INTRODUCTION
Federal criminal sentencing has changed dramatically since 1988. Fifteen years ago, judges determined if and for how long a defendant would go to jail. Since that time, changes in substantive federal criminal statutes, particularly the passage of an array of mandatory minimum penalties and the adoption of the federal sentencing guidelines, have limited significantly judicial sentencing power and have remade federal sentencing and federal criminal practice. The results of these changes are significantly longer federal prison sentences, as was the intent of these reforms, and the emergence of federal prosecutors as the key players in sentencing. Yet, at the same time, average sentence length appears to be falling slowly as judicial tendency to use the authority granted in the United States Sentencing Guidelines (the "Guidelines") to mitigate sentences through downward departures appears to be increasing.
This Article begins with the following two-part question: Why are (1) an apparent increase in judicial sentence mitigation through downward departure and (2) declining average sentence length two legacies of the 1980s' shift to longer prison sentences and increased prosecutorial control of sentence length? Although some might argue these legacies result from judicial resistance to loss of authority or general recognition among judges and prosecutors that federal sentences have become too long, this Article argues that a closer look reveals a much more complex and disturbing picture. Specifically, this Article will demonstrate the presence of two different federal criminal law systems operating in tandem.
The first system is that of narcotics prosecution wherein prosecutorial power often is unchecked and sentences often are unpredictable, but generally are quite harsh. Narcotics sentences have been decreasing steadily for almost ten years--a troubling instability--and there are wide disparities in sentences among similarly-situated defendants. Too many defendants receive sentences that are out of proportion to the wrongfulness of their conduct and too few will accept the risks that come with trying to enforce their rights in the face of often overwhelming prosecutorial power.
The second system involves the bulk of non-narcotics federal prosecution. Although there are many difficult questions presented by current prosecutions in fraud, robbery and other federal non-narcotics cases, the sentences in these areas are far more stable and predictable. Although harsh, problems of excessive severity are not as pervasive among sentences in the non-narcotics area, as there are many fewer mandatory minimum cases and judges are a more effective counterweight to instances of prosecutorial overreaching. Federal non-narcotics sentencing may not be the best of all possible systems of criminal sanctions, but it has proved reasonably stable and predictable over time. This Article will argue that much of the stability in non-narcotics sentences is the result of a defensible balance of power among prosecutors, judges and defense lawyers.
Meanwhile, the vast increase in prosecutorial power to control narcotics sentences is at the core of the problems with federal narcotics sentencing. The profusion of new narcotics and gun proscriptions, almost all of which carry mandatory minimum prison sentences, transformed the traditional prosecutorial power to charge into the contemporary prosecutorial power to determine the length of the sentence the defendant will serve. Although the Guidelines shifted sentencing power from judges to prosecutors across the whole range of federal crimes, this Article argues that mandatory minimum statutes have made this shift in power so extreme in narcotics cases that it has damaged the adversary system and wrought real injustice among those sentenced for narcotics crimes.
This Article has two descriptive and four analytic sections. Part I describes the changes that brought us from a system of almost unfettered judicial sentencing discretion to a system in which federal judges have very limited power to control the sentence in a particular case. Part II identifies some major trends in federal sentencing today and discusses some of the ways federal criminal practice has changed in response to changes in sentencing law. Part III describes the structure of federal criminal case processing and sentencing. Part III also analyzes how, given that structure, prosecutorial discretion has become the key force controlling sentencing in cases involving mandatory minimum statutes, and to a lesser degree, the Guidelines cases in general. Part IV analyzes how judicial power and defendants' bargaining power within the adversarial system can function as salutary checks on prosecutorial discretion, ensuring both efficiency and fairness. Part IV argues that we can predict certain patterns in a well-structured system of federal sentencing. Part V analyzes sentencing patterns in non-narcotics cases and argues that sentencing practices are consistent with the framework suggested in Parts III and IV. Part V then concludes that the Guidelines work reasonably well in the non-narcotics cases. Part VI examines sentencing practices and patterns in narcotics cases and discusses why those patterns signal deep flaws in our current system, flaws that can be traced to the mandatory minimum sentencing statutes. Part VI closes with another criticism of narcotics sentencing, arguing that regardless of the systemic impact, narcotics sentences are simply too long. The Conclusion joins the call for restoring some judicial sentencing power when the next great wave of criminal justice reform sweeps through our political culture.
I. THE SENTENCING REFORMS OF THE LATE 1980S AND THE PROBLEMS THEY ADDRESSED
Before discussing the current trends in federal sentencing, this Article offers a brief overview of how federal sentencing laws changed in the late 1980s and early 1990s in order to fully explicate the currently bifurcated federal criminal law system of non-narcotics cases on the one hand and narcotics cases on the other. This summary may help to situate the current situation for those less familiar with this area and also will signal some of my assumptions to those more familiar with the recent history of federal sentencing.
My first exposure to federal criminal sentencing was during the twilight of the age of unfettered judicial sentencing discretion. (1) In the mid-1980s, federal criminal sentencing was characterized by almost completely individualized and unreviewable judicial decision-making. The length of a federal criminal defendant's sentence was almost entirely in the control of the district judge, (2) although a negotiated plea agreement (3) could cap the possible sentence with a chosen statutory maximum, and prosecutors and defense lawyers developed local customs to influence their judges. (4) In most cases, however, judges could impose any combination of fines and restrictions on liberty from suspending the imposition of sentence, (5) with or without supervision by the probation department, to a sentence of imprisonment at the statutory maximum and the maximum authorized fine. Judges were not required to state reasons for their sentences and a "sentence within statutory limits was, for all practical purposes, not reviewable on appeal." (6) So long as the sentence was within the maximum and procedurally proper, the judge was not subject to review and could not be reversed (7) whether the imposition of sentence for bank robbery was suspended or the defendant was given twenty-five years in the penitentiary. (8)
Consistent with and providing a very limited check on judicial sentencing authority was the parole system. The parole system operated under the Parole Board, which acted under the authority of the executive branch; the Parole Board had the authority to release prisoners after they served a portion of their sentence. Thus, sentences actually served were not completely at the discretion of judges but judges were able to account for parole when they imposed sentence under the old law. Judicial discretion was the key feature of "old law" sentencing, as the sentencing system before the Guidelines has come to be known. (9)
Two related changes, the Guidelines and mandatory minimum sentencing statutes, have remade federal criminal sentencing. Prosecutors today have the discretion and the power to control many federal sentences. The formal, statutory beginning of this change was the Sentencing Reform Act of 1984 (10) ("SRA"), in which Congress mandated the creation of a new commission to develop guidelines. Impetus for the SRA came first from the liberal sentencing reform movement of the 1970s. (11) That movement criticized the standardless, unreviewable sentencing regime under the authority of federal judges as lawless and subject to the prejudices and whims of individual judges. Critics focused on extremely harsh sentences for relatively minor offenses and the probability of race and class discrimination in sentencing. (12) They urged an increase in the use of alternatives to incarceration, the development of sentencing standards, and appellate review to remedy these problems. (13)
Insightful commentators offer versions of how that liberal 1970s movement made a 1980s legislative alliance with conservative law-and-order interests who shared the goals of reining in judges and making sentences more uniform. (14) Riding the "crime wave" and controlling the legislative process, conservative interests were able to meet the demands of their liberal allies for rules and reviewability while achieving their goals of lengthening criminal sentences and eliminating the rehabilitative model of criminal sentencing.
The passage of the SRA created the United States Sentencing Commission (the "Commission"). Congress directed the Commission to draft the Guidelines. (15) The Guidelines (16) replaced a system of judicially determined discretionary sentencing with a system under which judges apply a complex set of rules keyed to the defendant's conduct and criminal history to determine a specific sentencing range--typically a minimum number of months of imprisonment and a maximum number of months of imprisonment--within which the judge must pick a specific sentence of imprisonment. The sentencing ranges under the Guidelines are set out in the sentencing table, which simultaneously increased sentences for many classes of offenses.
The Guidelines are a complex system and represent many compromises, (17) but one clear result of their adoption is that they have completely altered the balance of sentencing power between judges and prosecutors in at least two major ways. First, the Guidelines made the prosecutors' decisions about how much to investigative a given case (18) and about which particular criminal statute to use to charge a defendant, (19) crucial to the ultimate sentence. As this Article will discuss below, in many cases prosecutors can and do bargain with defendants over both the charges to which he or she will plead guilty and the facts upon which the prosecutor will argue the sentence should be based. These two decisions--charges and facts--can have significant impact upon the defendant's sentence. Second, and more fundamentally, the Guidelines replaced unreviewable judicial discretion with a system of rules subject to appellate review. The availability of appellate review has given prosecutors and defense lawyers a set of legally enforceable claims to replace the arguments and signals that lawyers only could hope would persuade judges deciding cases under the old sentencing system.
The most fundamental current constraint on judicial authority is that judges are bound, with certain limited exceptions, to impose sentences within the range determined by the Guidelines. That range is determined by the charges brought against the defendant and some aspects of the conduct underlying the charges. The legal requirement that a judge impose a specific sentence is subject to the exercise of discretion in two areas. The first area of discretion occurs in about two-thirds of cases, wherein judges determine that the sentence must be within the Guidelines range. (20) Once that determination is made, the judge retains the unreviewable authority to decide at what level in the applicable range to sentence the defendant. (21) The second area of discretion concerns the remaining one-third of all cases, which involve downward departures (22) or sentences imposed below the otherwise applicable Guidelines range. About half of the cases in the downward departure group (23) are sentenced outside the Guidelines at a lower level than that prescribed by the Guidelines because the defendant provided substantial assistance in the investigation or prosecution of others. (24) Only the prosecutor can trigger these departures; however, once the government requests a substantial assistance departure, the extent of the departure is up to the judge. (25)
The other half of the downward departure cases (26) involve non-substantial-assistance downward departures from the Guidelines. In these cases, judges impose sentences below the otherwise applicable Guidelines range because the judges found that in the case at hand "there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines that should result in a sentence different from that described." (27) These exercises of downward departure power are subject to review. (28)
Another set of laws, the mandatory minimum sentencing provisions that also came into effect in the late 1980s, significantly constrain judicial authority to depart downward in the narcotics cases to which they apply. In 1986, while the Commission drafted the Guidelines, Congress began another set of sentencing reforms with the passage of the Anti-Drug Abuse Act of 1986 (the Anti-Drug Abuse Act), (29) the first in a long series of mandatory minimum sentencing statutes that have undercut the Guidelines.
The Anti-Drug Abuse Act's provisions included statutory mandatory minimum penalties for those who trafficked in, imported, or possessed specified amounts of particular narcotics. (30) Statutory mandatory minimums are, as their name suggests, statutory provisions that require judges to impose not less than the specified sentence upon conviction. Typically the statutes permit harsher but not lesser sentences. (31) The decision to charge a defendant with a statute that carries a mandatory minimum is in the sole, unreviewable discretion of the prosecutor. (32) Once a defendant has been convicted of an offense carrying a statutory mandatory minimum sentence, there are only two ways the judge can mitigate the defendant's punishment. If the defendant is a first offender, and meets other conditions, he or she may be eligible for a modest sentence reduction through the "safety valve" provision. (33) If the defendant does not meet those conditions, or seeks the possibility of a greater sentence reduction, cooperation in the investigation or prosecution of others, commonly known as snitching or ratting, offers the only mechanism. (34)
Although a defendant must first decide that he or she wants to cooperate with the government, once he or she attempts to offer assistance, the decision whether or not to accept the cooperation and whether or not to make the motion that gives the judge the power to depart on the basis of cooperation once again is in the sole and unreviewable discretion of the prosecutor. (35) While they are constrained by norms and policies, in a very real sense federal prosecutors control the application of the federal mandatory minimum sentencing statutes.
Thus, it is clear that federal criminal sentencing changed dramatically between 1985 and the turn of the twenty-first century. Several trends came together to create a complex body of overlapping laws. Prior to the reforms there was only unreviewable judicial discretion and general norms channeling that discretion. Today, federal criminal practitioners and defendants must wade through a complex set of laws to analyze the impact of the Guidelines on a case, understand and evaluate the potential for departure from the Guidelines, and account for statutory mandatory minimums that may overlap with or supercede the Guidelines in a particular case. It is a complex structure that has unintended and unfortunate consequences that will be analyzed in the remainder of this Article.
II. FEDERAL SENTENCING AT THE TURN OF THE CENTURY
One of the interesting stories that might be told about federal criminal sentencing at the turn of the century is that the prosecutors and judges who enforce our sentencing laws do not use them as aggressively as the Commission and Congress intended. (36) More than one third of the sentences imposed are lower than (departures from) the Guidelines sentence. (37) The rate of downward departure has increased every year. (38) Similar trends can be found in state guidelines systems. (39) The rate of strict compliance with the Guidelines is even lower than these statistics suggest. (40) Prosecutors mitigate many sentences with charge and sentence fact bargaining; (41) such prosecutorial bargaining is not directly observable in the sentencing data. (42) What should we make of all this leniency in a sentencing system designed to encourage longer sentences?
The short answer is that the statistics show only apparent mitigation in many cases. While there are downward departures in more cases as time goes on, (43) average sentence length has remained the same for most offenses. (44) Rates of mitigation are independent of average sentence length because prosecutors and judges use their discretion to keep overall sentence length the same, while they increase individualization of sentences through increased departures. (45) In narcotics cases, however, average sentence length has been declining slowly and steadily for ten years, (46) although narcotics sentences remain quite harsh. (47) In this group of cases, declining sentence length is accompanied by increasing rates of downward departure. (48) Because narcotics cases account for about forty percent of all federal prosecutions, (49) narcotics cases overwhelm all the other offenses in the combined sentencing statistics, which show declining overall sentences because of the impact of the large group of narcotics sentences.
When we look at the narcotics and non-narcotics cases separately, we see that there is a more complicated story behind the appearance of prosecutorial and judicial sentence mitigation. It is a story of balance and imbalance among three forces: (1) prosecutorial discretion; (50) (2)judicial discretion; and (3) the bargaining power of criminal defendants. For most of the non-narcotics cases, prosecutors, judges and defendants have reached a rough equilibrium. First, prosecutors exercise significant control of sentence length through charging decisions and use sentencing phase mitigation to give the appearance of mitigation to induce pleas. Second, judges use their sentencing discretion in a similar fashion to help manage their dockets. Third, although defendants participate in the criminal process under compulsion, they are willing to plead guilty and accept these deals in a reasonably predictable manner. This Article thus argues that the current pattern of stable sentence length and slowly increasing rates of sentence mitigation represents one reasonable balance of forces in our criminal justice system.
Narcotics cases, however, are different. In narcotics cases, the harsh mandatory minimum statutes distort sentencing, hindering efficient and just enforcement of the narcotics laws. Mandatory minimum laws transformed prosecutorial discretion from a tool to allocate scarce enforcement resources into the driving force of the federal criminal process. Unfortunately, prosecutorial discretion is ill-suited for that role. We have lost the efficiency benefits of investigative and charging discretion; rather, prosecutorial discretion has become the unchecked power that allows individual prosecutors to choose among a vast range of punishments with almost no review, causing unwarranted disparity and excessive punishment.
Disparity results when prosecutors have unchecked power to choose among an array of applicable statutes with very different sentencing consequences. The pressure to individualize sentences leads to vastly different punishments for apparently similarly-situated defendants. (51) The negative effects extend beyond the problem of disparity and compromise other important values. In a system in which power is unchecked and trial is no longer a realistic option, evidence is rarely tested, prosecutorial power is not challenged, and normative questions go unasked. Our American devotion to limited power and checks and balances should make us worry about a system (52) in which the fairness of sentences is in the eye of the prosecutor. (53)
Furthermore, whatever the systemic impact, federal narcotics sentences simply are too long. Sentences of fifteen or twenty years and longer have become all too common for defendants who play relatively minor roles in large drug conspiracies. There is little evidence that we are winning the war on drugs this way but a great deal of evidence that we are destroying the lives of thousands of people.
Justice Breyer's comments bolster the view that narcotics sentences under the mandatory minimum laws are simply too harsh and that they are at the core of the general problem with narcotics sentences. Justice Breyer, concurring in Harris v. United States, (54) noted that "[m]andatory minimum statutes are fundamentally inconsistent with Congress' simultaneous effort to create a fair, honest and rational system...." (55) Citing remarks by two other Supreme Court Justices, a Senator, and academic commentary, Justice Breyer observed that mandatory minimum sentences increase disparity, result in unfair sentences, transfer power to prosecutors, and encourage subterfuge. (56)
However, the larger problem with mandatory minimum statutes is their basic incompatibility with the fundamental structures that undergird our criminal justice system: prosecutorial discretion, (57) overcriminalization, (58) and the counterbalancing force of the adversarial model along with the Sixth Amendment right to trial. (59) If we view sentencing as the final phase in the process of criminal law enforcement rather than an end in itself, it is apparent that harsh mandatory sentencing cannot work in our system. In short, criminal statutes designed to overcriminalize in the service of discretionary enforcement cannot and should not be rigidly and uniformly enforced. The high maximum sentences and profusion of crimes and theories of liability are intended to give prosecutors a wide range of choices and to promote efficiency. We cannot rigidly enforce these laws in every case to which they apply because there are simply too many cases. We should not try to enforce these laws in every case to which they apply, as justice will not be served by rigidly applying very broad laws. Yet the alleged benefits of mandatory minimum sentencing statutes--certainty, greater deterrence, and a version of fairness in which everyone gets exactly the same punishment without regard to individual circumstance--would accrue only if the mandatory minimum laws could be enforced rigidly against every wrongdoer. Because we cannot punish every wrongdoer, we should not attempt to punish some wrongdoers excessively. Moreover, experience demonstrates that in fact we are nowhere close to applying mandatory minimum laws uniformly, even to those wrongdoers who are apparently similarly-situated.
Importantly, the structural incompatibility of the Guidelines and statutory mandatory minimum sentences is not merely theoretical. The problems play out in the everyday enforcement of federal criminal law. As we will see, there are important differences between the sentencing practices for offenses that do not carry mandatory minimum sentences and those that do. Those differences can be seen in statistics and case studies that show federal criminal law working reasonably well with the Guidelines in many instances where the mandatory minimums do not apply. Sentencing for most categories of federal crime is characterized by stable average sentence lengths and slowly increasing apparent mitigation. That combination is a sign of balance among important systemic forces. Narcotics sentences, wherein mandatory minimums are at play, however, are not stable over time and sentencing practices are unstable. A closer look reveals an area rife with excessive punishment and injustice.
Problems of lack of stability in narcotics sentences are of recent origin and are the direct result of the changes in the past fifteen years discussed above. (60) The steady march of prosecutorial discretion, coupled with harsh statutory mandatory minimum penalties shifted the focus of federal criminal practice. A focus on sentencing litigation, as opposed to litigating Fourth-, Fifth-, and Sixth-Amendment issues in pretrial motions or litigating guilt or innocence at trial, dominates federal criminal law practice to a degree unimaginable since the late 1980s. Of course, guilty pleas have long been the norm; however, from the late 1960s to the mid 1980s, pretrial litigation was the strategic focus of many cases. (61) Motions to suppress evidence and to challenge indictments were important strategic devices for many defense lawyers. Such motions were used to learn about the evidence, to educate the judge about the favorable aspects of the case, and to gain bargaining leverage.
In the late 1980s, the balance of power shifted among defense lawyers, prosecutors and judges. Mandatory minimum sentences began to have real bite while suppression motions became harder to win as the Supreme Court stepped back from the rights revolution of the mid 1960s and 1970s. Thus, pretrial motions faded in strategic importance. This shift manifested not only in defendants' diminishing likelihood of winning pretrial substantial relief in the form of motions, but also in the fact that the tactical advantages to motion practice faded as judges lost sentencing power to prosecutors. …