Wisconsin, which has many different matrices and varies the number of criminal history categories depending on the offense matrix. The methods for ranking offense severity range from reliance on statutory grading to the use of social science scaling techniques. Most States begin with the offense classification used in statutes and then make finer distinctions within those grades. For example, Pennsylvania has three statutory grades for felonies and three statutory grades for misdemeanors. The commission made distinctions beyond the statutory classifications; the initial guidelines contained 10 categories measuring offense seriousness. Oregon and Pennsylvania subcategorized offenses so that the guideline measure of offense seriousness more closely reflected the severity of the crime.1 Articulated Offense Ranking Principles Some States developed principles to be used in measuring offense seriousness. In such States the sentencing commission begins the ranking process by identifying distinguishing factors such as the degree of injury to the victim, the amount of property loss, and the culpability of the offender. After the factors are ranked, specific offenses are attached to the rankings. The major advantage of this approach is that it allows for easy incorporation of newly created offenses into the rankings and provides a clear rationale to legislators, judges, and others as to the basis for the rankings. North Carolina, Louisiana, and Kansas have used this approach in ranking offenses. Number of Offense Levels The numbers of Offense Gravity Score levels generally range between 9 (North Carolina) and 15 (Washington). These numbers contrast with the United States Sentencing Commission's (USSC's) development of 43 categories. Commissions indicate that the number of categories is based on two goals: keeping the measure simple for implementation purposes and avoiding too few categories to allow for differentiation among the seriousness of offenses. In addition, States indicate that because the current measures for ranking offense seriousness still cluster offenses of different levels of seriousness, judges are encouraged to depart from the guidelines when a recommendation is inappropriate for a nontypical offender. Number of Matrices Another issue regarding offense measurement is whether a State adopts single or multiple matrices. Several States such as Florida, Wisconsin, and Kansas have multiple matrices. The Kansas guidelines contain one matrix for drug offenses and another for nondrug offenses. Kansas developed a separate matrix for drug offenses for a number of reasons: the complexity of distinguishing drug amounts and other problems perceived to be peculiar to drug convictions; the significant proportion (25 percent) of offenses represented by drug offenses; and the commission's desire to provide relatively harsh penalties for drug offenses. Wisconsin has developed 18 separate matrices so that the offense severity measure can be tailored to the type of offense. Multiple matrices allow the guidelines to incorporate additional severity factors, such as use of weapons, that are viewed as significant for a particular offense. A problem with the multiple matrix approach is that it may result in no comprehensive system of offense seriousness. For example, the dual matrices of the Kansas guidelines allow for disproportionality between drug and nondrug offenses. Other States have imitated the matrix approach developed by Minnesota in 1980 because of its simplicity. The matrix reads like a roadmap, with the offense severity measure along one side of the table and the prior record measure along the other. This approach has received criticism for oversimplifying the complexity of sentencing, however. A recent innovation that does not rely on matrices is the Swedish model. Sweden's criminal code did not establish numerical guidelines but set forth principles for the courts to consider in sentencing.2 Sentencing in this system consists of two phases. The court must first assess the seriousness of the offense on the basis of harmfulness (such as the defendant's culpability) and other offense-related factors. Then the court assesses whether the offense justifies imprisonment or a fine, based on the seriousness of the offense. If a fine is deemed inappropriate and imprisonment is required, the court determines whether probation is an appropriate alternative to imprisonment. Again, the criminal code guides this decision. Young offenders (under the age of 21) require special justification to be imprisoned. The court is expected to place on probation any offender who needs treatment. Thus, Sweden's criminal code favors a logical sequence of decisionmaking. This model could be applied to various U.S. sentencing options including probation, the range of intermediate penalties, county incarceration, boot camp, and State confinement. Comprehensiveness of Offense Coverage A significant issue in the measurement of offense severity is whether the measure integrates felonies and misdemeanors. Most States focus on providing guidelines for felonies; however, such focus may allow the court to sentence misdemeanor offenders more harshly than felony offenders, threatening the integrity of the guideline's attempt to achieve fairness and proportionality. However, the decision concerning the breadth of guideline coverage has important philosophical and practical implications. Because sentencing commissions are State agencies whose major focus is on State imprisonment, allowing total discretion for misdemeanors allows county jurisdictions greater freedom to consider local resources and set a local sentencing standard. Subclassification of Statutory Definitions One of the problems facing almost any commission attempting to rank offense seriousness is that of statutory definitions. Over the past century, statutes have tended to consolidate a broad range of criminal behaviors under one statutory definition.3 The court is thus provided with extensive discretion to impose the appropriate sentence for each defendant. The only limit on the court's authority for the worst case was the statutory maximum. Once a commission attaches presumptive guideline sentences to specific offenses, the breadth of the statutory definitions presents problems for the commission. For example, assigning a statutory offense a single rank on an offense seriousness scale can be a problem when the behaviors in any particular offense vary considerably. Guidelines can address this problem in one of four ways. USSC adjusts the rank of crimes on the basis of the offender's behavior. In the USSC approach, the offense of conviction is the starting point, but the final rank of the offense is adjusted significantly on the basis of an offender's conduct, regardless of whether the conduct was a part of the offense of conviction. This is referred to as "real offense" or "modified real offense" sentencing. The court can be allowed wider discretion to cover such circumstances. The difficulty with this second approach is that it relies on similarity of practices among judges to avoid significant sentencing disparity, reflected in either wider ranges in the guideline matrix (such as those in Pennsylvania's matrix) or wider discretion to depart for aggravating or mitigating circumstances. A third approach, used by Oregon and Pennsylvania, is referred to as the "subcategorization" of offenses to reflect different degrees of severity. For example, in Pennsylvania all burglaries historically have been classified as felony 1 offenses (the most serious statutory grading). The Pennsylvania sentencing commission drafted guidelines that subcategorize the single statutory burglary into four categories and then recommend different sentences on the basis of whether a residence was burgled and whether someone was present at the time of the burglary. It is significant that Pennsylvania included a factor that was part of the statutory definition (whether the structure was occupied) and added a real offense factor (type of structure) in its guidelines. The legislature later created a felony 2 burglary that parallels the least serious of the burglary subcategories (i.e., nonresidential burglary at which time no one was present). A fourth approach, adopted by Florida and Wisconsin, provides specific guidelines for various offense types. Florida's original guidelines established eight offense types and developed special guidelines for each. Although more complex, this system allows a commission to incorporate factors linked to particular offenses. Most guideline systems have rejected this model because of the problems of maintaining proportionality among offenses. Real Offense Versus Conviction Offense The sentencing decision generally begins with a focus on the conviction offense; this focus, however, may result in sentencing disparity, which guidelines aim to prevent. Many debate whether a guideline system appreciably diminishes disparity when it allows considerable latitude to the prosecutor to determine the guideline range based on charge bargaining.4 An alternative to conviction-based guidelines is real offense sentencing guidelines.5 No commission has embraced a real offense sentencing guideline system, perhaps primarily because of the appearance of unfairness. A fairness issue arises when sentences are based on a defendant's alleged conduct, which has a lower burden of proof than what is required to convict an offender.6 Another concern is that the role of plea bargaining in producing substantive justice may be undermined if defendants realize that the bargain will not influence the sentence.7 That is, plea bargaining carries perceived advantages. Guideline States have adopted an offense-of-conviction approach, presuming that this is the fairest system--although the difficulty presented by plea bargaining is recognized. USSC has addressed the issue of plea bargaining by beginning with the offense of conviction and then adding numerous factors for the court to consider as enhancements (such as degree of injury to victims, weapon use, and role in the offense) and as mitigating factors (such as substantial assistance to authorities). Criminal History Purpose Criminal history, or prior record, is the second major consideration in determining guideline sentences, the propriety of which has been hotly debated. Those who espouse a just desert or retributive philosophy argue that prior record should play a very limited role or no role in sentencing.8 However, all guideline systems include considerations of criminal history on the widely accepted premise that prior criminal activity demonstrates increased culpability and perhaps indicates future criminality. One reason to consider prior record in sentencing is to ensure awareness of an offender's increased culpability: A previously convicted offender who commits another crime has clearly been informed that criminal behavior is wrong but has relapsed despite such a warning. Another reason for considering prior record is risk assessment, or to predict future criminality. Sentencing commissions have not stated such a purpose of considering prior convictions; however, sentencing guidelines in the State of Washington weigh offenses differently depending on the current offense. For example, prior convictions for burglary count three points each against a current burglary offense, while prior convictions for murder count only two points. Measurement of Criminal History Data indicate that most States measure both the number and seriousness of prior convictions and that many States such as Minnesota consider additional factors such as an offender's status (e.g., being on probation) at the time of the current offense. The total number of prior record score categories varies among the States (e.g., Washington has nine categories and Tennessee has only five). The measurement of criminal history for purposes of guideline application raises many important issues. Among the most significant issues are how refined such a categorization should be and whether such a measure creates a typology of offenders. Some States (Minnesota, Washington, and Pennsylvania) have numerical scores that measure the number and seriousness of prior convictions. Most States with numerical prior record scores weight prior convictions as Pennsylvania does, with prior felony offenses assigned one, two, or three points depending on severity. Prior misdemeanors may count no more than two points in the total prior record score. Other guideline States (Oregon, Michigan, and Kansas) use descriptive categories that rely less on numerical scores or calculations. Instead, criminal history categories differentiate types of offenders such as those convicted of violent crimes and those convicted of multiple felonies. Such labels are intended to convey to the court the type of offender. One advantage of this approach is that the prior record categories contain a more homogeneous group of offenders in each criminal history category, increasing the likelihood that guidelines built on such measures can better reduce unwarranted disparity. For example, in States such as Minnesota and Pennsylvania, offenders may be violent or nonviolent and yet receive the same prior record score. USSC takes a unique approach to scoring and weighing prior convictions. The Federal guidelines incorporate the severity and length of prior sentences into the criminal history score. An offender who received a prison term for a specific conviction would have a higher criminal history score than an offender who received probation for the same offense. The question of perpetuating past disparity is then raised. Juvenile Adjudications One matter concerning measurements of prior record scores is whether to consider juvenile adjudications and, if so, to what extent. Some would argue that an offender entering adulthood should have a clean slate and that juvenile adjudications should not count. This position rests in part on the more informal process of the juvenile court and its emphasis on a juvenile's welfare and treatment. Moreover, it is argued that the juvenile court uses a lower standard for a finding of guilt than does the adult court. It has also been argued that juvenile records are less reliable than adult records across counties and that this inconsistency results in disparity in guideline systems relying on juvenile records. Some jurisdictions such as Pennsylvania, however, have noted an improvement in juvenile recordkeeping as a result of the incorporation of juvenile adjudications into their guidelines. Most States include juvenile adjudications in the measurement of prior records. An individual with a prior record has increased culpability, which is a justification used for including prior adult convictions. Juvenile records might also be important as assessments of offender risk of future criminality, although no State visited by researchers argued that guidelines created the prior record score as a predictive instrument. Lapsing of Prior Convictions Another issue raised in measuring criminal history is whether to lapse or expunge records of prior convictions after an offender remains crime free for some period of time. Several States, such as Minnesota and Washington, omit prior convictions after a specified number of years, whereas other States, such as Pennsylvania and Oregon, do not. Two basic arguments have been offered in favor of lapsing prior records. First, a long period of crime-free behavior may indicate diminished culpability/blameworthiness. Second, a long crime-free period may indicate that an old prior record has less value in predicting future criminality. Others, such as members of the Oregon and Pennsylvania sentencing commissions, argue that although the age at the time of a prior offense may be a mitigating factor, the principle of truth in sentencing demands that judges have information on all prior convictions at the time of sentencing. States that include juvenile adjudications in their guidelines almost always have a lapsing provision. For example, Minnesota lapses adjudications at age 21, Florida at age 23, and Maryland at age 25. Kansas lapses juveniles adjudications at age 25, except for serious personal offenses. Prior Record and Disparity Sentencing disparity is one significant issue for sentencing guideline systems. Disparity control requires that the guidelines be correctly applied to ensure that those with similar prior records are treated similarly. It is important to establish a policy to address prior convictions from other jurisdictions. Most States have required the courts to translate previous convictions into current codes. If the appropriate classification (e.g., whether the offense would be classified as a felony or misdemeanor) is ambiguous, then the court gives the defendant the benefit of the doubt and applies the least serious classification. Another disparity issue concerns the availability of records, which may be a particularly serious concern with the USSC standard of relying on previous sentences. Historically, State recordkeeping systems for dispositions and sentences have been inadequate. The more complex the measure of prior record, the greater is the risk of disparity. Defining Previous Conviction The definition of prior convictions affects which convictions are counted in the prior record calculations and thus affects sentence recommendations. At one extreme, the definition of previous conviction could require that commission of the offense, conviction for the offense, and sentence for the conviction all occur before the commission of the current offense. At the other extreme, a definition could include all offenses for which a defendant has been convicted, regardless of whether conviction and sentencing occurred before the current offense. These alternatives can produce very different prior record scores and dramatically different sentence recommendations. In some States, the definition of prior conviction has been subject to appellate review. In Minnesota the State supreme court has established that if a defendant is sentenced on the same day for three offenses, then the first two count in the prior record against the third offense. The role of criminal history in sentencing is thus inflated. In addition, a policy assuming that the commission of prior offenses increases the culpability of an offender--regardless of whether the offenses are committed before or after the current offense--is established. Other Factors Guideline systems across the country incorporate many factors in addition to the seriousness of the current offense and the prior convictions of a defendant. These factors include the possession/use of a weapon, the degree of physical injury to the victim, and whether a drug trafficking offense occurred near a school. If State statutes make such distinctions, sentencing commissions generally rely on the statute. However, some States historically have left such distinctions to the discretion of the courts, and commissions have incorporated them into guidelines. For example, the Pennsylvania and U.S. sentencing guidelines incorporate numerous factors not included in statutes but thought by commissions to be necessary for establishment of fair guidelines. The USSC guidelines incorporate factors that are not a part of the conviction offense but that the Commission thinks should be considered in sentencing. Data indicate that the USSC incorporates more factors than any other commission attempts to systematically include in its guidelines. Furthermore, the USSC considers these factors in greater detail than do the States. For example, in assessing bodily injury for aggravated assault with intent to commit murder, USSC breaks down the categories into (1) bodily injury, (2) serious bodily injury, and (3) permanent or life-threatening bodily injury. The base offense level receives 2, 4, or 6 penalty points depending on the extent of injury allegedly caused by the offender in commission of the crime. These enhancements to statutory factors reflect the intent of USSC to incorporate real offense sentencing factors into the guidelines. Minnesota chose not to create such enhancements, however, in keeping with its decision to focus on conviction behavior. In general, detailed statutory definitions require commissions to deal with fewer proportionality concerns through enhancements and real offense sentencing. Developing Sentencing Recommendations In establishing sentencing guideline ranges, commissions must consider many difficult topics, including the following: o The role of past sentencing practices in establishing the guideline sentence recommendations. o The amount of judicial discretion to be left to the court. o The role of guidelines in affecting correctional capacity. o The decision whether to link guideline sentence recommendations to mandatory penalties. o The relative role of enhancements to guideline ranges. o The extent to which the commission wants to specify nonincarceration options. State decisions have varied widely, depending on State confidence in the judiciary and in the ability of the guidelines to capture the appropriate sentencing standards for the typical case. Commissions can also be influenced by the political environment. This report does not allow for indepth exploration of each of these issues. However, the reader should be somewhat familiar with how various commissions address these issues. Descriptive Versus Prescriptive Guidelines As mentioned previously, the focus of the descriptive approach to guideline development is to reduce sentencing disparity by narrowing the range of variance of similarly defined cases under sentencing guidelines. This approach establishes a guideline that accounts for the variables in guideline construction and in the general sentencing practices historically used by sentencing judges. Most States, however, have rejected a purely descriptive approach because past practices do not necessarily reflect appropriate sentences. Furthermore, changes in resource availability (e.g., correctional capacity) and changes in societal definitions of the seriousness of various crimes necessitate reevaluation of past criteria. Nevertheless, all commissions use past practices in considering whether to change practices through sentencing guidelines. Most States have adopted more prescriptive models of sentencing. This approach encourages sentencing commissions to consider, aside from available correctional capacity, the roles of deterrence, incapacitation, rehabilitation, and retribution in the recommended guideline ranges. The prescriptive model leaves more decisions for the commission and is necessary for States that want to use the guidelines to help slow the growth of prison populations. Researcher interviews with State commissions indicate that past practice is one of the building blocks of all sentencing guidelines. Examining past sentencing practices provides the commission with (1) a benchmark against which to test the sentences being developed; (2) a measure to assess the potential impact of guidelines on sentencing and thus on correctional resources; and (3) an idea of the extent of disparity in sentencing. Past practices represent a very important foundation for the drafting of sentencing guidelines. Determining Sentence Lengths Historically, sentencing has been viewed as involving two decisions: should an offender be incarcerated and, if so, for how long? The first States adopting sentencing guidelines focused on these two decisions. Minnesota's guidelines, which apply only to felonies, focused almost exclusively on the State imprisonment decision and the length of such imprisonment. For offenders not recommended for State prison, the guidelines provided little guidance concerning appropriate lengths for county jail sentences or for community-based alternatives. In recent years many innovations in sentencing guidelines have been developed. The prison-crowding problem at both State and local levels pressured State sentencing commissions to search for means to encourage alternatives to incarceration. This interest was boosted by the book, Between Prison and Probation.9 The result was a new concept of alternatives to incarceration that were more than merely mechanisms for relieving prison crowding. These alternatives were also important in providing an intermediate range of penalties between prison and probation. Morris and Tonry's book significantly changed the concept of alternative sentencing programs, allowing people who write guidelines to more closely link the relatively expansive measures of offense seriousness and prior record with a range of penalties commensurate with basic guideline factors. Oregon and Louisiana developed alternatives that involved more punishment and restriction of liberty than did probation but that were more cost-efficient than prison or jail cells. Furthermore, these alternatives promised greater opportunities for offenders to contribute to their support and rehabilitation. Louisiana and Oregon mechanisms equate the amount of time needed in alternative programs with a period of incarceration. Data illustrate the concept of a penalty structure that ranges from probation to State incarceration, as developed by Louisiana's sentencing commission. One of the important roles for guidelines in the move toward more frequent use of nonincarcerative sanctions is their ability to identify offenders who are appropriate for such sanctions. Oregon and Louisiana have expressed all sentence recommendations apart from State prisons in terms of custody units or sanction units, and courts can select from a number of sentencing options such as jail, traditional probation, house arrest, community service, work release, or custodial treatment. Similarly, Washington's sentencing guidelines now express sentence recommendations apart from State prisons in terms of confinement time, but courts may decide the type of confinement or nonconfinement programs in which time is to be served. Having guidelines identify eligible offenders assures the public that such programs target appropriate offenders (such as nonviolent offenders) and those with the least serious prior records. Moreover, having guidelines target offenders for intermediate punishment sanctions reduces the risk that these programs will result in what is called net widening in which intermediate punishments target persons who would otherwise receive short jail sentences, with guidelines targeting such offenders for intermediate punishment. Width of Guideline Ranges Many questions have been raised concerning how much discretion guideline recommendations should allow to judges. Some States, such as Minnesota, provide very narrow ranges. Others, such as Pennsylvania, provide wide ranges. Most jurisdictions, however, have established ranges wider than Minnesota's but narrower than Pennsylvania's. There are both practical and philosophical concerns related to wide-ranging judicial discretion.10 First, wide ranges allow disparate sentences for similar offenders. For example, the Pennsylvania guideline for an offender convicted of rape (without prior record) would allow the judge to impose a minimum sentence ranging from 3 1/2 to 5 years. Similarly situated offenders can receive dissimilar sentences that still fall within the guidelines. Second, wide-ranging discretion makes it difficult to project how judges will sentence within the range. If judges impose sentences from the upper part of the range, the impact will likely be underestimated. Third, wide ranges allow consideration of factors that may be inappropriate in the sentencing process and may violate the intent of the guidelines. For example, wide ranges allow for county variations that contradict the guideline's intent to establish a statewide sentencing policy. Also, wide ranges allow for disparate sentences on the basis of factors such as employment, age, race, and gender. Link of Guideline Recommendations to Mandatory Penalties An issue of growing concern for sentencing commissions is whether guidelines should be commensurate with mandatory sentencing penalties adopted by legislatures. The question is: Should the commission take the prescriptive mandatory penalties as its benchmark and attempt to have a comprehensive and proportional system? The USSC guidelines adopted such a policy for drug offenses; this policy is perhaps the main reason for the significant impact of the Federal guidelines on the size of correctional populations. Other guideline systems, such as the one in Pennsylvania, view the guidelines independently of the mandatory statutes and have not attempted the linkage. The Pennsylvania commission deemed it necessary to act as an independent agency, with the guidelines reflecting differences not included in the mandatory statutes. This approach produced guidelines that were harsher than the mandatory statutes for some offenses but more lenient for others. Some commissions therefore reject the argument that guidelines should be commensurate and proportional to mandatory statutes because guidelines focus on different purposes (e.g., fairness) and provide for greater specification based on more factors. Furthermore, guidelines are written for the typical offender, whereas mandatory minimum penalties often address the worst cases. Multiple Convictions A significant proportion of sentences involve multiple convictions, but few States provide guidance on whether such sentences should be concurrent or consecutive. Most commissions leave this decision to the courts without any guidance. One notable exception, however, is the policy drafted by the Minnesota sentencing commission. Minnesota policy allowed for the imposition of consecutive sentences only under the following circumstances: (1) when at least one of the current offenses is a crime against the person, the sentence imposed is within the guidelines, and the sentence from a previous offense has not expired; (2) when the offender is convicted of multiple felonies against the person and is sentenced within the guidelines for the most serious offense; or (3) when the current conviction is for escape. In all other circumstances, imposition of consecutive sentences would be considered a departure requiring written justification. This policy provides separate retributive punishment in cases involving victims of serious violent crimes and provides for the severity of the sanction to be proportional to the severity of the offense. The Canadian sentencing commission suggested a different approach, recommending that discretion be left to the court regarding the imposition of concurrent or consecutive sentences for multiple convictions.11 However, the commission suggested a policy to limit the total period of confinement for offenders convicted of multiple transactions on the basis that the total sentence should be limited to prevent life sentences from being imposed for less serious offenses. The Kansas sentencing commission is one of the few to establish a policy governing the length of confinement for a "current conviction event." Kansas policy states that the total sentence cannot exceed twice the sentence recommended in the guidelines. In general, commissions have left this sentencing issue to the discretion of the court. In view of the significant number of cases involving multiple convictions, however, such a policy allows considerable discretion and the potential for considerable disparity. Sentencing Event One of the issues to be addressed by a sentencing commission is how to define a sentencing event in the guidelines. North Carolina defines a sentencing event as all sentences imposed on a defendant within a session of the court, generally within 1 week. Pennsylvania defines a sentencing event in terms of criminal transactions, on the basis of when the offenses occurred; for each separate transaction, a separate guideline form is required. Separate transactions occur when law-abiding behavior is exhibited between offenses; otherwise, the offenses are considered part of the same transaction. The definition of a sentencing event influences how information is collected and used. Data systems using criminal transactions must be corrected to make prison impact statements. In other words, prison impact assessments require the identification of how many individuals are being sent to prison and for how long. Aggravating and Mitigating Circumstances In contrast to mandatory minimum sentencing, one of the advantages of a guideline sentencing system is that it provides recommendations for typical cases and allows judges to depart from these recommendations when extenuating circumstances exist. When judges depart from guideline recommendations, however, they are required to justify their sentences. As indicated in data, almost all of the guideline States list aggravating and mitigating factors deemed appropriate for judges to consider in departing from presumed sentence recommendations. Even States such as Arkansas, where the guidelines are not subject to appellate review, have such a list. It should be noted that these lists are nonexclusive, leaving discretion to the court for consideration of other factors warranting departure. Incorporating a list of nonexclusive aggravating or mitigating factors helps provide some direction to the court and is intended to provide more consistency across the State for sentences that depart from the guidelines. This approach falls in line with the goal of providing fairer and more equitable sentencing, which is one of the reasons for States to implement guidelines. Furthermore, the lists are useful for assisting commissions if they revise the guidelines later. For example, the reasons can help determine why some offenses have high departure rates and how better recommendations can be made. The decision to include such a list, however, can become controversial for commissions. Pennsylvania is one of the few States that does not provide reasons for departure; this is particularly interesting in light of their enabling legislation requiring the guidelines to provide aggravated and mitigated sentence recommendations. The original set of proposed 1981 Pennsylvania guidelines included such a list, although the commission eventually decided to delete it. The decision to delay incorporation of aggravating and mitigating factors into the guidelines was based on the need for data to be used to develop such reasons. The Pennsylvania commission has revisited the issue of whether to incorporate a list into its guidelines. Aside from providing for more consistency among judges, a statewide survey had indicated that people who used the guidelines would appreciate having a list. Furthermore, the reasons for departure often received by the commission were considered inadequate. However, the commission from judges eventually decided not to include a list of reasons for departure because it would be a crutch for judges and unnecessary. There was also concern that it would result in more appellate litigation. Pennsylvania's stance is interesting in view of recent studies of race and gender, which find discrimination in departures from the guidelines.12 States that provide reasons for departure have neither indicated problems with such a decision nor advised against developing such a list. In the development of a list of appropriate departure reasons, caution should be exercised so that a balance is maintained between the number of aggravating and mitigating reasons and the guidelines do not appear to favor defense or prosecution. The number of formal reasons for departure varies by State, with most States having 10 or fewer reasons for both aggravation and mitigation. The number of reasons ranges from 4 (aggravating reasons in Kansas; mitigating reasons in Minnesota) to 31 (aggravating reasons in North Carolina). Washington reports more than 50 reasons cited by judges in mitigating guideline sentences. The type of factors allowed also varies by State. Minnesota has developed a list of reasons considered inappropriate for departure, including race, gender, and employment status of the defendant. However, other States allow reasons of employment and other socioeconomic factors to be used to justify a departure sentence. For example, North Carolina allows the following factors to be considered in making a departure: o Defendant supports her or his family. o Defendant has a positive employment history or is gainfully employed. o Defendant has a good treatment prognosis and a workable treatment plan. o Defendant has a support system in the community. Similarly, among the reasons allowed by the Washington sentencing commission as possible justifications for departures are the following: o Defendant's rehabilitation or treatment. o Defendant's remorse. o Defendant is employed or in school, or has had a commendable employment record or military service. o Defendant's age. o Defendant's assistance to law enforcement or in prosecution of codefendant. Clearly, how extensively these subjective factors are used to depart from the guidelines will have much to do with meeting the goal of reducing unwarranted disparity. Summary The drafting of sentencing guidelines is a complex and lengthy process. How the guidelines are written bears directly on sentencing disparity, the use of incarceration for various offenders, and prison crowding. The complexity results from balancing diverse and conflicting values with the knowledge obtained through social science research concerning sentencing decisions and their effectiveness in deterring, incapacitating, punishing, and treating offenders. Combined with limited correctional resources, this complex process of drafting sentencing guidelines creates a tremendous challenge for sentencing commissions. ------------------------------ Notes 1. Blumstein, A., J. Cohen, S.E. Martin, and M.H. Tonry (eds). Research on Sentencing: The Search for Reform. Washington, DC: National Academy Press. 1983. See also Bogan, K.M. "Constructing Felony Sentencing Guidelines in an Already Crowded State: Oregon Breaks New Ground." Crime and Delinquency 36(4)(1990), pp. 467-487. 2. von Hirsch, A. "Principles for Choosing Sanctions: Sweden's Proposed Sentencing Statute." New England Journal on Criminal and Civil Confinement 13(2)(1987), pp. 171-195. See also Ashworth, A.J. "Sentencing Reform Structures." In Crime and Justice: A Review of the Research, ed. M. Tonry, pp. 188-244. Chicago: University of Chicago Press. 1992. 3. Singer, R.G. Just Deserts: Sentencing Based On Equality and Desert. Cambridge, MA: Ballinger. 1979. 4. Alschuler, A.W. "Sentencing Reform and Prosecutional Power: A Critique of Recent Proposals for `Fixed' and Presumptive Sentencing." University of Pennsylvania Law Review 126(1978), pp. 550-577. See also Mistretta v. United States 448 U.S. 361 (1989) for a Federal examination of the issue. See also Savelsberg, J. "Law That Does Not Fit Society: Sentencing Guidelines as a Neoclassical Reaction to the Dilemmas of Substantivized Law." American Journal of Sociology 97(s)(1992), pp. 1346-1381. 5. Robinson, P.H. "Hybrid Principles for the Distribution of Criminal Sanctions." Northwestern Law Review 82(1987), pp. 19-42. See also Note 6. 6. Reitz, K.R. "Sentencing Facts: Travesties of Real-Offense Sentencing." Stanford Law Review 45(1993), pp. 523-573. 7. Eisenstein, J., and H. Jacob. Felony Justice: An Organizational Approach to Criminal Courts. Boston: Little, Brown & Company. 1977. See also Nardulli, P., R. Flemming, and J. Eisenstein. The Tenor of Justice: Criminal Courts and the Guilty Plea Process. Chicago: University of Illinois Press. 1988. 8. von Hirsch, A. Doing Justice: The Choice of Punishments. New York: Hill & Wang. 1976. See also Note 3. 9. Morris, N., and M. Tonry. Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System. Oxford: Oxford University Press. 1990. 10. Martin, S. "Interests and Politics in Sentencing Reform: The Development of Sentencing Guidelines in Pennsylvania and Minnesota." Villanova Law Review 29(1984), pp. 21-113. 11. Canadian Sentencing Commission. "Sentencing Reform: A Canadian Approach." Canada: Minister of Supply and Services. 1987. 12. Kramer, J.H., and D. Steffensmeier. "Race and Imprisonment Decisions." The Sociological Quarterly 34(2)(1993), pp. 357-376. See also Steffensmeier, D., J. Kramer, and C. Streifel. "Gender and Imprisonment Decisions." Criminology 31(3)(1993), pp. 411-446. ------------------------------ Chapter 6 The Impact of Sentencing Guidelines Structured sentencing reforms are designed to change how the criminal justice system operates. For example, sentencing guidelines are expected to reduce unwarranted disparity at the sentencing phase. Specifically, both the disposition decision (prison versus probation or other community sanctions) and duration (length of sentence or time to be served in prison or in the community) will become more predictable and equitable. Furthermore, some States have implemented guidelines to minimize the potential for prison crowding by requiring the imposition of imprisonment sanctions that can be managed in a more diligent manner. States that have adopted presumptive sentencing guideline schemes might look different on a number of key outcome measures, such as sentencing disparity, use of incarceration, and prison crowding, compared with States that have not adopted such reforms. This chapter reviews published studies and data that examine how extensively sentencing guidelines have affected these and other key issues. Methodological Issues in Assessing Impact Determining the impact of sentencing guidelines is complicated by a number of methodological issues. For example, measuring disparity and its sources can be very complex. Prison crowding and incarceration rates are the result not only of sentencing decisions but also of law enforcement, prosecutorial, and budget policies that are not controlled by the courts or sentencing guidelines. This section attempts to isolate the major factors that confound assessment of the impact of sentencing guidelines. The criminal justice community lacks consensus on the purpose and goals of structured sentencing. Consequently, it is not clear whether all States with guidelines should show effects in all of the dimensions listed above. Minnesota, Washington, North Carolina, and Oregon have stated that guidelines are to be used to regulate the use of prison space and to help avoid prison crowding. Because the structure of guidelines varies among the States, all sentencing guideline States cannot be held equally accountable for affecting a particular outcome measure. In assessing disparity in departure rates, a State with a rather broad range can succeed more easily than a State with a rather narrow guideline range. There is also a lack of stability in the structure and content of sentencing guidelines over time. Each year, many States modify their guidelines. These modifications may diminish or enhance a State's ability to achieve success in a particular area. For example, there have been several substantial modifications to the original Minnesota, Washington, and Pennsylvania sentencing guidelines. Such modifications in sentencing policy make it difficult to conduct meaningful time series analyses. The objective of establishing control States, or States that have not adopted structured sentencing, whose results can be compared with those of the structured sentencing States is not a clear-cut task. Many States not defined as true guideline States include varying degrees of structured sentencing, such as determinate and mandatory minimum sentencing provisions. Some nonguideline jurisdictions have eliminated the use of discretionary parole release, instituting many mandatory minimum sentencing provisions. In assessing the impact of guidelines on sentencing disparity, researchers relied exclusively on a review of the major studies published by independent researchers or sentencing commission staffs. To examine trends in the use of incarceration, prison crowding, and public safety, historical trend analyses were conducted for the Federal guideline system and the four States (Pennsylvania, Minnesota, Florida, and Washington) that have had guidelines in place for the longest time. These statistical trends were compared with nonguideline States over the past two decades. In selecting States for comparison with the four guideline States, the following methodology was used. For any outcome measure, trend data on the four guideline States and all the nonguideline States were analyzed for the years preceding the adoption of guidelines. Nonguideline States with similar historical trends on these outcome measures were selected as comparison States. For example, in assessing incarceration rate trends for Minnesota, Minnesota's incarceration rate from 1971 to 1980 was analyzed. States having similar incarceration rates over that time were used as comparison States. This selection process minimized the error of lumping all nonguideline States into a single group. Impact on Sentencing Disparity Although jurisdictions may vary in their reasons for initiating a sentencing guideline system, almost all might argue that a primary goal of sentencing guidelines is to reduce unwarranted disparity in sentencing practices. Every sentencing commission has claimed that its guidelines have achieved some reduction in sentencing disparities. In evaluating disparity reduction, the preferred research design would consist of a simple pre- and posttest comparison of two samples of offenders. The first sample represents offenders who were sentenced under the preguideline sentencing structure, with the second sample being similarly situated offenders who were sentenced under the newly instituted guidelines. Statistical comparisons between the two samples are made on comparable offense categories (as well as other relevant factors such as prior criminal record and victim attributes) to ensure that both samples are statistically equivalent. Analyses can then be made to determine whether the imposition of sentences has become more standardized for the guideline cases than was the case for preguideline samples (i.e., less variance in case disposition and sentence length). Although this design appears straightforward, several problems arise in its execution. A review of a number of the limited studies on sentencing disparity has pointed out three major methodological problems.1 Sentencing practices may have changed so drastically since the dissemination of guidelines that similarly labeled offenses may not mirror each other before and after the use of guidelines. For example, to circumvent the guidelines through plea bargaining, less serious second-degree aggravated assaults could result in less serious charges and residual convictions for these crimes, resulting in a more homogenous pool than existed before guidelines were enacted. This increase in homogeneity will give the appearance of increased uniformity in sentencing practices. However, the guideline assault cases will not truly represent the preguideline assault cases. In addition, because it is not possible to use an experimental design in which a pool of offenders is randomly assigned to either a guideline or a nonguideline system, reductions in disparity that would have occurred independently of the passage of guidelines cannot be controlled for. This is especially likely with the growing popularity of numerous legislative actions such as mandatory terms that require offenders convicted of specific crimes to be imprisoned and spend a specific amount of time incarcerated. A great deal of determinacy may already have been achieved before the guidelines were adopted. The studies also tend to rely on simplistic comparisons of crime and prior criminal history categories now used by the guidelines for both samples (before guidelines and under guidelines), rather than a more sophisticated analysis of sentencing factors that were relevant before the preguideline cohort. For example, an offender's employment status and age may have been relevant under an indeterminant/rehabilitative sentencing model, but now are excluded under guidelines. Excluding these factors from the comparative analyses almost ensures that the preguideline sample will appear more disparate than the guideline sample. Finally, few sophisticated and independent evaluations have been completed, with the exception of the Minnesota and the U.S. guidelines. Independent evaluations on disparity reduction have not yet been published for Florida, Oregon, Pennsylvania, Washington, and Delaware. These States have relied instead on their own analyses to conclude that disparity has been reduced. In some instances, the claim is based on a simple analysis of whether the court is complying with the current guidelines rather than on changes in sentencing practices based on a preguideline cohort. United States Sentencing Guidelines One of the most detailed and sophisticated studies on disparity was released by USSC in 1991. It examines the short-term impacts of the Federal guidelines system on unwarranted disparity in sentencing, use of incarceration, and prosecutorial discretion and plea bargaining.2 Using a variety of analytic techniques, the Commission concluded that Federal guidelines have reduced disparity and were generally well accepted by a majority of the judges, prosecutors, and probation officers. Only Federal public defenders and private attorneys believed otherwise. Nonetheless, the study has been the object of controversy, with the soundness of the Commission's methodology and its conclusions being questioned. The USSC study attempted to answer the following question: Does the range of sentences meted out for defendants with similar criminal records convicted of similar criminal conduct narrow as a result of guideline implementation?3 To answer this question, eight categories of offenses representing a cross-section of Federal crimes were selected for quantitative analyses of disparity reduction: four categories of robbery (with or without weapon, combined with no criminal history or moderate criminal history); two categories of embezzlement ($10,000-$20,000 loss and $20,000-$40,000 loss); heroin trafficking; and cocaine trafficking. Two samples representing pre- and postguideline cases were drawn. In assessing disparity, two measures of sentence lengths were examined: actual sentences imposed by the court and time served (or to be served). Actual time served was not always attainable because some offenders had not served their entire sentences. In those cases, length of stay was estimated by using the presumptive parole date for preguidelines cases and the sentence duration minus the maximum amount of credit for good behavior for guidelines cases. Researchers found that the range of sentences, as indicated by the middle 80 percent of cases, had narrowed for all eight crime categories, although only three were statistically significant. These results generally demonstrated that under the Federal guidelines, disparity had been reduced and that for some crimes, the reduction was substantial. However, a number of methodological issues limit the strength of these findings. Because USSC was attempting to ensure that pre- and post-guideline offender groups were properly matched on all relevant dimensions, the sample sizes for the disparity analyses were quite small. Although the total number of sample cases was approximately 6,000, the sample sizes for offense categories ranged from 13 to 44 cases (preguidelines) and 24 to 81 cases (postguidelines). Cases in which the judges departed from the guidelines because the offender provided substantial assistance to the Government were excluded from the analysis. In 1991 approximately 12 percent of all disposed cases resulted in "downward departures for substantial assistance;" the same type of departures reached 21 percent for drug offense cases, while the overall departure rate was 19 percent.4 Excluding these cases from the analysis understates the level of disparity, although the sentences are no longer considered guidelines. The fact that only three of the offense categories showed statistically significant reductions in disparity suggests that preguideline cases were already exhibiting a relatively high degree of uniformity in court disposition. Further declines in sentencing discrepancy might have occurred independently of the introduction of guidelines and may be the product of chance. The USSC survey of court officials found that while Federal prosecutors and probation officers believed disparity had decreased (51 and 52 percent, respectively), a majority of judges (56 percent), Federal defenders (68 percent), and private attorneys (50 percent) reported that disparity either increased or remained about the same.5 As noted by Senior Circuit Judge Gerald Heaney, assessing disparity reduction becomes more complicated because the decision as to whether a case is handled by either a State or Federal court was not part of the study.6 Heaney notes that local law enforcement agencies can file cases in Federal court for certain crimes (especially drug crimes) because convicted offenders are likely to receive longer sentences in the Federal courts. This discretionary local jurisdictional filing decision has not been analyzed by any of the studies of the Federal guidelines, yet it can represent a major form of disparity in criminal justice decisionmaking. USSC recognized that its study was preliminary. At the time of the study, guidelines had been in effect for only 2 1/2 years. Consequently, when the postguideline implementation cases were selected, only 43-75 percent of Federal offenders were sentenced according to the guidelines.7 The second major study of the Federal guidelines was conducted by the U.S. General Accounting Office (GAO), whose research consisted of reanalyzing the USSC data. Although employing different statistical methods, GAO agreed with USSC that preliminary evidence indicated a reduction in disparity. However, data limitations prevented a determination of whether overall sentencing disparity had been reduced. GAO was especially interested in whether the influence of demographic factors such as race, gender, and education had been diminished. Data limitations prevented this issue from being fully assessed: Congress was particularly interested in reducing or eliminating disparity caused by demographic factors such as an offender's race, gender, and education. However, limitations and inconsistencies in the data available for preguidelines and guidelines offenders made it impossible to determine how effective the sentencing guidelines have been in reducing overall sentencing disparity.8 The GAO analysis attempted to group the sentence differences into parts on the basis of discrimination and legitimate factors. To do this, GAO devised two regression models: a constrained model, reflecting sentences under the guidelines; and an unconstrained model, using only statistically significant variables available to sentencing judges and omitted from the guideline scoring system. Under the constrained model, 25 percent of the racial gap in sentences among bank robbers was caused by discrimination, whereas under the unconstrained model, 48 percent of the racial gap was caused by discrimination. The GAO findings suggest that although the Federal guidelines are responsible for reducing some racial disparity in sentencing, the problem of discrimination persists within the guideline range. Samuel Myers was especially concerned with the racial disparity issue.9 His review of the analyses of USSC and GAO led him to conclude that racial inequities persist, although not always in a consistent manner: GAO computed odds ratios for the chances that a black versus a white offender received a sentence which fell below the guideline minimum for the convicted crime, or at the top, bottom or middle of the guideline range.... [The] report suggests wide variations across crimes, with blacks more often than whites receiving top of the range and above the range sentences for robbery, but at the same time within range and bottom of the range sentences for cocaine distribution.... The GAO also performed a residual difference analysis of Federal sentences for black and white bank robbers sentenced in the fiscal year of 1990. On average, blacks served about ten percent longer sentences than whites. In jury trials, the reverse was true; whites received longer sentences.10 Two studies have shed further light on the issue of racial disparity under the USSC guidelines, casting doubt on the quantity ratio established by Congress in the mandatory minimum penalties.11 Both studies center on USSC penalties imposed for possession or sale of crack cocaine. Under the Anti-Drug Abuse Act of 1986 and before guidelines were implemented, Congress modified the penal code to allow for much more severe sentences for drug crimes involving crack cocaine than for crimes involving powdered cocaine. In general, persons convicted of possessing or trafficking in 50 grams or more of crack cocaine would receive a prison term of not less than 10 years. If the offender had a prior drug conviction, the sentence would be not less than 20 years. If the amount of crack cocaine was at least 5 grams but less than 50 grams, the sentence would be at least 5 years, or 10 years with prior drug crime convictions. These same penalties would apply to offenders if they had 100 times these amounts of powdered cocaine. Berk and Campbell analyzed Federal charging practices in the Los Angeles area for crack cocaine.12 In California the Federal crack cocaine penalties are much more severe than those under the State's determinate sentencing laws. Consequently, the prosecutor's discretion to try a case in Federal court represents a major source of unwarranted disparity. The researchers found that in Los Angeles, while African-Americans represent 58 percent (n=1,801) of all arrests and 53 percent (n=4,410) of State prosecutions for crack cocaine, they account for 83 percent (n=36) of Federal prosecutions. In contrast, Caucasians represented 3 percent (n=56) of arrests and 3 percent (n=222) of State prosecutions, but zero Federal prosecutions. McDonald and Carlson conducted a pre- and postguideline analysis of sentencing in the Federal courts by race. They found that prior to guideline implementation, there were no differences between race/ethnicity and sentences imposed: Fifty-four percent of all white as well as black offenders in these cases were given prison sentences that were comparable in length; a maximum of 50 months, on average, for whites, and 53 months for blacks. Hispanics were more likely to be imprisoned (69 percent), but their sentence lengths averaged approximately the same length: 52 months.13 Since guidelines were adopted, however, the proportion of African-Americans and Hispanics sentenced to prison by 1990 had grown to 78 and 85 percent, respectively, compared with a 72-percent increase for Caucasians. African-Americans sentenced to prison received an average sentence length of 71 months, compared with 50 months for Caucasians and 48 months for Hispanics. The researchers attribute most of these differences to the 1986 drug sentencing laws and the USSC guidelines. Between 1986 and 1990, the proportion of African-Americans sentenced for drug trafficking grew from 19 percent to 46 percent, compared with 26 percent to 35 percent for Caucasians. In evaluating the source of this disparity among African-Americans and Caucasians, the authors identified the source of disparity as the crack cocaine drug laws: Among cocaine traffickers, however, there were substantial differences. Blacks were imprisoned somewhat more frequently than Caucasians (96.7 percent versus 94.1 percent), and their sentences were significantly longer, on average: 102 months versus 74 months--a 37 percent difference.14 Because African-Americans constituted 83 percent of the crack cocaine trafficking cases, they were disproportionately sentenced to prison and received far longer prison terms than Caucasians: These and other analyses not reported here lead us to conclude that the source of the differences in sentencing of blacks and whites under the guidelines was not invidious discrimination by judges at the point of sentencing. Instead, the primary reasons were the predominance of blacks in Federal crack trafficking cases, and Congress's decision to punish crack cocaine severely.15 The widespread use of mandatory minimum penalties illustrates how sentencing reforms external to the guidelines themselves can negate gains in reducing disparity. The USSC conducted a study of the effects of these mandatory minimum penalties on sentencing disparity and found that they were not applied in a uniform manner. On the basis of its analysis, USSC reached the following conclusion: Despite the expectation that mandatory minimum sentences would be applied to all cases that meet statutory criteria of eligibility, the available data suggest that this is not the case. The lack of uniform application creates unwarranted disparity in sentencing, and compromises the potential for the guidelines sentencing system to reduce disparity.16 The use of departures within the Federal system is another issue for consideration. After 5 years of nationwide guideline implementation, one of the more disturbing trends in the Federal system has been an increase in the annual rate of downward departures based on substantial assistance provided to authorities by offenders (also known as cooperation). Before substantial assistance departures can be made, a motion must be filed by the Government in court, outlining the defendant's assistance. In 1989 the substantial assistance departure rate was 5.8 percent, rising to 7.5 percent in 1990, 11.9 percent in 1991, 15.1 percent in 1992, and 16.9 percent in 1993.17 Depending on which defendants the government decides should receive motions for downward departure, considerable disparity could be brought back into the system. The U.S. Sentencing Commission is conducting a major study of these departures, with a report expected in January 1996. Collectively, results of all the studies described in this chapter suggest that the USSC guidelines have reduced disparity but that more research is needed to better understand the overall impact on sentencing disparity reduction. The extent of racial disparity in the use of incarceration has worsened under the mandatory minimum-driven drug guidelines, with dramatically increased penalties for which African-Americans are disproportionately arrested and convicted. Minnesota Sentencing Guidelines Much research has been conducted on the effects of Minnesota's guidelines, with the most thorough study having been completed by the Minnesota Sentencing Guidelines Commission (MSGC) in 1984. The MSGC evaluation used a preguidelines sample (1978) and several postimplementation samples (1981-1983) to assess impact on sentencing practices. Using a measure of grid variance, MSGC found that greater uniformity was achieved by the guidelines than by the previous indeterminate system. Part of this achievement could be attributed to the low level of departures from the prescribed guideline-based dispositions. Initially, only 6.2 percent of the cases departed from the guidelines and did so in no singular direction. However, departure rates increased slightly during the succeeding 2 years, from 6.2 to 8.9 percent. An article published by Stolzenberg and D'Alessio provides further evidence that the early gains in reducing disparity have slowly eroded. In their analysis of sentencing practices between 1977 and 1989, the authors found that disparity in the length of sentence decision has been reduced by 60 percent.18 However, for the prison disposition decision, substantial gains realized immediately after guideline implementation have not been maintained. The estimated reduction for the in/out decision is only 18 percent and is approaching preguidelines levels. Regarding sentencing neutrality, "there were only marginal gains."19 While dispositions for African-Americans and Native Americans were more uniform than they were before the guidelines, reductions in sentencing variations for these minorities were less than reductions for Caucasians. Minority offenders continued to receive somewhat harsher sanctions than Caucasian offenders, even when offense severity and criminal history are controlled for. Employment status continued to be associated with sentence length. Sentences for imprisoned women averaged 7 months less than those for similarly situated male prisoners. Nonetheless, it is generally agreed that racial and gender disparity have declined modestly under the guidelines.20 Washington Sentencing Guidelines The Washington State Sentencing guidelines were intended to ensure long-term imprisonment for violent offenders, community sanctions for property offenders, and greater equity in the sentencing process. Neutrality in sentencing was a major goal stated in the Sentencing Reform Act (SRA), which required application of the guidelines to all offenders equally, without regard to any offender characteristics except for offense severity and criminal record. Reviewing the State's first 10 years' experience with sentencing guidelines, the Washington Sentencing Guidelines Commission (WSGC) concluded: The high degree of compliance with sentencing guidelines has reduced variability in sentencing among counties and among judges. Moreover, the great majority of sentences fall within the standard ranges, and they tend to be gender and ethnicity neutral.21 This conclusion was supported by earlier findings that little disparity was found in the lengths of sentences imposed for nondeparture sentences (87 percent of all sentences fell within the standard sentencing range).22 In general, the guidelines have produced the desired results: higher probabilities and longer sentences for persons convicted of violent crimes, less imprisonment for property crimes, and high compliance with the guidelines. African-American offenders received longer sentences than Caucasians, but these differences were attributed to ethnic differences in the offense seriousness level and overall offender score patterns. Results were confirmed by imposing the average sentence lengths received by Caucasians in various crime and criminal history categories on similarly situated African-American offenders. Although evidence failed to show systematic ethnic disparity among nondeparture cases, substantial disparity according to ethnicity and race was revealed in the use of the First-Time Offender Waiver (FTOW) and the Special Sexual Offender Sentencing Alternative (SSOSA)23. Both of these sentencing dispositions can be used by the judge to depart from the guideline-based sentences. Although these are not viewed as departures, they are frequently invoked. For example, in FY 1992, 13 percent of all sentences were FTOW, with another 2 percent being SSOSA.24 In 1987, only 8 percent of all cases received such a sentence. Caucasians were three times as likely as African-Americans and half as likely as other minorities to receive a FTOW sentence below the standard sentencing range. Caucasians were also four times as likely as African-Americans and twice as likely as other minorities to receive a suspended sentence under SSOSA. The disparity in FTOW and SSOSA departures could be partly explained by the lack of treatment programs and other resources provided for minority defendants, but nontreatment alone cannot account for most of the disparity in the frequency of these departures. The Washington report also examined gender disparity in sentencing and found that females systematically received shorter sentences, even when studies controlled for crime severity and criminal history. Females were more than twice as likely as males to receive a downward departure through the FTOW. Oregon Sentencing Guidelines Oregon's Criminal Justice Council (OCJC) has also compared cases sentenced prior to and after guidelines went into effect in 1986. The first-year report on Oregon's sentencing guidelines compared the two samples and reached the following conclusion: . . . [G]uidelines have increased uniformity in sentencing substantially. 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.25 This statement was based on a computation of the same grid variance used by Minnesota and Washington in their evaluations. As was done in Minnesota, researchers applied the guideline grid criteria to both the pre- and postguideline implementation samples. For the 1986 sample, the grid variance equalled 0.100, while the guideline cases had a grid variance of 0.045. On the basis of these two statistics, the OCJC concluded that "... since the guidelines went into effect, the degree of dispositional uniformity has increased by 55 percent."26 According to the OCJC, this level of improvement in dispositional uniformity was similar to levels reported by Minnesota and Washington (54 percent and 73 percent, respectively).27 No pre- and posttests were conducted on variance reductions in sentence length or estimated time served. The first report found that sentencing disparity by race and gender had been reduced. However, "significant racial and gender disparity continues" for those cases that are based on the two sentencing options for which judges have discretion: optional probation and sentencing departures (approximately 9 percent of all cases sentenced were one of these two options). The third-year report was based on 1992 cases sentenced under the guidelines and notes that the departure rate steadily increased (from 6 percent in 1990 to 10 percent in 1992). The report also makes a more definitive statement that racial and gender disparity persists under the guidelines: Generally, men were convicted of more serious offenses and had more serious criminal histories than women. Minorities were convicted of more serious offenses and had more serious criminal histories than whites. Under guidelines, the presumptive sentence is based on an offender's crime seriousness and criminal history classification. Thus, male and minority offenders were more likely to have a presumptive sentence of prison. And, actual imprisonment rates were higher for men and minorities.28 The report also found that: Sentencing disparity is not entirely due to differences in current and prior conviction offenses. More details are needed on the crimes of conviction to determine if the specifics of the offender's criminal acts are the source of the disparity in sentencing.29 While differences in sentencing severity have been reduced among regions of the State, they persist. Pennsylvania and Delaware Sentencing Guidelines Neither Pennsylvania nor Delaware have conducted a pre- and postguidelines implementation study; they have focused instead on consistency within the State and/or compliance to guideline criteria. In Delaware, three studies have found that the guidelines have succeeded in sentencing more violent offenders to prison and increasing the number of nonviolent offenders sentenced to intermediate sanctions.30 In Pennsylvania, the Commission staff found high levels of compliance and disparity among the counties having declined since the guidelines were implemented.31 The results of several analyses of sentencing under Pennsylvania's guidelines have shown a pattern similar to that in Washington and Oregon.32 Although the factors prescribed by guidelines (offense severity, prior record) were overwhelmingly the strongest predictors of sentencing outcomes, significant differences exist by race, gender, region of the State (e.g., rural versus urban), and sentencing differences between those who plead guilty and those convicted by jury trial are especially great. Dispositional departures below guidelines were a substantial contributor to differences in gender and race. Summary of Disparity Findings There is substantial evidence to show that sentencing disparity has been reduced by presumptive guidelines in several jurisdictions. However, there is also evidence to show that some of the early progress achieved in these States may be slowly eroding. Tonry concludes that although sentencing disparity has "most likely" been reduced, " . . . the evaluation research evidence on this question is less definitive than it appears or than its celebrants claim."33 One common theme among researchers' findings is that guideline departures--especially departures below guidelines--significantly contribute to disparity and especially race and/or gender sentencing disparities.34 That is, Caucasians and females are more likely to receive departures below guidelines and less likely to receive departures above guidelines. Similarly, Caucasians are more likely to receive options that are not officially labeled as departures but serve the same function. Data summarize the departure rates for five States and the Federal guidelines as assembled by Tonry.35 Standard sentences are shown applied to 75-80 percent of all cases, and the vast majority of departures are below the guidelines. As long as departure rates remain at this level and in a downward direction, the question of disparity reduction remains somewhat clouded. Impact on Plea Bargaining Some feared that guidelines would give prosecutors a disproportionate level of discretion over sentencing and result in more charge bargaining.36 However, limited research as well as anecdotal evidence indicates that this fear is unfounded. For example, Miethe reports that, after guidelines were instituted in Minnesota, prosecutors did not engage in greater charge bargaining and sentence bargaining cases did not show greater race or class disparity than those prior to guidelines.37 Cirillo,38 Ulmer and Kramer,39 and Ulmer40 found that Pennsylvania's "looser" guidelines allowed many "windows of discretion" beyond the charging stage, such as mitigated- and aggravated-range sentences and relatively easy departures, that allowed judges ample sentencing discretion and power in the courtroom. In Washington, there has been no noticeable increase in trial rates.41 Evidence exists, however, that mandatory minimum penalties produce greater charge bargaining, disproportionate prosecutorial power, and hidden disparities. The USSC concluded that mandatory minimums are not being applied uniformly and have granted more discretion to prosecutors.42 Austin found that Florida's habitual sentencing law was applied in only 20 percent of the cases that met the criteria for being "habitualized" and that African-Americans were nearly twice as likely as Caucasians to receive such a sentence.43 There is a trade-off between allowing courts to alter sentences to reflect unusual circumstances and reducing disparities in sentencing. By radically constricting judicial discretion, mandatory minimum penalties severely constrain the sentencing process and move the locus of disparity to the charging stage, where it is less visible. Impact on Incarceration Rates The States with presumptive guidelines and the USSC have shown higher incarceration rates since guidelines were implemented. However, all States, regardless of their sentencing structure, have experienced increases in their incarceration rates. Contributing to this growth is an increase in the certainty and length of incarceration for persons convicted of violent crimes and drug violations. At the same time, some States' guidelines reduce the length of imprisonment for nonviolent offenders to offset the effects of more punitive sentences for violent, drug, and repeat offenders on prison population growth. The following pages contain a comparison of trends in incarceration rates in four States that have sentencing guidelines with trends in States without guidelines. It should be noted that prison population growth is driven by two basic forces--admissions and length of stay (LOS). Sentencing guidelines affect admissions and LOS, but they are not the only factors. Prison admissions are also influenced by demographic trends, crime rates, arrest practices, and prosecutorial policies. In some guideline States, the offender's LOS is regulated by good-time (both statutory and discretionary) and parole boards that remain outside the domain of sentencing commissions. Also, there may be considerable legislative activity, such as mandatory prison terms and/or adoption of "truth in sentencing" provisions, that affect admissions and LOS. For these reasons, sentencing guidelines or sentencing structures are not the sole cause of incarceration rates. Data summarize rates of incarceration (the number of persons in prison on any given day per 100,000 U.S. population) for the Nation and the four guideline States. On a national level, incarceration rates have increased dramatically over the past two decades. The four guideline States have also increased their incarceration rates, but not always in line with the national trends. The guideline States vary substantially in their preguideline incarceration rates. Florida has traditionally had very high rates, Pennsylvania and Washington reported moderate rates, and Minnesota has historically had a low incarceration rate. After each State adopted its guidelines, these trends persisted, with some variations. Washington has experienced a very modest increase since its guidelines were introduced in 1983. This low-growth pattern, however, masks a long-term trend that has not yet materialized. Washington's guidelines, along with several State supreme court rulings, were intended to reduce the rate and length of imprisonment for persons convicted of property and drug possession crimes. Counteracting that trend was the policy of substantially increasing prison terms for inmates convicted of violent and sex crimes. The overall effect was an immediate decline in prison population followed by a substantial increase that is projected to continue well into the future as the violent offenders sentenced for longer terms begin to "stack up" in the prisons. Florida has substantially increased its incarceration rate since adopting guidelines. The Florida trends would have been even higher were it not for a consent decree and subsequent legislation that required the State to maintain its inmate population at 98.5 percent of its prison-bed capacity. This restriction resulted in Florida's well-publicized early release program (referred to as Controlled Release) to keep the prison population within its rated capacity.44 Pennsylvania, while maintaining relatively moderate incarceration rates compared to Florida, also experienced an increased rate after adopting guidelines. This growth pattern is not surprising since Pennsylvania's guidelines were not designed to address prison crowding. Minnesota, which did link its guidelines to correctional resources, has shown the lowest rate of growth. In fact there was virtually no growth until 1985. Since then, the State's incarceration rate has increased. This may be explained in part by major revisions to the guidelines that were adopted in 1988--revisions that were intended to make the guidelines more punitive for offenders convicted of violent crimes.45 The impact of guidelines on incarceration rates can be seen by contrasting two States that linked their guidelines to correctional resources (Minnesota and Washington) with States that have comparable preguideline incarceration rate trends. Both Minnesota and Washington fared far better than the nonguideline States. These data prove that guidelines can have a moderating effect on the use of imprisonment if they are designed to accomplish that goal. Impact on Prison Crowding As incarceration rates have increased, what has been the effect on prison crowding? Do sentencing guidelines have any effect on prison overcrowding? Do they exacerbate or relieve the problem? To answer these questions, the rates of crowding were examined, using data provided by the U.S. Department of Justice's Bureau of Justice Statistics (BJS). Unfortunately, national and individual State prison capacity data were not reported in a standardized format until 1983. Consequently, it is not possible to compare the magnitude of overcrowding before and after the implementation of guidelines, as was done for incarceration rates. The BJS data use two measures of prison capacity. One is based on the State's highest bed capacity figure, which may include a State's temporary, operating, or emergency bed capacity. The lowest capacity figure often reflects the design capacity of the system. However, in some instances, States have reported the same figure for both lowest and highest bed capacity.46 Data report the crowding ratios (prison population divided by prison capacity) for the four guideline States since 1983 (ratios above 100 reflect a crowding situation). Here again there are important differences among the four States. Pennsylvania, which did not require the guidelines to take capacity into account, has had the worst overcrowding problem; its prison population has always exceeded its prison capacity, starting at 122 percent and climbing to approximately 150 percent. Prison crowding in Washington has fluctuated along with its prison population due to its guidelines, two State supreme court decisions, and numerous amendments to the SRA that increased penalties. As indicated earlier, there was a considerable drop in the inmate population during the first few years of the SRA up through 1988. Supreme court decisions and the guidelines allowed many inmates to serve shorter prison terms or no terms at all. Consequently, Washington had excess capacity and became well known throughout the Nation as a State with excessive capacity that would rent cells to other States to keep its prisons full and make money. However, since 1988, as the long-term effects of its sentencing guidelines have taken hold, and its prison population has grown, Washington's crowding situation has worsened. This trend is expected to continue to deteriorate over the next decade. Furthermore, the Washington State Guidelines Commission noted that jail crowding has worsened under the guidelines and has concluded that the objective of making "frugal use of the State's resources" has not been met.47 Florida has succeeded in keeping its prison system from overcrowding, despite its sentencing guidelines. As noted earlier, a court order that later became law required that the prison population not exceed bed capacity. As a result, the system remained uncrowded because pressures created by the guidelines and mandatory minimum sentencing laws were offset by a massive early release program.48 Only Minnesota has been able to provide adequate capacity for its prison inmates over a sustained period of time. However, as the prison population began to exceed its rated capacity beginning in 1991, its early success seemed to wane. Moreover, Minnesota never had a prison crowding problem to be solved. An earlier review by Austin pointed out that before guidelines were enacted, Minnesota was not experiencing overcrowding: Prior to the implementation of the guidelines, Minnesota's prison population had never exceeded its bed capacity and has traditionally reported one of the nation's lowest state incarceration rates. Due to a surplus of beds, the state has been housing several hundred inmates from Wisconsin and the Federal Bureau of Prisons in their facilities since 1980. A new maximum security facility (Oak Park Heights) was brought on line in 1983 which further added capacity to the State prison system. Thus, in many ways, the inclusion of prison capacity as a basis for setting guideline criteria may have reflected a long standing Minnesota tradition to not overcrowd its prisons regardless of sentencing structure.49 Minnesota's success in avoiding prison overcrowding, then, is not linked solely to guidelines for two reasons. First, before guidelines were enacted, no prison crowding problem existed. Second, since the guidelines were modified significantly in 1989 to increase prison terms for certain offenses, incarceration rates have increased and the prison system has become more crowded. These same modifications were made to appease a strong movement to abolish the guidelines and institute mandatory minimum penalties--a movement that was barely defeated. The reform also created two discretionary release programs under the control of the Department of Corrections that threaten the new-found goal of "truth in sentencing."50 Data pool the results of Minnesota and Washington with the rest of the Nation's prison systems and show little difference between guideline and nonguideline States. The finding that guidelines by themselves are not a sufficient condition for controlling prison crowding is clearly shown in data, which present States that have prison populations below their rated capacity. Of the 15 States listed, 5 are guideline States. Of the five guideline States, two (Louisiana and Utah) have had voluntary/advisory guidelines since 1992. Clearly, prison crowding can be controlled with or without guidelines. Projected Prison Populations for Guidelines States To what extent will guideline States experience future prison population growth and crowding? To answer this question, researchers examined prison population projections that were available for seven different sentencing guideline States and the Federal Bureau of Prisons. These population projections are the latest official forecasts of prison growth for each system. In each case, the prison population is expected to grow, and most will continue to be overcrowded unless capacity is increased. The specific trends for each jurisdiction are as follows: o The population of the Federal Bureau of Prisons is projected to increase by 22,546 inmates from 1995 to 1999 and by 33,635 inmates by the year 2002. The Federal system is projected to decrease the level of overcrowding by increasing its bed capacity by 37,714 beds from 1995 to 1999 and by 52,500 beds for the year 2002. Nonetheless, the Federal system will have 4,657 fewer beds than required by the year 2002. o Florida is projected to have 83,802 inmates by 1998 from its projected 1995 population of 74,669. However, its bed capacity is projected to reach only 59,556 in 1998, resulting in a shortfall of 24,246 beds. o Minnesota's prison population will grow to 5,848 inmates in the year 2002, but the State will have a capacity of only 4,105 beds. Unless substantial increases are made to the State's bed capacity, Minnesota will soon break its long tradition of being uncrowded. o North Carolina's prison population is projected to increase from 24,968 inmates in 1995 to 32,084 inmates by the year 2003. The 1995 capacity of 24,136 beds is projected to increase to 28,247 beds by 1997 and remain constant, resulting in a bed shortage of 3,837 in the year 2003. o The prison population in Oregon is projected to grow from 6,555 in October 1993 to 7,333 inmates in July 1995. o Pennsylvania is anticipating prison population growth from 29,667 in 1995 to 33,154 by the year 2000. The projected bed capacity for the same year is 24,910, resulting in a shortfall of 8,244 beds. o The prison population for Tennessee will grow from 14,571 inmates in 1993 to 17,328 in 1996. The projected capacity will be 15,979 in 1996, creating a shortfall of 1,349 beds. o Washington's prison population is projected to grow from 10,683 in 1994 to 14,121 inmates by the year 2000. Unless more bed capacity is planned and constructed, there will be a shortfall of 2,528 beds. At least two of these guideline States have been forced to use early release programs in the past to relieve overcrowding. As indicated earlier, Florida has a program called Controlled Release. Although a number of legislative restrictions keep many inmates from being released early, the program awards credits to some inmates on a discretionary basis to accelerate release dates and thereby reduce lengths of stay. Tennessee reduces time served by using a Safety Valve program that reduces lengths of stay until first parole eligibility.51 In summary, unless prison capacity is increased significantly or current sentencing laws are modified, guideline States will be forced either to live with overcrowded facilities or implement early release programs that invariably preclude achieving the objective of reducing disparity. However, it can be argued that these trends would have been far worse had guidelines not been in place. Although there are no published studies to verify this claim, data provided by the North Carolina sentencing commission show that, without guidelines, the State's future crowding problem would be far greater. Impact on Public Safety and Crime Rates The issue of whether sentencing guidelines (or other forms of structured sentencing) should result in reduced crime rates remains somewhat controversial. Certainly, a number of State guideline systems promote public safety as one of the purposes of sentencing. How sentencing commissions define public safety and incorporate it into their guideline systems, however, may be quite different from how public safety is understood by legislative bodies and the general public. For example, some argue that removing violent offenders from the community increases public safety; others argue that targeting violent offenders or those likely to commit multiple offenses for long prison terms reduces the crime rate and therefore increases public safety. Sentencing commissions argue that although guidelines can identify offenders who pose the greatest threat to society and provide sufficient penalties to remove them from society for long periods of time, sentencing is one of the final stages in a complex criminal justice system and should not shoulder the burden for reducing crime rates. Decisions at the reporting, arrest, charging, and plea bargaining stages play equally important roles in determining whether crime rates will change. Despite politicians' hopes that sentencing reform, and in particular the increased use of incarceration through sentencing, will reduce crime,52 States should not be lulled into the belief that this reform can solve the crime problem or even necessarily reduce State crime rates. By selectively incapacitating those offenders who pose the greatest threat to society, some crimes may be averted and public protection enhanced. Generally, States with guidelines strive to ensure public safety by targeting violent and career offenders for lengthy prison stays. However, no State has formally claimed that guidelines will measurably reduce crime rates or set as one of its guideline goals the reduction of crime rates. The attempt to link crime reduction to sentencing structure is further weakened by the realization that guidelines tend to discriminate among offenders for sentencing dispositions based on two factors unrelated to criminal careers. The first-time youthful offender is likely to receive a very brief period of incarceration even though that person poses a far greater risk to public safety because of age, sex, employment, marital status, education, drug use, and residence, factors unrelated to criminal careers but not incorporated in the guidelines. At the same time, an older offender with a long history of arrests and convictions is more likely to receive a far longer prison term even though that person is now "burning out" of his/her criminal career. Such a sentencing policy will fill a State's prison system with less risky and older inmates, thus diminishing possible incapacitation and subsequent crime effects. In his recent study of Federal prisoner recidivism rates, Harer points out that a significant portion of drug offenders, who are now receiving far longer prison terms under the mandatory minimum sentencing guidelines for drug offenses, are very low risks to public safety. He questions these lengthy mandatory prison terms as an effective crime control strategy. . . . [I]t does seem clear that by imposing longer sentences on low-risk drug traffickers, there has been a substantial increase in retribution with little, if any, reduced sentencing disparity, or increased incapacitation, deterrent, or rehabilitation value and a huge increase in taxpayer dollars spent.53 Because sentencing policies deal with a small subset of those who enter the criminal justice system, increases in incarceration are likely to affect crime rates only marginally. Consequently, the influence of sentencing guidelines on crime rates is extremely limited. It may be useful to demonstrate this limited influence by examining the relationship between adoption of sentencing guidelines and changes in crime rates. Although such an examination cannot attribute a causal relationship, it may illuminate the degree to which any sentencing reform can reliably affect crime rates. Data show crime rates for four guideline States: Minnesota, Pennsylvania, Washington, and Florida. These States were selected based on the same criteria as described in the previous section. The crime measures used were extracted from the Uniform Crime Reports (UCR), which include eight crime categories reported by the public to the police and tabulated by the FBI. Violent crime, as defined in the UCR, is composed of four offenses: murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault. The other major crime category, property crime, includes burglary, larceny-theft, motor vehicle theft, and arson. The rates described in this analysis are based on the number of crimes per 100,000 total population. Minnesota and Pennsylvania had lower crime rates than Washington and Florida. Between 1970 and 1980, crime rates increased. Since then, crime rates have remained relatively constant or have declined slightly. In general, these trends are similar to national trends. The dip in crime rates from 1980 to 1983 was similar for all guideline States even though they implemented guidelines in differing years. After years of steady increase starting in 1983, Washington has experienced a moderate decline since 1987, the year it implemented its guidelines. Comparisons of total crime rates between guideline and nonguideline States that had similar crime rates show few differences. The slight increase after 1983 for the guideline States is wholly attributable to Florida (total and violent crime rates) and Minnesota (violent crime rates only). As expected, this analysis suggests that the adoption of guidelines or other sentencing models has had little measurable impact on overall crime rates. This lack of impact, however, should not be viewed as a failure of sentencing reforms or guidelines. With improved reliability in State sentencing information during and following guideline implementation, States are better able to identify those violent and career criminals who invoke the greatest fear and to focus scarce correctional resources on those offenders. Summary This chapter has examined the available evidence regarding the impact of guidelines on disparity, plea bargaining, incarceration rates, prison growth, prison overcrowding, and crime rates. For those States that have conducted evaluations, clear evidence exists that guidelines have succeeded in changing historic sentencing patterns. The probability of prison and the length of imprisonment have increased for persons convicted of violent crimes and/or repeat offenders but have decreased for persons convicted of property crimes. Following is a summary of the general findings contained in this chapter. Sentencing Disparity While a number of methodological problems exist in evaluating the impact of guidelines on disparity, those States that have examined disparity show improvement under the guidelines. Compliance rates are high, and substantial evidence suggests that sentencing disparity has been reduced in States with guidelines. However, the growing frequency of departures and potential increased reliance on mandatory minimum penalties may threaten those gains in disparity reduction. Plea Bargaining Limited research and anecdotal evidence indicate that guidelines have not increased charge bargaining and, contrary to some expectations, have not shifted sentencing discretion to prosecutors. Evidence suggests, however, that mandatory minimum penalties produce greater charge bargaining, increased prosecutorial discretion, and hidden disparities, particularly among racial and ethnic groups. Imprisonment Rates States with guidelines have significantly increased their incarceration rates as have all other States. However, the growth in imprisonment rates was far lower for the two States (Minnesota and Washington) that had established guidelines with the goal of managing prison populations. Consequently, evidence indicates that guidelines can have a moderating effect on the use of imprisonment if they are designed to accomplish that end. Prison Overcrowding Guideline States, like most other States, have experienced significant prison overcrowding. Only Minnesota and Florida have had some success in keeping prison crowding under control, possibly because of factors beyond sentencing guidelines. By statute, Florida must provide sufficient bed capacity in its prisons, and Minnesota had no overcrowding problem at the time its guidelines were implemented. In general, guideline systems attempt to regulate prison population at the front end through determinate sentencing policies rather than at the back end through discretionary release. Each State must decide on the purposes of sentencing; however, without some attention to prison overcrowding, the elimination of parole and other discretionary release mechanisms likely will result in serious prison overcrowding. Prison Populations As in the rest of the Nation, prison populations are projected to grow in States with sentencing guidelines. To avoid overcrowding, these States will have to add more prison capacity or adjust their guidelines to reduce prison lengths and/or prison admissions. Without these changes, overcrowding conditions probably will continue, and some States may be forced to reintroduce discretionary release practices. Crime Rates Although no guideline State suggests that its sentencing structure reduces crime rates, some proponents of structured sentencing hope for a beneficial impact on crime. This limited analysis finds no evidence that guidelines or other forms of structured sentencing have measurably reduced crime rates. Few differences in crime rates exist between guideline and nonguideline States. Under guidelines, however, States attempt to provide greater public safety through sentencing policies tailored to ensure that violent and career offenders are more likely to be incarcerated and incarcerated for longer periods of time. In summary, considerable progress has been made through sentencing guidelines, but much work remains. Tonry, a long-time evaluator of sentencing reform, concluded: Some commissions have operated much as Judge Frankel hoped they would; they have achieved and maintained specialized institutional competence, have to a degree insulated sentencing policy from short-term "crime of the week" political pressures, and have maintained a focus on comprehensive systemwide policy making. Guidelines promulgated by commissions have altered sentencing patterns and practices, and have shown that policies can be linked to correctional and other resources, thereby enhancing governmental accountability and protecting the public purse.54 Simply stated, the success of guidelines depends on their design, purpose, and implementation. Guidelines will not solve all the problems of dispensing justice and reducing crime. As this chapter shows, however, some guideline systems have been instrumental in altering sentencing policies, setting resource priorities, reducing disparity, and helping to manage prison populations. It is important to note the relative absence of research on these important topics. The lack of rigorous studies that employ pre- and postguideline samples of disparity is striking. Only three States and the USSC have provided the field with such studies. Given the high priority that States place on reducing unwarranted disparity and given the fact that many States are adopting guideline systems, now is the time to launch a series of such studies. A related need is to conduct more comprehensive studies on whether guidelines have altered plea bargaining practices and whether they enhance the influence of prosecutors by shifting discretion from the courts to the prosecution. Such studies should consist of both quantitative and qualitative methods (interviews and observations of the plea bargaining process) to determine whether disparity has been reduced throughout the entire system or has simply been displaced. The growing number of African-Americans and Hispanics being incarcerated also needs to be addressed in terms of whether guidelines and/or mandatory minimum penalties are contributing to racial disparity at the points of charging and sentencing. Finally, studies are required to document the extent to which the attributes of the prison and nonprison populations have been altered and the consequences of such changes on prison management. If guidelines result in nonviolent offenders' receiving very short or no prison terms while violent and repeat offenders are incarcerated more frequently and for longer periods of time, prison populations will look very different in the future. How these changes will affect the number and type of prison beds, along with the need for special housing units and services (medical, administrative segregation, and protective custody populations), needs to be understood. ------------------------------ Notes 1. Tonry, M. "Sentencing Commissions and Their Guidelines." In Crime and Justice: A Review of the Research, vol. 17, ed. M. Tonry, pp. 137-196. Chicago: University of Chicago Press. 1993. 2. 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, 1991. 3. Note 2, p. 279. 4. U.S. Sentencing Commission. Annual Report-- 1991. Washington, DC, 1992. 5. Note 4, Table 29 (not available in this format). 6. Heaney, G.W. "The Reality of Guidelines Sentencing: No End to Disparity." American Criminal Law Review 28(2)(1991), pp. 161-232. 7. Note 2. 8. U.S. General Accounting Office. Sentencing Guidelines--Central Questions Remain Unanswered. Washington, DC, 1992. 9. Myers, S.L., Jr. "Racial Disparities in Sentencing: Can Sentencing Reforms Reduce Discrimination in Punishment?" University of Colorado Law Review 64(3)(1993), pp. 781-808. 10. Note 9, pp. 793-794. 11. Berk, R., and A. Campbell. "Preliminary Data on Race and Crack Charging Practices in Los Angeles." Federal Sentencing Reporter 6(1)(1993), pp. 36-38. See also McDonald, D, and K.E. Carlson. "Why Did Racial/Ethnic Sentencing Differences In Federal District Courts Grow Larger Under the Guidelines?" Federal Sentencing Reporter 6(4)(1994), pp. 223-226. 12. Note 11. 13. McDonald, D, and K.E. Carlson. "Why Did Racial/Ethnic Sentencing Differences in Federal District Courts Grow Larger Under the Guidelines?" Federal Sentencing Reporter 6(4)(1994), pp. 223- 226. 14. Note 13, p. 224. 15. Note 13, pp. 225-226. 16. U.S. Sentencing Commission. Special Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System. Washington, DC, 1991. 17. Personal communication with Phyllis Newton, U.S. Sentencing Commission. 18. Stolzenberg, L., and S.J. D'Alessio. "Sentencing and Unwarranted Disparity: An Empirical Assessment of the Long-Term Impact of Sentencing Guidelines in Minnesota." Criminology 32(2)(1994), pp. 301-310. 19. Parent, D.G. Structuring Criminal Sentences: The Evolution of Minnesota's Sentencing Guidelines. Stoneham, MA: Butterworth. 1988. 20. Miethe, T.D., and C.A. Moore. "Socioeconomic Disparities Under Determinate Sentencing Systems: A Comparison of Pre- and Post Guideline Practices in Minnesota." Criminology 23(1985), pp. 337-363. See also Miethe, T.D., and C.A. Moore. "Racial Differences in Criminal Processing: The Consequences of Model Selection on Conclusions About Differential Treatment." Sociology Quarterly 27(1986), pp. 217-237. See also Miethe, T.D., and C.A. Moore. "Can Sentencing Reform Work? A Four-Year Evaluation of Determinate Sentencing in Minnesota." Paper presented at the annual meeting of the American Society of Criminology, November 1988. See also Moore, C.A., and T.D. Miethe. "Regulated and Non-Regulated Sentencing Decisions: An Analysis of First-Year Practices Under Minnesota's Felony Sentencing Guidelines." Law and Society Review 20(2)(1986), pp. 253-277. See also Miethe, T.D. "Charging and Plea Bargaining Practices Under Determinate Sentencing: An Investigation of the Hydraulic Displacement of Discretion." Journal of Criminal Law and Criminology 78(1)(1987), pp. 155-176. See also Frase, R.S. "Implementing Commission-Based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota." Wake Forest Law Review. 1993. 21. Washington State Sentencing Guidelines Commission. A Decade of Sentencing Reform: Washington and Its Guidelines 1981-1991. Olympia, 1992. 22. Fallen, D.L. Sentencing Practices Under the Sentencing Reform Act. Olympia: Washington State Sentencing Guidelines Commission. 1987. 23. First-time offenders not convicted of a sex offense or a violent offense can be sentenced under a broader range of sentence conditions permitted by the First-Time Offender Waiver. The Special Sexual Offender Sentencing Alternative allows a determinate sentence to be suspended so that treatment programs can be imposed. If an offender fails to comply with the sentence conditions, the suspension will be revoked and the original sentence imposed. 24. Note 21, p. 21. 25. Ashford, K., and C. Mosbaek. First Year Report on Implementation of Sentencing Guidelines: November 1989 to January 1991. Portland: Oregon Criminal Justice Council. 1991. 26. Note 25, p. 17. 27. Note 25, p. 72. 28. Mosbaek, C. Third-Year Report on Implementation of Sentencing Guidelines: 1992. Portland: Oregon Criminal Justice Council. 1993. 29. Note 28. 30. Gebelein, R.S. "Sentencing Reform in Delaware." Overcrowded Times 2(2)(1991), pp. 5, 12-13. See also Quinn, T.J. "Delaware Sentencing Guidelines Achieving Their Goals." Overcrowded Times 3(1)(1990), pp. 1-2. See also Quinn, T.J. "Voluntary Guidelines Effective in Delaware." Overcrowded Times 3(1)(1992), pp. 1, 9, 11. 31. Pennsylvania Commission on Sentencing. Sentencing in Pennsylvania: Annual Report. University Park, 1991. See also Kramer, J.H., and R.L. Lubitz. "Pennsylvania's Sentencing Reform: The Impact of Sentencing Commission-Established Guidelines." Crime and Delinquency 31(1985), pp. 481-500. 32. Kramer, J.H., and D. Steffensmeier. "Race and Imprisonment Decisions." The Sociological Quarterly 34(2)(1993), pp. 357-376. See also Steffensmeier, D., J. Kramer, and C. Streifel. "Gender and Imprisonment Decisions." Criminology 31(3)(1993), pp. 411-446. See also Ulmer, J. Trial Court Communities Under Sentencing Guidelines: Sentencing, Case Processing, and Organizational Relations. Ph.D. dissertation, The Pennsylvania State University. Ann Arbor, MI: University Microfilms International. 1993. See also Kramer, J.H., and J. Ulmer. "Sentencing Disparity and Guideline Departures." Criminology, in review. 33. Note 1. 34. Note 1. See also Kramer, J.H., and D. Steffensmeier. "Race and Imprisonment Decisions." The Sociological Quarterly 34(2)(1993), pp. 357- 376. See also Steffensmeier, D., J. Kramer, and C. Streifel. "Gender and Imprisonment Decisions." Criminology 31(3)(1993), pp. 411-446. See also Kramer, J.H., and J. Ulmer. "Sentencing Disparity and Guideline Departures." Criminology, in review. See also Griswold, D. "Deviation From Sentencing Guidelines: The Issue of Unwarranted Disparity." Journal of Criminal Justice 15(1987), pp. 317-329. 35. Tonry, M. "Malign Neglect--Race, Crime and Punishment in America." Unpublished manuscript. 36. Alschuler, A.W. "Sentencing Reform and Prosecutional Power: A Critique of Recent Proposals for `Fixed' and Presumptive Sentencing." University of Pennsylvania Law Review 126(1978), pp. 550-577. See also Alschuler, A.W. "Departures and Plea Agreements Under the Sentencing Guidelines." Federal Rules Decision 117(1988), pp. 459-476. See also Tonry, M. "Sentencing Guidelines and Their Effects." In The Sentencing Commission and Its Guidelines, ed. A. von Hirsch, K. Knapp, and M. Tonry. Boston: Northeastern University Press. 1987. See also Savelsberg, J. "Law That Does Not Fit Society: Sentencing Guidelines as a Neoclassical Reaction to the Dilemmas of Substantivized Law." American Journal of Sociology 97(s)(1992), pp. 1346-1381. 37. Miethe, T.D. "Charging and Plea Bargaining Practices Under Determinate Sentencing: An Investigation of the Hydraulic Displacement of Discretion." Journal of Criminal Law and Criminology 78(1)(1987), pp. 155-176. 38. Cirillo, V. "Windows of Discretion in the Pennsylvania Sentencing Guidelines." Villanova Law Review 31(1986), pp. 1309-1349. 39. Ulmer, J., and J.H. Kramer. "Sentencing Guidelines and Case Processing in Two Court Communities." Paper presented at the annual meeting of the Midwest Sociological Society, Kansas City, MO. 1992. 40. Ulmer, J. Trial Court Communities Under Sentencing Guidelines: Sentencing, Case Processing, and Organizational Relations. Ph.D. dissertation, The Pennsylvania State University. Ann Arbor, MI: University Microfilms International. 1993. 41. Tonry, M. "Structured Sentencing." In Crime and Justice: A Review of the Research, vol. 10, ed. M. Tonry and N. Morris. Chicago: University of Chicago Press. 1988. 42. U.S. Sentencing Commission. Special Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System. Washington, DC, 1991. 43. Austin, J. Reforming Florida's Unjust, Costly and Ineffective Sentencing Laws. San Francisco: National Council on Crime and Delinquency. 1993. 44. Note 43. 45. Dailey, D. "Minnesota: Sentencing Guidelines in a Politically Volatile Environment." Federal Sentencing Reporter 6(3)(1993), pp. 144-146. 46. From a prison management perspective, relying on the highest capacity figure may understate the State's crowding problem because prisons generally require reserve capacity (5 percent) to operate efficiently and safely. For instance, prison dormitories and cells need periodic maintenance and extra cells are needed for protective custody, disciplinary cases, and other emergencies. Therefore, operating a prison at its highest capacity may already indicate a crowding situation. 47. Washington State Sentencing Guidelines Commission. A Decade of Sentencing Reform: Washington and Its Guidelines 1981-1991. Olympia, 1992. 48. Austin, J. Reforming Florida's Unjust, Costly and Ineffective Sentencing Laws. San Francisco: National Council on Crime and Delinquency. 1993. 49. Austin, J. "The Use of Early Release and Sentencing Guidelines to Ease Prison Crowding: The Shifting Sands of Reform." Paper presented at the National Academy of Sciences meeting on prison overcrowding, Chicago. 1986, pp. 45-46. 50. Dailey, D. "Minnesota: Sentencing Guidelines in a Politically Volatile Environment." Federal Sentencing Reporter 6(3)(1993), pp. 144-146. 51. Only inmates granted parole will benefit from the Safety Valve Program because the sentence discharge date is not affected. 52. Harer, M.D. Recidivism Among Federal Prison Releases in 1987: A Preliminary Report. Washington, DC: Federal Bureau of Prisons. 1994. See also Irwin, J., and J. Austin. It's About Time: America's Imprisonment Binge. Belmont, CA: Wadsworth Publishing. 1993. See also Note 35. 53. Note 52, p. 117. 54. Note 1. ------------------------------ Chapter 7 Summary The past decade has brought significant activity in sentencing reform. Most States have instituted reforms to increase the certainty of sentencing decisions by adopting mandatory sentencing provisions, more sweeping determinate sentencing, or sentencing guideline systems (either voluntary/advisory or presumptive). The impetus for many of these reforms has been the need to control sentencing disparity, increase truth in sentencing, and control correctional population growth. These reforms enable States to establish an overall policy on sentencing goals and to limit their investment in incarceration facilities. In view of the trend toward sentencing commissions over the past 15 years and the projected growth of prison populations, continued interest in structured sentencing models is likely. This report summarized the nature and results of this reform movement. This closing chapter summarizes the major findings and policy implications for how structured sentencing reforms may proceed in the future. Major Findings Current Sentencing Practices o An unprecedented flurry of structured sentencing reforms has occurred over the past two decades. The Federal Government and 16 States have implemented or are implementing either presumptive or voluntary/advisory sentencing guidelines. Another five States have adopted determinate sentencing systems. o Another popular form of structured sentencing is mandatory minimum sentencing laws. All States have some version of these laws, which target habitual offenders ("three strikes and you're out") and the crimes of possession of a deadly weapon ("use a gun--go to prison"), drunk driving, and possession and distribution of drugs. o Most States continue to allow inmates to earn good-time credits either to reduce the inmate's sentence or to advance the inmate's parole eligibility date. o Despite the level of criticism directed at parole, most States, including those that have adopted determinate and sentencing guideline models, retain some form of discretionary release and postrelease supervision. o Although several States have converted to guideline-based sentencing, most States do not use sentencing guidelines. Furthermore, five States have tried and failed to adopt sentencing guidelines. Sentencing Guideline Models The purposes, structures, requirements, and impacts of sentencing guidelines are summarized in the following paragraphs: Purposes and Goals. o Virtually all guideline commissions were asked to fulfill the multiple goals of punishment ("just deserts"), deterrence, incapacitation, and rehabilitation. o Few jurisdictions explicitly stated the goal of establishing sentencing neutrality with respect to race, gender, or social or economic status. o Only a few guideline commissions were required to consider the impact of the guidelines on the need for future correctional resources (i.e., prison beds). Structure of Sentencing Guidelines and Sentencing Commissions. o The structure of sentencing guidelines varies dramatically in the criteria used to determine sentencing disposition and sentence length. o The most common format for guidelines is a matrix that relies on two variables: offense severity and prior criminal history. o The number and type of members selected for the sentencing commission vary considerably, with no apparent advantage to either a large or small number of members. o In general, membership includes judges, prosecutors, defense attorneys, and private citizens. Corrections officials, legislators, and law enforcement officials are also often members of these commissions. Implementation Issues. o The process for implementing guidelines is neither quick nor inexpensive, often requiring 2 years to complete. o Before guidelines are implemented, detailed data on current sentencing patterns need to be collected and analyzed. A jurisdiction will also need a simulation model to estimate the impact of the proposed guidelines on prison, parole, probation, and jail populations. o The annual budget for guideline commissions and support staff ranges from $250,000 to $500,000. Considerable support is required to monitor and analyze compliance with the guidelines and to project the impact of modifying them. Impact of Sentencing Guidelines. o States that have implemented presumptive guidelines have reported high compliance rates and have succeeded in changing historical sentencing trends. o In general, offenders convicted of violent crimes or repeat offenders are far more likely to be imprisoned and to serve longer prison terms under sentencing guidelines. Conversely, first-time offenders charged with property crimes are less likely to be imprisoned and generally serve shorter prison terms. o Partly because of guidelines, persons convicted of drug crimes (both possession and sale) are now far more likely to be imprisoned and to serve lengthy prison terms. This trend has directly increased the rate of imprisonment for African-American and Hispanic offenders. o Guidelines have helped reduce sentencing disparity, but disparity reductions have eroded somewhat over time. o Limited evidence indicates that guidelines designed to take correctional resources into account have lowered rates of growth in incarceration and have helped to control prison crowding. However, prison crowding remains a problem for most States. o Guideline States, like nonguideline States, will probably have overcrowded prisons over the next decade unless they embark on substantial prison construction programs or reduce prison terms for violent and drug offenders. Policy Implications Reducing Disparity While Maintaining Discretion A jurisdiction's decisions about its criminal justice laws will have a major impact on the quality of justice and the costs to its citizens. Unless structured by law or a commission, these decisions are left to individual judges who must determine not only what they perceive as just sentences, but also the best use of State and local correctional resources. The major question is whether the technology of structured sentencing, and presumptive sentencing guidelines in particular, can overcome well-established organizational values that may facilitate and protect inequitable sentencing practices. One area of concern is the ability to individualize sentencing and consider a wide range of sentencing purposes while maintaining an equitable sentencing system. Most States have attempted to control the individualization of judicial practices while also maintaining that sentences should consider the full range of traditional sentencing purposes. Pennsylvania and Washington provide the most explicit examples of this effort to structure sentencing while maintaining adequate court discretion to fashion sentences to individual offenders. Pennsylvania requires judges to consider offenders' rehabilitative potential and community protection as well as the guidelines. In Washington, the enabling legislation mandates that the guidelines incorporate multiple goals of retribution, incapacitation, rehabilitation, and frugal use of correctional resources. Recent innovations in Louisiana and Oregon have expanded the range of sentencing options and allowed more commensurate punishments. This new approach, which illustrates that sentences can be very different yet equal, is an important move in the evolution of sentencing guidelines. Building on these innovations, other commissions may develop guidelines that will increase fairness, equity, and proportionality and still provide for a full range of sentencing options. Displacement of Discretion As noted earlier, one of the key issues involved in the attempts to control sentencing discretion is the displacement of discretion from the court to the prosecutor. One must be concerned that guidelines have merely shifted discretion from parole boards, prison officials, and judges to prosecutors. Unfortunately, little systematic evidence exists to document the extent to which this has occurred. Clearly, research is needed in this area of sentencing reform. Prison Crowding To date structured sentencing reforms have not demonstrated any appreciable effects on prison crowding. This is not to say that they could not have such an impact in the future. However, until the legislators and sentencing commissions become immune to "get tough on crime" pressures there is little reason to believe that structured sentencing models will solve the prison crowding problem. Moreover, as State prisons remain crowded, they will continue to use discretionary early release programs. Depending upon how such programs are structured, they may interfere with the goal of reducing disparity in time to be served by similarly situated offenders. Recommendations (Note: the recommendations in this section are those of the Advisory Board.) The purpose of this project was not to advocate one particular form of sentencing policy, whether it be presumptive guidelines, voluntary/advisory guidelines, determinate sentencing, or indeterminate sentencing. Any number of sentencing reform measures can reduce disparity, incarceration rates, and prison crowding. The question is how best to achieve these goals. The most promising structured sentencing model to address these issues is sentencing guidelines developed by sentencing commissions. Sentencing guidelines have helped States set overall policies as to the goals of sentencing and, in some instances, to set limits on what they are willing to invest in prisons. In view of the strong move toward sentencing commissions and the continued growth of prison populations, it is likely that there will be continued interest in the use of sentencing guidelines. Despite this endorsement of the guideline approach, a note of caution is warranted. This report has provided examples where guidelines did not improve the quality of justice. Simply stated, guidelines may or may not work depending on how they are designed and implemented, and for what purposes. Importantly, one should not expect guidelines to solve all of the problems regarding the dispensation of justice and the reduction of crime. The following policy recommendations from the Advisory Board are offered for States to consider in developing structured sentencing policies: o Commission representation participation. Broad participation in the guideline process is desired as it cultivates commitment to the promulgated guidelines and improves their quality. o Resources. Adequate financial support is necessary if a commission is to study past sentencing practices in the State, study what other States have done, prepare reports and proposals for the commission to consider, pay for the commission to meet regularly, hire consultants when necessary, provide ongoing feedback to the court and other governmental agencies, and monitor and evaluate the impact of guidelines. o Ongoing monitoring and support. Commissions that have successfully implemented guidelines have realized increases in budget and in staff size in the postimplementation years. This is due to continually increasing responsibilities imposed on the commissions by the legislature and the cost of maintaining and evaluating guidelines. o Appellate review. Without an enforcement mechanism, guidelines are merely voluntary/advisory and as such may have little impact on changing sentencing practices. The common procedure is to provide for appellate review and for the appellate review to be initiated by either the defense or prosecution. o Mandatory minimum sentences. It is clear from the experiences of many States that the increased use of mandatory minimum penalties is interfering with achievement of the dual goals of reducing disparity and controlling correctional population growth. A State should resist such provisions if they affect large proportions of its sentenced population. o Controlling the use of departures. The extent to which subjective factors are used to depart from the guidelines will have much to do with meeting the goal of reducing unwarranted disparity. States should specify as much as possible the type of departures that are acceptable. Moreover, commissions should carefully monitor the frequency and direction of departures by, among other things, crime category, race, and gender. o Role of the Federal Government. States that have recently adopted or are interested in adopting guidelines require assistance and guidance. In particular, the Federal Government can play an important role by providing a modest amount of assistance to States by the following actions: o Establish and maintain a national clearinghouse on structured sentencing. o Convene annual meetings of State sentencing commissions to share information and research findings. o Establish a funding program to assist States in technologies necessary to support the development or modification of existing sentencing structures. Assistance is especially required in the following areas: o Conducting comprehensive studies of existing sentencing practices that can be used to guide States in the formulation of new sentencing structures. o Developing new methods or providing training to States in existing analytic methods for measuring disparity. o Designing information systems that can be used to monitor compliance with sentencing criteria. o Developing and improving criminal court and correctional population simulation technology (jails, prisons, probation, and parole). o Developing and improving methods of designing and pilot-testing sentencing guideline methods. Finally, more research is needed to assess whether guidelines and other forms of structured sentencing are reducing sentencing disparity. As indicated above, a number of States have implemented, or are about to implement, sentencing guidelines. Independent process evaluations and impact evaluations of these new structured sentencing reforms would be valuable to the field. A better understanding is also needed concerning the effect of reforms external to the guidelines on sentencing disparity, especially along racial lines. Topics to be addressed include the impact of mandatory minimum provisions for certain drug crimes on disparity and the effect of guidelines on shifting discretion from the courts to the front end of the system (arrest, charging, and plea bargaining). Such studies will help clarify how best to correct undesirable and unequal sentencing practices. ------------------------------ Sources for Further Information For more information on Structured Sentencing, contact: Bureau of Justice Assistance 633 Indiana Avenue NW. Washington, DC 20531 202-514-5947 Bureau of Justice Assistance Clearinghouse Box 6000 Rockville, MD 20849-6000 800-688-4252 National Council on Crime and Delinquency 1325 G Street, Suite 1020 Washington, DC 20005 202-638-3080 ------------------------------ Bibliography Allen, F.A. The Decline of the Rehabilitative Ideal: Penal Policy and Social Purpose. New Haven, CT: Yale University Press. 1981. Alschuler, A.W. "Departures and Plea Agreements Under the Sentencing Guidelines." Federal Rules Decision 117(1988), pp. 459-476. Alschuler, A.W. "Sentencing Reform and Prosecutional Power: A Critique of Recent Proposals for `Fixed' and Presumptive Sentencing." University of Pennsylvania Law Review 126(1978), pp. 550-577. American Bar Association. Standards for Criminal Justice--Sentencing Alternatives and Procedures. 3d ed. Boston: Little, Brown & Company. 1993. American Friends Service Committee. Struggle for Justice: A Report on Crime and Punishment in America. New York: Wang. 1971. Ashford, K., and C. Mosbaek. First Year Report on Implementation of Sentencing Guidelines: November 1989 to January 1991. Portland: Oregon Criminal Justice Council. 1991. Ashworth, A.J. "Sentencing Reform Structures." In Crime and Justice: A Review of Research, ed. M. Tonry, pp. 188-244. Chicago: University of Chicago Press. 1992. Austin, J. Reforming Florida's Unjust, Costly and Ineffective Sentencing Laws. San Francisco: National Council on Crime and Delinquency. 1993. Austin, J. "The Use of Early Release and Sentencing Guidelines to Ease Prison Crowding: The Shifting Sands of Reform." Paper presented at the National Academy of Sciences meeting on prison overcrowding, Chicago. 1986. Berk, R., and A. Campbell. "Preliminary Data on Race and Crack Charging Practices in Los Angeles." Federal Sentencing Reporter 6(1)(1993), pp. 36-38. Blumstein, A., J. Cohen, S.E. Martin, and M.H. Tonry (eds). Research on Sentencing: The Search for Reform. Washington, DC: National Academy Press. 1983. Bogan, K.M. "Constructing Felony Sentencing Guidelines in an Already Crowded State: Oregon Breaks New Ground." Crime and Delinquency 36(4)(1990), pp. 467-487. Bureau of Justice Statistics. National Corrections Reporting Program, 1990. Washington, DC: U.S. Department of Justice, Office of Justice Programs. 1993. Bureau of Justice Statistics. Prisoners. Washington, DC: U.S. Department of Justice, Office of Justice Programs. 1984-1993. Bureau of Justice Statistics. Sourcebook of Criminal Justice Statistics--1992. NCJ-143496. Washington, DC: U.S. Department of Justice, Office of Justice Programs. 1993. Canadian Sentencing Commission. "Sentencing Reform: A Canadian Approach." Canada: Minister of Supply and Services. 1987. Carrow, D.M., J. Feins, B.N.W. Lee, and L. Olinger. Guidelines Without Force: An Evaluation of the Multijurisdictional Sentencing Guidelines Field Test. Cambridge: Abt. 1985. Casper, J., and D. Brereton. "Evaluating Criminal Justice Reforms." Law and Society 18(1984), pp. 121-144. Cirillo, V. "Windows of Discretion in the Pennsylvania Sentencing Guidelines." Villanova Law Review 31(1986), pp. 1309-1349. Cohen, J., and J. Helland. "Methodology for Evaluating the Impact of Sentencing Guidelines." Unpublished paper. Pittsburgh: Urban Systems Institute, School of Urban and Public Affairs, Carnegie-Mellon University. 1982. Cohen, J., and M.H. Tonry. "Sentencing Reforms and Their Impacts." In Research on Sentencing: The Search for Reform, ed. A. Blumstein, J. Cohen, S.E. Martin, and M.H. Tonry, pp. 305-459. Washington, DC: National Academy Press. 1983. Dailey, D. "Minnesota: Sentencing Guidelines in a Politically Volatile Environment." Federal Sentencing Reporter 6(3)(1993), pp. 144-146. Del Sole, J. "Appellate Review in a Sentencing Guidelines Jurisdiction: The Pennsylvania Experience." Duquesne Law Review 31(1993), pp. 419-504. Dershowitz, A. "Criminal Sentencing in the United States: A Historical and Conceptual Overview." Annals of the American Academy of Political and Social Science 423(1976), pp. 117-132. Dillingham, S.D. "Remarks: The Attorney General's Summit on Law Enforcement Responses to Violent Crime: Public Safety in the Nineties." Presentation, Washington, D, 1991. Eisenstein, J., R.B. Flemming, and P.F. Nardulli. Contours of Justice: Communities and Their Courts. Boston: Little, Brown & Company. 1988. Eisenstein, J., and H. Jacob. Felony Justice: An Organizational Approach to Criminal Courts. Boston: Little, Brown & Company. 1977. Fallen, D.L. Sentencing Practices Under the Sentencing Reform Act. Olympia: Washington State Sentencing Guidelines Commission. 1987. Fallen, D.L. Preliminary Evaluations of Washington's Sentencing Guidelines. Olympia: Washington State Sentencing Guidelines Commission. 1986. Federal Bureau of Investigation. Uniform Crime Reports for the United States--1992. Washington, DC: U.S. Department of Justice. 1993. Frankel, M.E. Criminal Sentences: Law Without Order. New York: Hill & Wang. 1972. Frankel, M.E. Struggle for Justice. New York: Hill & Wang. 1971. Frase, R.S. "Implementing Commission-Based Sentencing Guidelines: The Lessons of the First Ten Years in Minnesota." Wake Forest Law Review. 1993. Frase, R.S. "Sentencing Reform in Minnesota, Ten Years After. Reflections on Dale G. Parent's Structuring Criminal Sentences: The Evolution of Minnesota's Sentencing Guidelines." Minnesota Law Review 75(1991), p. 727. Gebelein, R.S. "Sentencing Reform in Delaware." Overcrowded Times 2(2)(1991), pp. 5, 12-13. Goodstein, L., J.H. Kramer, and L. Nuss. "Defining Determinacy: Components of the Sentencing Process Ensuring Equity and Release Certainty." Justice Quarterly 1(1)(1984), pp. 47-73. Gottfredson, M., and D. Gottfredson. "Guidelines for Incarceration Decisions: A Partisan Review." University of Illinois Law Review 2(1984), pp. 291-317. Gottlieb, D.J. "Kansas Adopts Sentencing Guidelines." Overcrowded Times: Solving The Prison Problem 4(3)(1993), pp. 1, 10-13. Griset, P. Determinate Sentencing: The Promise and the Reality of Retributive Justice. Albany: State University of New York Press. 1991. Griswold, D. "Deviation From Sentencing Guidelines: The Issue of Unwarranted Disparity." Journal of Criminal Justice 15(1987), pp. 317-329. Harer, M.D. Recidivism Among Federal Prison Releases in 1987: A Preliminary Report. Washington DC: Federal Bureau of Prisons. 1994. Heaney, G.W. "The Reality of Guidelines Sentencing: No End to Disparity." American Criminal Law Review 28(2)(1991), pp. 161-232. Irwin, J., and J. Austin. It's About Time: America's Imprisonment Binge. Belmont, CA: Wadsworth Publishing. 1993. Knapp, K. "Allocations of Discretion and Accountability Within Sentencing Structures." Colorado Law Review 64(1993), pp. 679-705. Knapp, K. "Implementation of the Minnesota Guidelines: Can the Innovative Spirit Be Preserved?" In The Sentencing Commission and its Guidelines, ed. A. von Hirsch, K. Knapp, and M. Tonry. Boston: Northeastern University Press. 1987. Kramer, J.H., and R.L. Lubitz. "Pennsylvania's Sentencing Reform: The Impact of Sentencing Commission-Established Guidelines." Crime and Delinquency 31(1985), pp. 481-500. Kramer, J.H., R.L. Lubitz, and C.A. Kempinen. "Sentencing Guidelines: A Quantitative Comparison of Sentencing Policy in Minnesota, Pennsylvania and Washington." Paper presented at the annual meeting of the American Society of Criminology, San Diego, November. 1989. Kramer, J.H., and J. McCloskey. "The Role of Political Pressure in Developing Sentencing Guidelines for Drug Offenses." Paper presented at the 1988 annual meeting of the Academy of Criminal Justice Sciences, San Francisco. 1988. Kramer, J.H., and A.J. Scirica. "Complex Policy Choices: The Pennsylvania Commission on Sentencing." Paper presented at the annual meeting of the Academy of Criminal Justice Sciences, Las Vegas, April 1985. Kramer, J.H., and D. Steffensmeier. "Race and Imprisonment Decisions." The Sociological Quarterly 34(2)(1993), pp. 357-376. Kramer, J.H., and J. Ulmer. "Sentencing Disparity and Guideline Departures." Criminology, in review. Kress, J.M. Prescription for Justice: The Theory and Practice of Sentencing Guidelines. Cambridge: Ballinger. 1980. Lagoy, S.P., F.A. Hussey, and J.H. Kramer. "A Comparative Assessment of Determinate Sentencing in the Four Pioneer States." Crime and Delinquency 24(1978), pp. 385-400. Lieb, R. Personal interview. 1993. Lieb, R. "Washington Prison Population Growth Out of Control." Overcrowded Times 4(1)(1993), pp. 1, 13-14, 20. Lubitz, R. "North Carolina Legislature Considers Sentencing Change." Overcrowded Times 4(2)(1993), pp. 1, 9-10. Martin, S. "Interests and Politics in Sentencing Reform: The Development of Sentencing Guidelines in Pennsylvania and Minnesota." Villanova Law Review 29(1984), pp. 21-113. Martin, S. "The Politics of Sentencing Reform: Sentencing Guidelines in Pennsylvania and Minnesota." In Research and Sentencing: The Search for Reform, vol. 2, ed. A. Blumstein, J. Cohen, S. Martin, and M. Tonry. Washington, DC: National Academy Press. 1983. McCloskey, J. "The Effectiveness of Independent Sentencing Commission Guidelines: An Analysis of Appellate Court Decisions in Two Jurisdictions." Paper presented at the annual meeting of the American Society of Criminology, San Diego. 1985. McDonald, D, and K.E. Carlson. "Why Did Racial/Ethnic Sentencing Differences in Federal District Courts Grow Larger Under the Guidelines?" Federal Sentencing Reporter 6(4)(1994), pp. 223- 226. Messinger, S.L., and P.D. Johnson. California's Determinate Sentence Stature. History and Issues in Determinate Sentencing: Reform or Regression? Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. 1978. Miethe, T.D. "Charging and Plea Bargaining Practices Under Determinate Sentencing: An Investigation of the Hydraulic Displacement of Discretion." Journal of Criminal Law and Criminology. 78(1)(1987), pp. 155-176. Miethe, T.D., and C.A. Moore. Sentencing Guidelines: Their Effect in Minnesota. Washington, DC: U.S. Department of Justice, Office of Justice Programs, National Institute of Justice. April 1989. Miethe, T.D., and C.A. Moore. "Can Sentencing Reform Work? A Four-Year Evaluation of Determinate Sentencing in Minnesota." Paper presented at the annual meeting of the American Society of Criminology, November 1988. Miethe, T.D., and C.A. Moore. "Officials' Reactions to Sentencing Guidelines." Journal of Research in Crime and Delinquency 25(2)(1988), pp. 170-187. Miethe, T.D., and C.A. Moore. "Racial Differences in Criminal Processing: The Consequences of Model Selection on Conclusions About Differential Treatment." Sociology Quarterly 27(1986), pp. 217- 237. Miethe, T.D., and C.A. Moore. "Socioeconomic Disparities Under Determinate Sentencing Systems: A Comparison of Pre- and Post-Guideline Practices in Minnesota." Criminology 23(1985), pp. 337-363. Mistretta v. United States 488 U.S. 361 (1989). Moore, C.A., and T.D. Miethe. "Regulated and Non-Regulated Sentencing Decisions: An Analysis of First-Year Practices Under Minnesota's Felony Sentencing Guidelines." Law and Society Review 20(2)(1986), pp. 253-277. Morris, N. The Future of Imprisonment. Chicago: University of Chicago Press. 1974. Morris, N., and M. Tonry. Between Prison and Probation: Intermediate Punishments in a Rational Sentencing System. Oxford: Oxford University Press. 1990. Mosbaek, C. Third Year Report on Implementation of Sentencing Guidelines: 1992. Portland: Oregon Criminal Justice Council. 1993. Myers, S.L., Jr. "Racial Disparities in Sentencing: Can Sentencing Reforms Reduce Discrimination in Punishment?" University of Colorado Law Review 64(3)(1993), pp. 781-808. Nardulli, P., R. Flemming, and J. Eisenstein. The Tenor of Justice: Criminal Courts and the Guilty Plea Process. Chicago: University of Illinois Press. 1988. Nelson, Blake. "The Minnesota Sentencing Guidelines: The Effects of Determinate Sentencing on Disparities in Sentencing Decisions." Law and Inequality 10(3)(1992), p. 217. North Carolina Sentencing and Policy Advisory Commission. Report of the 1993 Session for the General Assembly of North Carolina. Raleigh, 1993. Olson, G. "Forecasting the Long-Term Impact of Washington's Sentencing Guidelines: A Roller Coaster Ride." Crime and Delinquency 38(3)(1992), pp. 330-356. Parent, D.G. "What Did the United States Sentencing Commission Miss?" Yale Law Journal 101 (1992), p. 1773. Parent, D.G. Structuring Criminal Sentences: The Evolution of Minnesota's Sentencing Guidelines. Stoneham, MA: Butterworth. 1988. Pennsylvania Commission on Sentencing. Sentencing in Pennsylvania: Annual Report. State College, 1994. Pennsylvania Commission on Sentencing. Sentencing in Pennsylvania: Annual Report. University Park, 1991. Pincoffs, E.L. Philosophy of Law. Belmont, CA: Wadsworth. 1990. Quinn, T.J. "Voluntary Guidelines Effective in Delaware." Overcrowded Times 3(1)(1992), pp. 1, 9, 11. Quinn, T.J. "Delaware Sentencing Guidelines Achieving Their Goals." Overcrowded Times 1(3)(1990), pp. 1-2. Reitz, K.R. "Sentencing Facts: Travesties of Real-Offense Sentencing." Stanford Law Review 45(1993), pp. 523-573. Rich, W.D., L.P. Sutton., T.R. Clear., and M.J. Saks. Sentencing by Mathematics: An Evaluation of the Early Attempts To Develop Sentencing Guidelines. Williamsburg, VA: National Center for State Courts. 1982. Robinson, P.H. "Hybrid Principles for the Distribution of Criminal Sanctions." Northwestern Law Review 82(1987), pp. 19-42. Ross, H.L. Confronting Drunk Driving: Social Policy for Saving Lives. New Haven, CT: Yale University Press. 1992. Ross, H.L. Deterring the Drunk Driver. Lexington, MA: Lexington Books. 1984. Savelsberg, J. "Law That Does Not Fit Society: Sentencing Guidelines as a Neoclassical Reaction to the Dilemmas of Substantivized Law." American Journal of Sociology 97(s)(1992), pp. 1346-1381. Shane-DuBow, S., A.P. Brown, and E. Olsen. Sentencing Reform in the U.S.: History, Content and Effect. Washington, DC: U.S. Department of Justice. 1985. Shane-DuBow, S. Personal communication. 1993. Singer, R.G. Just Deserts: Sentencing Based on Equality and Desert. Cambridge, MA: Ballinger. 1979. Sparks, R.F., B.A. Stecher, J. Albanese, and P.L. Shelly. Stumbling Toward Justice: Some Overlooked Research and Policy Questions About Statewide Sentencing Guidelines. Report to the National Institute of Justice, U.S. Department of Justice. Washington, DC: U.S. Government Printing Office. 1982. Spohn, C., and J. Horney. Rape Law Reform: A Grassroots Revolution and Its Impact. New York: Plenum. 1992. Steffensmeier, D., J. Kramer, and C. Streifel. "Gender and Imprisonment Decisions." Criminology 31(3)(1993), pp. 411-446. Stolzenberg, L., and S.J. D'Alessio. "Sentencing and Unwarranted Disparity: An Empirical Assessment of the Long-Term Impact of Sentencing Guidelines in Minnesota." Criminology 32(2) (1994), pp. 301- 310. Tonry, M. "Malign Neglect--Race, Crime and Punishment in America." Unpublished manuscript. Tonry, M. "Sentencing Commissions and Their Guidelines." In Crime and Justice: A Review of Research, vol. 17, ed. M. Tonry, pp. 137-196. Chicago: University of Chicago Press. 1993. Tonry, M. "Mandatory Penalties." In Crime and Justice: A Review of Research, vol. 16, ed. M. Tonry. Boston: Northeastern University Press. 1992. Tonry, M. "The Politics and Processes of Sentencing Commissions." Crime and Delinquency 37(3)(1991), pp. 307-329. Tonry, M. "Structured Sentencing." Crime and Justice: A Review of Research, vol. 10, ed. M. Tonry and N. Morris. Chicago: University of Chicago Press. 1988. Tonry, M. "Sentencing Guidelines and Their Effects." In The Sentencing Commission and Its Guidelines, ed. A. von Hirsch, K. Knapp, and M. Tonry. Boston: Northeastern University Press. 1987. Twentieth Century Fund Task Force on Criminal Sentencing. Fair and Certain Punishment. New York: McGraw-Hill. 1976. Ulmer, J. Trial Court Communities Under Sentencing Guidelines: Sentencing, Case Processing, and Organizational Relations. Ph.D. dissertation, The Pennsylvania State University. Ann Arbor, MI: University Microfilms International. 1993. Ulmer, J. "Virginia's Sentencing Guidelines." Report prepared for the Structured Sentencing Project. 1993. Ulmer, J., and J.H. Kramer. "Sentencing Guidelines and Case Processing in Two Court Communities." Paper presented at the annual meeting of the Midwest Sociological Society, Kansas City, MO. 1992. U.S. General Accounting Office. Sentencing Guidelines--Central Questions Remain Unanswered. Washington, DC, 1992. U.S. Sentencing Commission. Annual Report--1991. Washington, DC, 1992. 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, 1991. U.S. Sentencing Commission. Special Report to the Congress: Mandatory Minimum Penalties in the Federal Criminal Justice System. Washington, DC, 1991. Van den Haag, E. Punishing Criminals: Concerning a Very Old and Painful Question. New York: Basic Books. 1975. von Hirsch, A. "The Politics of Just Deserts." Canada Journal of Criminology 32(1990), p. 397. von Hirsch, A. "Principles for Choosing Sanctions: Sweden's Proposed Sentencing Statute." New England Journal on Criminal and Civil Confinement 13(2)(1987), pp. 171-195. von Hirsch, A. "Equality, Anisonomy, and Justice: A Review of Madness and the Criminal Law." Michigan Law Review 82(4)(1984), pp. 1093-1112. von Hirsch, A. Doing Justice: The Choice of Punishments. New York: Hill & Wang. 1976. von Hirsch, A., and K. Hanrahan. "Determinate Penalty Systems in America: An Overview." Crime and Delinquency 27(1981), pp. 289-316. von Hirsch, A., K. Knapp, and M. Tonry (eds.). The Sentencing Commission and Its Guidelines. Boston: Northeastern University Press. 1987. Walker, S. Taming the System: The Control of Discretion in Criminal Justice 1950-1990. Oxford: Oxford University Press. 1993. Washington State Sentencing Guidelines Commission. A Decade of Sentencing Reform: Washington and Its Guidelines 1981-1991. Olympia, 1992. Wilkins, L.T., J.M. Kress, D.M. Gottfredson, J Calpin, and A.M. Gelman. Sentencing Guidelines: Structuring Judicial Discretion--Report on Feasibility Study. Washington, DC: U.S. Department of Justice. 1978. ------------------------------ Appendix A Information on States Used for Impact Analysis Sentencing guideline States selected for analysis and the years of pre- and postcomparison were as follows: Minnesota, 1980; Pennsylvania, 1982; Florida, 1983; Washington State, 1984. Although Washington's law was adopted in 1983, it was not fully implemented until 1984. The nonguideline States selected as comparison for each of the four guideline States (for each of three categories) were as follows: Violent Crime Rates Minnesota: Maine, Montana, Nebraska, Idaho Pennsylvania: Connecticut, Indiana, Virginia, Arkansas, Mississippi, Rhode Island Florida: Nevada, California, Maryland Washington: Connecticut, Rhode Island, Oklahoma, Tennessee, Ohio, Alabama Total Crime Rates Minnesota: Virginia, Montana, Wyoming, Idaho, Oklahoma, Indiana, Louisiana, Kansas Pennsylvania: Arkansas, Nebraska, New Hampshire, Kentucky Florida: California, Arizona, Nevada, Colorado Washington: Delaware, New York, Alaska, Oregon, Hawaii Incarceration Rates Minnesota: Virginia, Montana, Wyoming, Idaho, Oklahoma, Indiana, Louisiana, Kansas Pennsylvania: Utah, Vermont, Maine, Rhode Island, West Virginia, Iowa, Connecticut, South Dakota Florida: Maryland, Nevada, South Carolina, Georgia, Texas, North Carolina Washington: California, Kentucky, Mississippi, New York, Oregon, Ohio, Alaska