Fifteen Years of Guidelines Sentencing: An Assessment of How Well the Federal Criminal Justice System Is Achieving the Goals of Sentencing Reform
From The United States Sentencing Commission; November 2004 (The full study is available in PDF Format at: www.ussc.gov/15_year/15year.htm)
The Sentencing Reform Act of 1984 [hereinafter the SRA] ushered in a new era of sentencing in federal courts. Prior to implementation of the SRA, federal crimes carried very broad ranges of penalties, and federal judges had the discretion to choose the sentence they felt would be most appropriate. They were not required to explain their reasons for the sentence imposed, and the sentences were largely immune from appeal. The time actually served by most offenders was determined by the Parole Commission, and offenders, on average, served just 58 percent of the sentences that had been imposed. The sentencing process, a critical element of the criminal justice process, was opaque, undocumented, and largely discretionary. Because of its impenetrability to outside observers, there was a sense that the process was unfair, disparate, and ineffective for controlling crime.
In order to inject transparency, consistency, and fairness into the sentencing process, Congress passed the SRA, which established the United States Sentencing Commission [hereinafter the Commission] and charged it with establishing guidelines for federal sentencing. The guidelines were promulgated in 1987, but district and circuit court rulings prevented their full implementation until the Supreme Court, in Mistretta v. United States, 488 U.S. 361 (1989), affirmed the constitutionality of the Commission and its work in crafting guidelines. As a result, in 1991, when the Commission issued its report, 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 [hereinafter called the Four-Year evaluation], there was relatively little data from which the Commission could evaluate the effects of the guidelines. Today the Commission is in a better position to evaluate the success of the guidelines system and identify areas for further refinement. This report focuses on three specific assessments:
1) the guidelines' impact on the transparency, certainty, and severity of punishment, 2) the impact of the guidelines on inter-judge and regional disparity, and 3) research on racial, ethnic, and gender disparities in sentencing today.
Introduction To The Sentencing Reform Act And The Guidelines Goals And Evaluation Criteria
The SRA was the result of nine years of bipartisan deliberation and compromise and, as such, reflects the varied and, at times, competing sentencing philosophies of its many sponsors and supporters. It set forward the following goals for sentencing reform:
1. elimination of unwarranted disparity;
2. transparency, certainty, and fairness;
3. proportionate punishment; and
4. crime control through deterrence, incapacitation, and the rehabilitation of offenders.
The goals of the new system identified in the SRA provide the best criteria for judging whether sentencing reform has been successful. These goals can be divided into two groups. The first group, the goals of sentencing reform, include certainty and fairness in punishment and the elimination of unwarranted disparity. Research on the effectiveness of the system at achieving these goals is the subject of this report. The second group, establishment of policies that will best accomplish the purposes of sentencing -- which are usually summarized as just punishment, deterrence, incapacitation, and rehabilitation -- is the subject of previous Commission-sponsored research as well as ongoing research at the Commission.
Development Of The Guidelines
The guidelines promulgated by the Commission were based on the directives in the SRA and other statutory provisions, as well as on a study of past sentencing practices. The Commission analyzed detailed data from 10,000 presentence reports and additional data on over 100,000 federal sentences imposed in the immediate preguidelines era. The Commission determined the average prison term likely to be served for each generic type of crime.
These averages helped establish "base offense levels" for each crime, which were directly linked to a recommended imprisonment range. Aggravating and mitigating factors that significantly correlated with increases or decreases in sentences were also determined statistically, along with each factor's magnitude. These formed the bases for "specific offense characteristics" for each type of crime, which adjusted the offense level upward or downward. The Commission deviated from past practice when it determined there was a compelling reason, such as past under-punishment of white collar offenses, and when Congress dictated increased severity for an offense category. The Commission also factored offenders' criminal history into the guidelines as a way to identify offenders most likely to recidivate.
Real Offense Guidelines
The statute-defined elements of many federal crimes fail to provide sufficient detail about the manner in which the crime was committed to permit individualized sentences that reflect the varying seriousness of different violations. In addition, the many, sometimes overlapping provisions in the federal criminal code create the potential that similar offenses will be charged in many different ways. To better reflect the seriousness of each offender's actual criminal conduct, and to prevent disparate charging practices from leading to sentencing disparity, the original Commission developed guidelines that are based to great extent on offenders' real offense behavior rather than the charges of conviction alone. Some of the mechanisms to help ameliorate the effects of uneven charging include: 1) the multiple count rules, 2) cross-references among guidelines, and 3) the relevant conduct rule. In a real offense system, the offender's actual conduct proved at the sentencing hearing -- not only the elements of the counts of conviction -- factor into the sentence imposed within the statutory penalty range established by the legislature for the offenses of conviction.
Certainty and Severity of Punishment
Truth-in-sentencing, mandatory minimums, and sentencing guidelines. In some sense, the success of the guidelines at achieving certainty of punishment has never been at issue, because the establishment of "truth-in-sentencing" with the elimination of parole accomplished it at a stroke.
Under the guidelines, punishment became not only more certain but also more severe. The proportion of probation sentences declined, use of restrictive alternatives such as home confinement increased, and the rate of imprisonment for longer lengths of time climbed dramatically compared to the preguidelines era. While mandatory minimum penalties had some direct and indirect effects on these trends, careful analysis of sentencing trends for different types of crimes demonstrates that the sentencing guidelines themselves made a substantial and independent contribution.
Overall trends in the use of imprisonment and probation. Between 1987 and 1991, as the full impact of the sentencing guidelines gradually emerged in federal courts, the use of simple probation was cut almost in half. It continued to decline throughout the guidelines era. By 2002, the percentage of offenders receiving simple probation was just a third what it had been in 1987.
The use of imprisonment spiked in the early years of guidelines implementation and then resumed a long gradual climb, reaching 86 percent of all offenders by 2002, about 20 percent higher than it had been in the preguidelines era. Some of the decrease in the use of simple probation following implementation of the guidelines is explained by increased use of intermediate sanctions, especially for "white collar" crimes. These offenders historically were more likely to receive simple probation, but under the guidelines they increasingly are subject to intermediate sanctions, such as home or community confinement or weekends in prison, and imprisonment.
In addition to an increase in use of imprisonment, the guidelines era is marked by longer prison terms actually served. Longer prison terms result both from the abolition of parole, which requires offenders to serve at least 85 percent of the sentence imposed, and also by increases in the sentences that are imposed for many types of crimes. Between November 1987 and 1992, the average prison term served by federal felons more than doubled. Since fiscal year 1992, there has been a slight and gradual decline in average prison time served, but federal offenders sentenced in 2002 will still spend almost twice as long in prison as did offenders sentenced in 1984, increasing from just under 25 to almost 50 months in prison for the typical federal felon.
The abolition of parole, the enactment of mandatory minimum penalty provisions, and changes in the types of offenders sentenced in federal court, along with implementation of the guidelines, all contributed to increased sentence lengths. The influence of each of these factors varies among different offenses.
Drug trafficking offenses have comprised the largest portion of the federal criminal docket for over three decades. With the overall growth in the federal criminal caseload, the number of offenders convicted of drug trafficking or use of a communication facility to commit a drug offense has grown every year, reaching 25,835 offenders in 2002, or 40.4 percent of the total criminal docket. Only 592 additional drug offenders, less than 1 percent, were convicted of simple drug possession, as opposed to trafficking. As a result of the large number of drug offenders, overall trends in the use of incarceration and in average prison terms are dominated by drug sentencing.
In developing sentences for drug trafficking offenders, the Commission was heavily influenced by passage of the Anti-Drug Abuse Act of 1986 [hereinafter ADAA] which created five-and ten-year mandatory minimum penalties based on the weight of the "mixture or substance containing a detectable amount" of various types of drugs. Finding the correct quantity ratios among different drugs and the correct thresholds for each penalty level has proven problematic. The Commission previously reported that the ratios among certain types of drugs contained in the ADAA, and incorporated into the guidelines' Drug Quantity Table, fail in some cases to reflect the relative harmfulness of different drugs. This is particularly true for the 100-to-1 drug quantity ratio between powder and crack cocaine. The quantity thresholds linked to five- and ten-year sentences for crack cocaine have been shown to result in penalties that are disproportionately long given the relative harmfulness of crack and powder cocaine, and results in lengthy incarceration for many street-level sellers and other low culpability offenders. As a result, the Commission has recommended to Congress revision of the mandatory minimum penalty statutes and the guidelines. Congress has not yet acted on this recommendation.
There has been a dramatic increase in time served by federal drug offenders following implementation of the ADAA and the guidelines. The time served by federal drug traffickers was over two and a half times longer in 1991 than it had been in 1985, hovering just below an average of 80 months. In the latter half of the 1990s, the average prison term decreased by about 20 percent but remained far above the historic average. The decrease in time served during the late 1990s is a result of a trend toward less serious offenses and a greater incidence of mitigating factors in cases sentenced. The overall pattern is repeated for each drug type, although the severity levels are highest for crack cocaine, followed by powder cocaine, heroin, and other scheduled narcotics. Marijuana offenses received the shortest prison terms.
Economic offenses -- which include larceny, fraud, and non-fraud white collar offenses -- constitute the second largest part of the federal criminal docket. A wide variety of economic crimes are prosecuted and sentenced in the federal courts, ranging from large-scale corporate malfeasance to small-scale embezzlement to simple theft. The Commission's study of past sentencing practices revealed that in the preguidelines era, sentences for fraud, embezzlement, and tax evasion generally received less severe sentences than did crimes such as larceny or theft, even when the crimes involved similar monetary loss. A large proportion of fraud, embezzlement, and tax evasion offenders received simple probation. In response, the guidelines were written to reduce the availability of probation and to ensure a short but definite period of confinement for a larger percentage of these "white collar" cases, both to ensure proportionate punishment and to achieve adequate deterrence.
The most striking trend in economic offenses is a shift away from simple probation and toward intermediate sentences that include some type of confinement. The use of imprisonment for economic offenders also has increased steadily throughout the guidelines era. These data demonstrate some success in achieving the Commission's goal of assuring a "short but definite period of confinement" for white collar offenders. The guidelines ensure that offenses involving the greatest monetary loss, the use of more sophisticated methods, and other aggravating factors are given imprisonment.
Prior to fiscal year 1994 there were relatively few immigration cases sentenced in the federal courts. In the first three years of the 1990s the number of cases ranged between 1,000 and 2,000 annually. Beginning in 1995, however, the number of cases began to climb, and after the implementation of Operation Gatekeeper -- the Immigration and Naturalization Service's southwest border enforcement strategy -- the number began to soar, reaching a peak of just under 10,000 cases in 2000. Along with the phenomenal growth in the size of the immigration offense docket, a series of policy decisions by Congress and by the Commission have steadily increased the severity of punishment for the two most common classes of immigration offenses: alien smuggling and illegal entry.
Use of imprisonment has increased substantially for these offenses and is affected by the fact that many immigration offenders are non-resident aliens. Lacking a legal home in the United States, many are detained prior to sentencing. Immediate deportation has also become a frequent response to those individuals arrested for illegal entry. Legislative and Commission changes to these penalties have focused on increasing the guidelines offense levels. This has pushed more offenders into the zones of the Sentencing Tables in which probation and alternative sentences are unavailable. In addition to the increased use of incarceration, the average length of time served for both alien smuggling and illegal entry have increased considerably. Illegal entry offenders experienced the first wave of sentence increases in the early 1990s as the guideline amendments enacted in those years became effective. Alien smuggling experienced a steep increase in 1998, as the amendment promulgated pursuant to the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 took effect.
Firearm Trafficking And Possession
The federal criminal code contains a variety of provisions proscribing the possession, use, and trafficking of firearms. In the last two decades, congressional attention has focused on 18 U.S.C. § 924©, which provides for a mandatory minimum penalty for offenders who use, carry, or possess a firearm in relation to a drug trafficking or violent crime. In 1984, the statute was amended to require at least five years' imprisonment, to be served consecutive to the sentence for the underlying offense. In 1986, the statute's scope was expanded to include drug trafficking offenses, and additional penalties were added. In 1998, in response to Bailey v. U. S., 516 U.S. 137 (1995) -- a U. S. Supreme Court decision that narrowly construed the "use" criteria -- the statute's scope was again expanded to include "possession in furtherance" of the underlying offense. Penalties were also increased for brandishing or discharging a firearm during a crime.
Federal statutes also define two other broad types of firearm offenses. Federal law regulates transactions in firearms and imposes record-keeping and other requirements designed to facilitate control of firearm commerce by the various states. In addition, possession of a firearm by certain classes of persons, such as felons, fugitives, or addicts, is prohibited, as is "knowing transfer" of weapons to these persons. Under the guidelines, the certainty and severity of punishment for all these offenses have greatly increased.
For firearm traffickers, the use of probation has been steadily reduced to about one-quarter of its preguidelines level, replaced by imprisonment and, to a lesser extent, intermediate sanctions.
After a period of volatility and decline in trafficking sentences in the first years of guidelines implementation, time served began a steady climb in fiscal year 1992, after the Commission enacted a major revision to the firearms guideline. The subsequent amendments to the guideline have continued to increase sentence severity. By 2000, prison terms were about double what they had been in the preguidelines era. For illegal possessors, probation has been replaced almost completely by imprisonment. The penalty increases for possession offenses were equally dramatic, doubling average time served between 1988 and 1995.
Some of the changes observed for firearm offenses may have been a consequence of more serious cases generated by Department of Justice [hereinafter the Department] initiatives. But the most significant factor driving the penalty increases appears to have been the guideline amendments.
These revisions have dramatically increased offense levels, particularly for offenders with prior convictions and for those who used more dangerous types of weapons. These changes in sentences for illegal firearm transactions and possession represent one of the most substantial policy changes initiated largely by the Commission.
Unlike the state courts, the federal courts sentence relatively few offenders convicted of violent crimes. In 2002, murder, manslaughter, assault, Kidnapping, robbery, and arson constituted less than four percent of the total federal criminal docket. Due to the unique nature of federal jurisdiction over these types of crime, a sizeable proportion of murder, assault, and especially manslaughter cases involve Native American defendants. The most common federal violent crime is bank robbery, which has long been of special concern to federal law enforcement.
For most violent offenses, rates of imprisonment have always been high and have remained so under the guidelines. Only manslaughter, the violent offense for which Native Americans are most highly represented, contained room for significant growth in incarceration rates. The use of alternatives to imprisonment for manslaughter cases has been steadily reduced under the guidelines, and now occurs in less than ten percent of cases. Kidnapping and murder have incarceration rates between 90 and 100 percent, with arson and assault somewhat lower. The imprisonment rate for bank robbers climbed from the mid to the high 90s under the guidelines.
Average prison sentences imposed on violent offenders decreased at the time of guidelines implementation, but due to the abolition of parole, the time served increased significantly. The greatest increases have been for murder, Kidnapping, bank robbery, and arson. The more stable prison term lengths for manslaughter partly reflect the large number of these offenders who receive relatively short prison terms rather than an alternative sanction.
Although sex offenses account for a very small percentage of cases in the federal docket, just 1.3 percent in 2002, Congress has legislated frequently on this issue during the guidelines era, particularly regarding offenses against minors. Much like policymaking in the area of drug trafficking, Congress has used a mix of mandatory minimum penalty increases and directives to the Commission to change sentencing policy for sex offenses. In the PROTECT Act of 2003, Congress, for the first time since the inception of the guidelines, directly amended the Guidelines Manual and developed unique limitations on downward departures from the guidelines in sex cases.
The guidelines treat separately three types of sexual offenses. Criminal sexual abuse involves offenses such as aggravated rape, statutory rape, or molestation. Sexual exploitation involves the production, distribution, or possession of child pornography. Promotion offenses involve inducing, enticing, or persuading commission of an illegal sex act, or traveling or transporting persons to commit such acts, or otherwise promoting illegal commercial sex acts.
The percentage of offenders receiving imprisonment increased for both sexual abuse and sexual exploitation offenders in the guidelines era, and dramatically so for sexual exploitation offenders. Fewer than ten percent of either type of offender receives probation or intermediate sanctions. The average length of time served for sexual exploitation has increased by 20 months from its preguidelines level. Sentences imposed on sexual abuse offenders show the same decreases observed for violent offenders, but time actually served has remained fairly constant throughout the period of study.
Inter-judge and Regional Disparity
Evidence of disparity in preguidelines sentencing. In the debates leading to passage of the SRA, Congress identified differences among judges and, to a lesser extent, differences among geographic regions in sentencing practices as particularly common sources of unwarranted disparity. Research demonstrated that philosophical differences among judges affected the sentences they imposed. The data showed that some judges were consistently more severe or more lenient than their colleagues, and that judges varied in their approaches to particular crime types. Several studies found geographical variations in sentencing patterns, suggesting that different political climates or court cultures can affect sentences. Regional differences arise not just from the exercise of judicial discretion, but also from differences in policies among U. S. attorneys.
Increased Transparency And Predictability Of Sentences Under The Guidelines
The guidelines have made sentencing more transparent and predictable. The SRA requires judges to document in open court the facts and reasons underlying the sentences they impose, which are then reviewable on appeal. Defendants and prosecutors are better able to predict sentences based on the facts of the case than in the discretionary, preguidelines era. By making sentencing policies more transparent, the guidelines make it easier to debate and evaluate the merits of particular policies. The effects of changes in sentencing policy can also be anticipated more precisely. The prison impact model developed by the Sentencing Commission, and further elaborated by the Bureau of Prisons [BOP], has proven very accurate at projecting the need for prison beds and supervision resources, making management of correctional resources easier.
Statistics provide a method for quantifying the increased understanding of sentencing made possible by guidelines. Most of the "variance" -- the deviation of sentences around the average -- among sentences in the preguidelines era was unaccounted for in statistical studies. Only 30 to 40 percent of the variance could be explained by characteristics of the offense or offender, leaving open the possibility of considerable arbitrary variation. Today, approximately 80 percent of the variance in sentences can be explained by the guidelines rules themselves. This greater transparency makes it easier to dispel concerns that sentences vary arbitrarily among judges, or that irrelevant factors, such as race or ethnicity, significantly affect sentences.
Evaluation research has been made easier by another benefit of sentencing reform -- the creation of a specialized expert agency with a substantial research mission. The Commission has developed and maintains huge databases on the sentences imposed in each fiscal year, as well as specialized data sets focused on particular issues. These represent the richest sources of information that have ever been assembled on federal crimes, federal offenders, and sentences imposed. As a result, we are in a better position to evaluate whether unwarranted inter-judge, regional, or racial discrimination affects sentences today.
The Effect Of Guidelines Implementation
The effect of the guidelines on unwarranted disparity is best evaluated by comparing, among judges who receive similar types of cases, the amount of variation in sentences before and after guidelines implementation. Researchers both inside and outside the Commission have made this comparison using the "natural experiment" created by the random assignment of cases to judges in many courthouses. The most recent and best of these studies found significant reductions in the unwarranted influence of judges on sentencing under the guidelines compared to the preguidelines era.
Studies of disparity divide judges' influence into "primary judge effects" (greater severity or leniency among judges in all types of cases, represented by differences in their average sentence) and "interaction effects" (greater severity or leniency in particular types of cases). Two judges with similar average sentences may greatly differ in their treatment of particular offenses. Interaction effects can reduce or even cancel the primary judge effect, with one judge sentencing drug offenses more severely than "white collar" offenses and another doing the opposite.
In the Commission's study, the influence of several different factors were compared, including the primary judge effect, interaction effects, city effects, as well as the general type of offense involved and whether an offender had any prior criminal conviction. General offense type accounted for the most variation in sentences both before and after guidelines implementation (between 15% & 20%) followed by interaction effects, city effects, and judge effects. The primary judge effect was relatively small in both the preguidelines and guidelines era, but was reduced by between a third and half under the guidelines (e.g., from 2.32 to 1.24 percent among judges who sentenced in both time periods). Interaction effects were about three to five times larger than primary judge effects. Interaction effects were reduced for most judges under the guidelines, although not among judges who sentenced in both time periods. The influence of judges was reduced by the guidelines for drug, fraud, firearm, and larceny offenses, though immigration or robbery offenses did not show a reduction. Notably, regional differences in drug trafficking cases were increased from the preguidelines to the guidelines era.
Disparity Arising at Presentencing Stages
The SRA focused primarily on sentencing, but Congress, the Commission, and other observers recognized that sentencing could not be considered in isolation. Decisions regarding what charges to bring, decline, or dismiss, or what plea agreements to reach can all affect the fairness and uniformity of sentencing. Congress directed the Commission to develop mechanisms to monitor and, if necessary, control some of the negative effects of plea bargaining, particularly through policy statements establishing standards for judicial review and rejection of plea agreements that undermine the guidelines. In addition, the Commission developed the real offense system of relevant conduct and multiple count rules to reduce the effects of charging variations on the sentencing of offenders who engage in similar conduct. The Judicial Conference of the United States developed procedures for presentencing investigations designed to inform judges of the effects of charging and plea bargaining decisions. The Department also took steps to help ensure that sentencing uniformity was not thwarted at the presentencing stages. The Department's efforts were recently renewed, demonstrating continuing recognition that presentencing decisions can undermine sentencing uniformity.
Congress has previously directed the Commission to study plea bargaining and its effects on disparity. Because fewer statistical data are available to investigate decisions made at presentencing stages, their effects are difficult for the Commission to monitor and precisely quantify. However, a variety of evidence developed throughout the guidelines era suggest that the mechanisms and procedures designed to control disparity arising at presentencing stages are not all working as intended and have not been adequate to fully achieve uniformity of sentencing.
The Commission, as well as outside observers, have reported that plea bargaining is re-introducing disparity into the system. The Commission in 1989, 1995, and again in 2000 compared descriptions of the offense conduct contained in samples of presentence reports with the conduct for which the offenders were charged and sentenced. Each time a large proportion of qualifying offenders (in some cases large majorities) were not charged with potentially applicable penalty statutes. While some offenders are charged in a manner that results in sentences above the guideline range that would otherwise apply to the case, in other cases the charges selected cap the statutory range below the guideline range that would properly apply to the offender's real offense conduct.
Charging decisions that limit the normal operation of the guidelines result in sentences that are disproportionate to the seriousness of the offense and disparate among offenders who engage in similar conduct. Surveys of judges and probation officers have suggested other forms of plea bargaining, such as fact bargaining, that can result in disparity. A majority of chief probation officers reported in a survey sponsored by the Commission's Probation Officer's Advisory Group that the facts included in plea agreements were complete and accurate in the majority of cases. However, 43 percent reported this was true just half the time or less. Probation officers in some districts reported that prosecutors tried to limit information used in applying the guidelines in some cases. The Federal Judicial Center found in a nationwide survey that more than a quarter of responding judges reported that plea stipulations understated the offense conduct somewhat or very frequently, while another 12 percent said they did so about half the time. Judges reported that they did sometimes "go behind" the plea agreements to examine underlying conduct, but they reported doing so "infrequently."
Field studies in several districts have demonstrated other ways that plea bargaining can result in sentencing disparity. An early study sponsored by the Commission estimated that plea agreements circumvented the guidelines in 20 to 35 percent of cases through charge, fact, or date bargaining. Some commentators have called circumvention of the guidelines through plea agreements a form of "hidden departure," in which prosecutors and courts create incentives for guilty pleas and defendant cooperation beyond the incentives contained in the guidelines themselves. In some cases, the sentence recommended in plea agreement appears to the parties and to the court more fair and effective at achieving the purposes of sentencing than the sentence required by strict pursuit of every potentially applicable charge or sentence enhancement.
Other Sources of Disparity Under the Guidelines
Several mechanisms within the guidelines system have been identified by commentators as continuing sources of disparity. These include variation in the rates of departure, including departures for substantial assistance to the government, or the extent of such departures. In addition, the guidelines give judges discretion over placement of the sentence within the guideline range, including, in some cases, whether to use a sentencing option such as probation.
The Commission analyzed the influence of each of these mechanisms on sentencing variations. Among these mechanisms, substantial assistance departures accounted for the greatest amount of variation in sentence lengths -- 4.4 percent. Other downward departures contributed 2.2 percent, while upward departures contributed just 0.29 percent. Only 0.07 percent of the variation was explained by use of the guideline range above the guideline minimum. Because data is unavailable on the types of assistance offered by defendants, or the nature of the mitigating circumstances present in cases, it is not possible to determine how much of these sentencing variations represent unwarranted disparity.
Even though the rate of substantial assistance and other downward departures is similar -- 17.1 percent and 18.3 percent, respectively -- substantial assistance departures account for more variability in sentence length because the extent of departure for substantial assistance is on average greater. Commission research found varying policies and practices in different U. S. attorney's offices regarding when motions for departures based on substantial assistance were made, and in the extent of departure recommended for different forms of assistance.
Racial, Ethnic, and Gender Disparity
Growing caseload of minority offenders and a gap in sentencing. The proportion of the federal offender population consisting of minorities has grown over the past fifteen years. While the majority of federal offenders in the preguidelines era were White, minorities dominate the federal criminal docket today. Most of this shift is due to dramatic growth in the Hispanic proportion of the caseload, which has approximately doubled since 1984. Most notably, while the gap in average sentences between White and minority offenders was relatively small in the preguidelines era, the gap between African-Americans and other groups began to widen at the time of guidelines implementation, which was also the period during which large groups of offenders became subject to mandatory minimum drug sentences. The gap was greatest in the mid-1990s and has narrowed only slightly since then. The Commission had conducted a great deal of research to investigate possible reasons for this gap, including the possible influence of discrimination or of changes to the sentencing laws themselves.
The SRA sought to eliminate all forms of unwarranted disparity, including disparity based on irrelevant differences among offenders. Different treatment based on such characteristics is generally called discrimination. Discrimination may reflect intentional bias toward a group, or may result from unconscious stereotypes or fears about a group, or greater empathy with persons more similar to oneself. Discrimination is generally considered the most onerous type of unwarranted disparity and sentencing reform was clearly designed to eliminate it. Concern over possible discrimination in federal sentencing remains strong today. No sentencing issue has received more attention from investigative journalists or scholarly researchers.
The studies agree on a general point: racial and ethnic discrimination by judges, if it exists at all, is not a major determinant of federal sentences compared to the seriousness of offenders' crimes and their criminal records. But the studies disagree over whether discrimination continues to affect sentencing at all. Many of the earlier studies were plagued by methodological problems, including a lack of good data on legally relevant considerations that might help explain differences in sentences and a failure to take account of statutory minimum penalties. Many of these problems can be overcome by using a "presumptive sentence" model.
The Commission studied whether race, ethnicity, or gender affects federal sentences after controlling for the influence of legally relevant considerations, including the guidelines rules and mandatory statutory penalties. Across five recent years, a typical Black male or Hispanic male drug trafficker had somewhat greater odds of being imprisoned when compared to a typical White male drug trafficker. No differences were found in non-drug cases. The odds of a typical Black drug offender being sentenced to imprisonment are about 20 percent higher than the odds of a typical White offender, while the odds of a Hispanic drug offender are about 40 percent higher. Differences in odds are difficult to translate into plain language, but further analysis examining the proportional reduction in error achieved by using race and ethnicity suggest that in only a handful of cases in any given year does being Black or Hispanic influence the decision whether to incarcerate. Some of these differences might be explained by legally relevant considerations for which we have no data.
For offenders whom judges choose to incarcerate, the question becomes: do similar offenders receive similar prison terms? For Black offenders, the results are once again limited to drug trafficking offenses and to male offenders. The typical Black drug trafficker receives a sentence about ten percent longer than a similar White drug trafficker. This translates into a sentence about seven months longer. A similar effect is found for Hispanic drug offenders, with somewhat lesser effects also found for non-drug and female Hispanic offenders. These findings indicate that all types of Hispanic offenders are placed above the minimum required sentence more frequently than similar White offenders, or receive somewhat lesser reductions when receiving a downward departure. The same is true of Black drug trafficking offenders and Black males.
While any unexplained differences in the likelihood of incarceration or in the lengths of prison terms imposed on minority and majority offenders is cause for concern, there is reason to doubt that these racial and ethnic effects reflect deep-seated prejudices or stereotypes among judges.
Most noteworthy is that the effects, which are found only for some offense types and for males, are also unstable over time. Separate year-by-year analyses reveals that significant differences in the likelihood of imprisonment are found in only two of the last five years for Black offenders, and four of the last five for Hispanic offenders. The effects for sentence length disappear for both Black and Hispanic offenders in the most recent year for which data are available. Offense-to-offense and year-to-year fluctuations in racial and ethnic effects are difficult to reconcile with theories of enduring stereotypes, powerlessness, or overt discrimination affecting sentencing of minorities under the guidelines. In addition, the effects that we observe may be due in part to differences among groups on factors that judges legitimately may consider when deciding where to sentence within the guideline range or how far to depart, but on which we have no data.
Unlike race and ethnic discrimination, the evidence is more consistent that similar offenders are sometimes treated differently based on their gender. Gender effects are found in both drug and non-drug offenses and greatly exceed the race and ethnic effects discussed above. The typical male drug offender has twice the odds of going to prison as a similar female offender. Sentence lengths for men are typically 25 to 30 percent longer for all types of cases. Additional analyses show that the effects are present every year.
Rules Having Questionable Adverse Impacts.
Discrimination by sentencing judges cannot explain the growing gap between African-American and other offenders observed during the guidelines era. Another possibility is sentencing rules that have a disproportionate impact on a particular demographic group. Research has shown that differences in the types of crimes committed by members of different groups and in their criminal histories explains much of the gap in average sentences among them. Rules that are needed to achieve the purposes of sentencing are considered fair, even if they adversely affect some groups more than others. But if a sentencing rule has a significant adverse impact and there is insufficient evidence that the rule is needed to achieve a statutory purpose of sentencing, then the rule might be considered unfair toward the affected group.
In its cocaine reports, the Commission addressed crack cocaine defendants -- over eighty percent of whom are Black -- who are given identical sentences under the statutes and the guidelines as powder cocaine offenders who traffic 100 times as much drug (the so-called 1-to-100 quantity ratio). The average length of imprisonment for crack cocaine was 115 months, compared to 77 months for the powder form of the drug. The Commission reported that the harms associated with crack cocaine do not justify its substantially harsher treatment compared to powder cocaine. For these reasons, the Commission recommended that cocaine sentencing be reconsidered. If the Commission's recommendations were adopted, the gap between African-American and other offenders would narrow significantly. Other rules in the statutes and guidelines have adverse impacts on particular groups. The efficacy of these rules for advancing the purposes of sentencing should be carefully assessed.
Summary and Conclusions
Significant achievement of the goals of sentencing reform. In general, the guidelines have fostered progress in achieving the goals of the Sentencing Reform Act. Sentencing is more transparent, based on articulated reasons stated in open court and reviewable on appeal. Punishment is more certain and predictable, allowing the parties to better anticipate the sentencing consequences of case facts, and allowing the system to better predict the impact of changes in policy on prison populations and correctional resources. Sentence severity has been increased for many types of crime, in some cases substantially. Most important, the guidelines do not admit consideration of factors, such as race or ethnicity, that are irrelevant to the purposes of sentencing. There is less inter-judge disparity for similar offenders committing similar offenses.
Sentencing reform has had its greatest impact controlling disparity arising from the source at which the guidelines themselves were targeted -- judicial discretion. Disparity arising from the decisions of other participants in the sentencing system, or from the process of sentencing policymaking itself, has been less successfully controlled. Statutory minimum penalties are invoked unevenly and introduce disproportionality and disparity when they prevent the guidelines from individualizing sentences. Presentencing stages, such as charging and plea negotiation, lack the transparency of the sentencing decision, making research more difficult. But significant evidence suggests that presentencing stages introduce disparity in sentencing. There is still work to be done to achieve the ambitious goals of sentencing reform in all respects.
Partial implementation of the components of sentencing reform. Part of the reason not all the goals of sentencing reform have been fully achieved is that not all of the components of guidelines implementation put in place at the dawn of the guidelines era have been fully implemented or have worked as intended. Probation officers conduct presentencing investigations to the best of their abilities given limited resources. Judges conscientiously apply the guidelines to the facts as they know them. Appellate review corrects guideline misapplications and alerts the Commission to areas of ambiguity where clarification of the guidelines is needed. But neither appellate review nor guidelines amendments have prevented, at least through the 2002 data currently available, significant variations in departure rates. Neither Department policy nor judicial review of plea agreements has prevented plea bargaining from sometimes circumventing proper application of the guidelines needed to ensure similar treatment of offenders who commit similar crimes.
The SRA also outlined three major components of sentencing policy development: 1) utilization of research and criminological expertise developed by the Commission, 2) collaboration among policymakers and front-line implementers in the courts, and 3) political accountability through legislative directives and review. The Commission has worked to be responsive to the concerns of Congress, and its priorities and policymaking agenda have been greatly influenced by congressional directives and other crime legislation. In some cases, the results of research and collaboration have been overridden or ignored in policymaking during the guidelines era through enactment of mandatory minimums or specific directives to the Commission.
The Commission is uniquely qualified to conduct studies using its vast database, obtain the views and comments of various segments of the federal criminal justice community, review the academic literature, and report back to Congress in a timely manner. These are the processes set out in the SRA, which established the Commission as the clearinghouse for information on federal sentencing practices and the forum for collaboration among policymakers, implementers, and other stakeholders. As an independent agency in the Judiciary, but with frequent interaction with the three branches of government, the Commission is well-positioned to develop fair and effective sentencing policy as long as it continues to receive the resources and support it needs to carry out its vital mission.
Prior to November 1, 1987, the implementation of federal sentencing guidelines, sentencing in the federal courts was very different. Crimes typically carried broad statute-defined ranges of possible penalties and sentencing judges had discretion to choose the penalty within the statutory range they felt would best achieve the purposes of sentencing. Judges were not required to explain the reasons for their sentences, and the sentences themselves were largely immune from appeal. If prison time was ordered, the time defendants actually served depended only partly on the sentence imposed by the judge. Release dates generally were determined by the United States Parole Commission and defendants typically served just 58 percent of the sentence that had been imposed (BJS, 1987).
These factors contributed to a widespread perception that sentences imposed and sentences and prison terms served under the old "indeterminate" sentencing system were unfair, disparate, and ineffective for controlling crime. Respect for law enforcement and the entire criminal justice process was undermined when offenders served only a fraction of the sentence imposed by the judge. The Sentencing Reform Act of 1984 [SRA] sought to establish sentencing practices that would eliminate unwarranted disparity, assure certainty and fairness, reflect advances in criminological knowledge, achieve proportionate punishment, and control crime through the deterrence, incapacitation, and rehabilitation of offenders.
The SRA established the U.S. Sentencing Commission, composed of federal judges and other experts in the field of sentencing, and charged it with the task of promulgating sentencing guidelines for federal courts. After eighteen months of deliberations, the Commission issued the initial set of guidelines, which took effect on November 1, 1987. Four years later, in December 1991, the Commission submitted its report to Congress, 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 to Congress.
The gradual implementation of the guidelines, which applied only to offenses committed after their enactment, and numerous court challenges delayed full implementation until the early 1990s. Furthermore, the guidelines were accompanied by changes of policy and practice that took time to be fully established. Thus, when the Commission released its Four-year Evaluation, it noted the report was a "preliminary assessment of some short-term effects" (USSC, 1991a) rather than a comprehensive examination of the effects of the guidelines on federal sentencing practices.
Twenty years after the SRA was passed and with fifteen years of data on sentences imposed under the guidelines, the Commission is in a better position to evaluate how well the changes brought by the SRA have achieved the ambitious goals Congress set for federal sentencing. This report will update the Four-year Evaluation and outline areas for further research in the continuing evolution of sentence reform.
Overview of the Fifteen-Year Evaluation
Fifteen Years of Guidelines Sentencing is one of a series of publications describing the results of the Commission's fifteen-year anniversary evaluation of the guidelines. In addition to this report, the Commission has published three other monographs: Cocaine and Federal Sentencing Policy (May 2002), the third in a series of Commission reports on cocaine sentencing; A Survey of Article III Judges on the Federal Sentencing Guidelines, Final Report (February 2003), which provides all the findings of the Commission's survey conducted as part of the Fifteen-Year Evaluation; and Downward Departures from the Federal Sentencing Guidelines (October 2003). These reports are available at the Commission website, www.ussc.gov. In addition, the Commission is releasing on its website a research series on the recidivism of federal offenders. Two reports, Recidivism and the First Offender and Measuring Recidivism: The Criminal History Computation of the Federal Sentencing Guidelines are currently available.
Fifteen Years of Guidelines Sentencing undertakes a survey of the federal sentencing system in light of the goals for sentencing reform established by Congress in the SRA. It draws upon a diverse pool of research, including work from both inside and outside the Commission. A bibliography of the published research bearing on the effectiveness of the guidelines is included in this report as Appendix A. The report picks up where the Four-Year Evaluation left off. The Commission targeted three primary areas for special consideration in this report: 1) the guidelines' impact on the transparency and rationality of sentencing, and the certainty and severity of punishment, 2) the impact of presentencing stages and inter-judge and regional disparity, and 3) research on racial, ethnic, and gender disparities in sentencing today. In all three areas, evidence indicates that in the fifteen years under sentencing guidelines, we have made progress toward meeting the goals of sentencing reform.
As policymakers reconsider the federal sentencing system's purposes and effectiveness, the Commission believes improvements in the system can best be achieved by careful consideration of the best available evidence concerning what works in sentencing policy, what doesn't work, and what we still do not know. The Fifteen-Year Evaluation was designed to inform this debate by summarizing the current state of "knowledge of human behavior as it relates to the criminal justice process." 28 U.S.C § 991(b)(1)(C).
END of Executive Summary
Fifteen Years of Guidelines Sentencing; from USSC, November 2004
Table of Contents (.PDF)
(Note: For downloadable PDF format files, you will need the Free Adobe Acrobat Reader.)