Disclaimer: This document was developed by staff for discussion purposes only and does not represent the views of any commissioner. It should not be interpreted as legislative history to any subsequent Commission action. The discussion draft is provided to facilitate public comment on improving and simplifying the sentencing guidelines.
Any punishment, if it is to be reasonable, must be meted out based substantially on an offender's conduct. The scope of an offender's conduct to be considered in coming up with a particular punishment thus becomes a critical determinant of the punishment. In the criminal law, if a sentencing judge considers only the set of criminal acts detailed within the four corners of the charging document that formed the basis of the conviction, the sentence will often be quite different than if the same judge considers related uncharged misconduct or even unrelated uncharged misconduct. If uncharged misconduct is considered, punishment is based on facts proven outside procedural protections constitutionally defined for proving criminal charges, introducing an argument of unfairness that has been repeated often by critics of "real offense sentencing." Defining the appropriate scope of conduct on which to base punishment has been a tug-of-war of fairness and justice for many years for both courts and sentencing commissions.
The scope of conduct considered at sentencing will also affect, at least to some extent, the complexity of a sentencing system. The scope can be as limited as the conduct defined by the elements of the offense or as broad as any wrongdoing ever committed by the defendant or the defendant's partners in crime. All things being equal, a large scope of considered conduct will require more fact-finding than a more limited scope. Generally, then, if a sentencing judge considers only a limited set of facts in determining a sentence, her/his job will be simpler than if she/he considers a much greater set of facts. In the latter case, not only will the number of factual disputes for the judge be greater, but more legal issues will likely be introduced as well. However, as will be discussed ahead, the way relevant conduct is applied, we believe, has a far greater impact on complexity, as well as on fairness, than simply its scope.
Besides fairness and complexity, the scope of conduct considered at sentencing may have serious implications for the balance between prosecutorial and judicial power in sentencing. For example, if the scope of considered conduct is confined to the offense of conviction, many argue that the sentencing system will provide relatively more power to prosecutors to control sentences. If the scope of considered conduct is broader -- more like real offense sentencing -- the prosecutor's charging decisions seem to be much less important.
Finding the right balance among fairness, complexity, and the role of the prosecutor has been a struggle for sentencing commissions generally and, amid the mandate of the Sentencing Reform Act, for the federal commission specifically. It has most often been described simply as a debate between real-offense and charge-offense sentencing. This paper briefly explores this issue and the Commission's response to it: the relevant conduct guideline. Section II discusses the federal criminal code and how the code and the Sentencing Reform Act, in many ways, eliminate the possibility of a pure offense of conviction sentencing system. Section III and IV review the history of the relevant conduct guideline, how critics and the Commission's training staff view the guideline. Section V looks at how state systems have defined the scope of conduct to be considered at sentencing, and how those systems use and apply this conduct to set sentences. Finally, section VI provides some analysis and outlines broad options the Commission has in addressing relevant conduct as well as research questions the Commission may look to answer in order to help choose the appropriate option for refinement.
II. The Federal Criminal Code Compels A Provision Like Relevant Conduct
The federal criminal code has been criticized as a hodgepodge of statutes passed at various times and for disparate and wide-ranging reasons. There have been considerable efforts over the past several decades to reform the federal criminal code so as to provide a more coherent structure. As of now, the code remains a mix of some very specific statutes and some very general and broad statutes, many of which were drafted largely with jurisdictional concerns in mind. The jurisdictional concerns result from the limits the Constitution places on the reach of the federal government into criminal matters. As a result, for much of the federal criminal code, offenses do not contain elements that significantly differentiate culpability among classes of offenders.
For example, the mail fraud statute prohibits using the mails to commit a fraud. The statute does not differentiate those offenders who commit large frauds from those who commit small frauds, those who target vulnerable victims from those who do not, or those who abuse their positions of trust from those who do not. Because the Sentencing Reform Act mandates that the Commission's guidelines differentiate sentences among offenders of different culpabilities, the guidelines, to some degree, must consider aggravating and mitigating factors beyond the elements of the offense in setting sentences for many, if not most federal offenses. Otherwise, a person committing a $1,000 fraud would be sentenced in much the same way as someone committing a $1,000,000 fraud.
As a result, the guidelines must define the scope of conduct beyond the elements of the offense of conviction from which these aggravating and mitigating factors will be gleaned. Similarly, because conspiratorial and accomplice liability are charged and proven so often in the federal system, the guidelines must define the scope of such liability in determining sentences. The point is that in some way, the federal sentencing guidelines must define the conduct to be considered at sentencing beyond the elements of the offense.
III. History of the Relevant Conduct Guideline
Deemed the "cornerstone" of the federal sentencing guidelines, relevant conduct defines the scope of behavior that must be considered in every federal case. Relevant conduct, as it is now defined, can include uncharged conduct, acquitted conduct, conduct described in dismissed counts, and conduct of co-conspirators. Because its application is so critical to the determination of the severity of federal sentences, it has been the subject of significant scrutiny and litigation.
When the Commission was first constructing the guidelines, it sought to develop a pure real offense system. United States Sentencing Commission, Guidelines Manual, Chapter 1, Part A(3), "The Basic Approach," (November 1987) pp. 2-4. It did so for the explicit reason that a charge offense system "affords prosecutors [the potential] to influence sentences by increasing or decreasing the number [and content] of counts in an indictment." Id. The Commission was concerned not only with sentence disparity as a result of judicial discretion but also disparity as a result of "inappropriate manipulation" of the charging decision by prosecutors. As the Commission noted in its discussion of real offense versus offense of conviction sentencing, "the Commission will closely monitor charging and plea agreement practices and will make appropriate adjustments should they become necessary." The Commission believed that to achieve certainty and uniformity, it was mandated to get to the "real" facts of a case irrespective of the prosecutorial charging decision. It also believed that under pre-guidelines practice, sentencing judges could and did consider whatever facts they wanted to, whether related to the offense of conviction or not. Finally, the Commission drew on the fact that the Parole Commission did in fact consider all real-offense conduct in making parole decisions.
The early Commission tried to devise a sentencing system that would use real-offense behavior and would separately account, in a detailed and formulaic way, for as many harms caused by defendants as was practicable. The early commissioners, however, found that a pure real offense system that separately accounts for all harms would be intolerably complicated.
To make such a system work, even to formalize and rationalize the status quo, would have required the Commission to decide precisely which harms to take into account, how to add them up, and what kinds of procedures the courts should use to determine the presence or absence of disputed factual elements. The Commission found no practical way to combine and account for the large number of diverse harms arising in different circumstances; nor did it find a practical way to reconcile the need for a fair adjudicatory procedure with the need for a speedy sentencing process given the potential existence of hosts of adjudicated "real harm" facts in many typical cases. Id.
The complexity that the Commission found was due not only to the scope of relevant conduct but also to the fact that the Commission wanted to account for all the real offense facts through detailed sentencing formulas.
As a result, the commissioners reluctantly moved away from a real offense system toward an offense of conviction system. To be true to their mandate, though, they moved only as far as they thought they needed to create a "workable" system. The guidelines still needed a real-offense component, and as a result, the Commission still needed a formulaic way to get to the real-offense facts irrespective of what was in the prosecutor's charging document. Hence, the creation of "relevant conduct" and the modified real offense system. Under this system, the offense of conviction provides the starting point -- the Chapter Two guideline -- for calculating sentences. In applying the appropriate Chapter Two guideline, however, relevant conduct allows for consideration of real offense facts: facts beyond those directly related to the offense of conviction.
The relevant conduct guideline defines the scope of conduct to be considered at sentencing in two ways. For one set of offenses, notably robbery and offenses against the person, section (a)(1) of the relevant conduct guideline limits the scope of conduct to be considered at sentencing to acts that occurred during the commission of the offense of conviction, in preparation for that offense, or in the course of attempting to avoid detection or responsibility for the offense. This is somewhat close to an offense of conviction scheme. This is not, however, close to an elements of the offense of conviction scheme. Section (a)(1) requires the consideration of facts beyond the elements of the offense but, as stated in the text, directly related to the offense of conviction. The conduct used to determine the sentence goes beyond the elements of the offense but is limited to conduct occurring around the offense of conviction. Under section (a)(1), all acts committed by a defendant, aided and abetted by him/her, and reasonably foreseeable in furtherance of his/her jointly undertaken criminal activity are considered part of relevant conduct so long as the acts are related, as described above, to the offense of conviction.
For a second set of offenses - so-called "aggregatable offenses" including drug, fraud, and firearms offenses - however, section (a)(2) broadens relevant conduct to include conduct that is part of the same course of conduct or common scheme or plan as the offense of conviction. This is the provision that allows consideration of uncharged conduct, acquitted conduct, and conduct described in dismissed counts. Sentences for the offenses that use this broader definition of relevant conduct were considered by the early Commission to be quantity driven or "aggregatable." The Commission believed that before the guidelines, in sentencing these offenses, judges considered the real and complete quantity of the contraband involved in the illegal activity irrespective of how the prosecutor charged the offense (i.e., how much of the contraband was actually described in the charging document) and irrespective of whether a jury acquitted on one count or another of a multiple count indictment. This mixed sentencing system - a system that is predominantly charge-based for certain offenses but predominantly real offense-based for so-called "aggregatable offense" - in and of itself has caused confusion and complexity for many practitioners. The Commission determined that continuing this practice was the appropriate way of fulfilling the mandate of the Sentencing Reform Act.
The Commission believed, however, that the non-aggregatable offenses were very similar to state law criminal conduct, and thus the Commission thought that it was more appropriate to use a sentencing system tied more to the offense of conviction for these offenses. The aggregatable offenses were thought to be more uniquely federal. Because the Commission found that pre-guidelines sentencing practice considered conduct beyond the offense of conviction most often for these offenses and because, as stated above, the parole guidelines -- which the sentencing guidelines were in part replacing -- were based on real offense conduct, the Commission determined that sentences for aggregatable offenses should be based more on real offense conduct.
In addition, it should be noted that since their initial development, the Commission has introduced into the guidelines a significant number of cross-references to other guidelines. These cross-references allow relevant conduct, rather than the offense of conviction, to determine the appropriate Chapter Two guideline from which the sentencing calculation begins. As the number of cross-references increases, real-offense conduct becomes more important in the sentencing determination and the offense of conviction becomes less important. In other words, by introducing more cross-references over the recent years, the Commission has moved the guidelines closer to a real-offense system.
IV. How Critics and the Commission Training Staff View the Relevant Conduct Guideline
A. The View of the Critics
Most of the outside criticism of the relevant conduct guideline surrounds the issue of fairness and section (a)(2) which brings into consideration acts not encompassed by a count of conviction that are part of the same course of conduct or common scheme or plan as the offense of conviction. Critics charge that relevant conduct, and specifically section (a)(2), encompasses too much unconvicted conduct, that sentences can be driven by unconvicted conduct, and as a result the full constitutional protections surrounding the criminal justice system, for practical purposes, are lost. These critics point out that there is no grand jury review of relevant conduct, no need to set out relevant conduct in a charging document, and lesser procedural or evidentiary protections surrounding its proof. Few critics, however, suggest that relevant conduct alone is responsible for the guidelines' complexity.
B. The View of the Training Staff
Since the initial set of guidelines were issued in 1987, the Commission's training staff has found that the relevant conduct guideline has been among the most troublesome for application and that the guideline's application has been very inconsistent across districts and circuits. In attempts to remedy this situation, the relevant conduct guideline has been amended nearly every year since the guidelines were promulgated. The training staff believes that there are several reasons for the application problems. First, in defining relevant conduct and in so doing, describing sentencing liability, the Commission used legal terms of art that had been traditionally used to describe criminal liability. For example, the Commission intended that relevant conduct include specific acts or omissions the defendant "aided and abetted." Because the Commission used the terms "aided and abetted," which have a specific and broader meaning in the criminal law than the meaning intended by the Commission, many users focus not on the specific acts the defendant aided and abetted, as the Commission seems to have intended, but rather on the entire principal crime that the defendant aided and abetted. As a result, the training staff believes that application has been inconsistent and in many cases not what the Commission intended. The definition of conspiratorial liability under the guidelines poses similar problems.
Second, because the Commission defined sentencing liability for conspiracies more narrowly than traditional criminal law conspiratorial liability and because the Commission's definition of sentencing liability for conspiracies is intricate and fact specific, the training staff believes that applying this definition has been a struggle for attorneys, probation officers, and courts since the advent of the guidelines. Specifically, unlike criminal conspiratorial liability, relevant conduct limits sentencing conspiratorial liability to "jointly undertaken criminal activity." This prong of relevant conduct often requires courts to hold significant hearings to determine what part of a defendant's criminal law conspiratorial liability "the particular defendant agreed to jointly undertake (i.e., the scope of the specific conduct and objectives embraced by the defendant's agreement)" as well as all reasonably foreseeable conduct of others in furtherance of the jointly undertaken activity. USSG 1B1.3n2. Because this determination is case- and fact-specific, and because the determination can drive a guideline sentence, it is litigated in many cases. Commission research shows that after the drug guideline, relevant conduct is the most frequently appealed guideline issue. These data further show that most of the appeals surround the definition of conspiratorial liability.
Third, the training staff believes that several aspects of the way the relevant conduct guideline is drafted make for difficulties in application. For example, in setting out the offenses for which the "same course of conduct, common scheme or plan" rules apply, the Commission refers to offenses "for which 3D1.2(d) would require grouping of multiple counts." This has confused some attorneys and probation officers who think that this section applies only if there are in fact multiple counts. The training staff has also found that because of the structure of the guideline, many users applying 1B1.3(a)(2) do not realize that the criteria from 1B1.3(a)(1) also must be met for proper application.
The training staff can cite other examples of application difficulties with the relevant conduct guideline. Many of these application problems might be addressed without changing the fundamental policy choices concerning the current modified real-offense system. However, it is not clear that any changes to address these application problems would significantly simplify the guidelines in general or the relevant conduct guideline in particular. Most of the yearly amendments to the relevant conduct guideline were made attempting to clarify the guideline and make its application easier. Some argue that since many significant amendments to relevant conduct have been made recently and because the concepts surrounding relevant conduct are inherently complex, that courts are still struggling to catch-up and interpret these changes. This might suggest that if no substantive policy changes are to be made, that simplifying relevant conduct may mean simply leaving the guideline alone and allowing courts to interpret and adjust to it.
IV. State Guideline Systems
Like the federal criminal code, most state criminal codes, for many classes of crimes, do not differentiate among offenders of differing culpabilities. As a result, most state guidelines systems consider conduct beyond the elements of the offense of conviction in determining sentences. In fact, most state guideline systems consider as much or more of a defendant's conduct than the federal sentencing guidelines. However, sentences under these systems are primarily determined by the scope of conduct that occurred during the offense of conviction. Under most of these systems, the judge is then able to adjust the sentence for conduct that goes beyond the offense of conviction.
The North Carolina sentencing guidelines are a good example. Like the federal guidelines, the North Carolina guidelines determine sentences based on a grid. The sentencing judge first determines the offense severity level, which is fixed by the offense of conviction. Next, the judge determines the defendant's prior criminal record. These determinations define the grid location which contains three sentencing ranges: a presumptive sentencing range, an aggravated range, and a mitigated range. The judge next determines whether there are aggravating or mitigating factors present in the case and whether the aggravating factors outweigh the mitigating factors or vice versa. If aggravating factors outweigh mitigating, the judge sentences in the higher aggravating range. If mitigating factors predominate, the judge sentences in the lower mitigating range. If neither aggravating or mitigating factors predominate, the judge sentences in the presumptive range.
The North Carolina guidelines' list of aggravating factors include specific factors related to the offense of conviction (e.g., "whether the offense was especially heinous, atrocious, or cruel") and a catch-all: "[a]ny other aggravating factor reasonably related to the purposes of sentencing." Similarly, the list of mitigating factors include a catch-all: "[a]ny other mitigating factor reasonably related to the purposes of sentencing." The catch-all aggravating factor has been interpreted by the North Carolina Supreme Court to allow consideration of events that were part of an uncharged course of misconduct. North Carolina v. Farlow, 336 N.C. 534 (1994). In other words, under the North Carolina system, uncharged conduct, and the other elements of the federal relevant conduct, can be considered at sentencing. However, the consideration is limited by the structure of the sentencing calculus so that the final sentence is driven primarily by the offense of conviction.
Almost all other state guidelines allow for consideration of uncharged conduct in determining sentencing. However, most of these systems, like the North Carolina guidelines, determine sentences first and primarily through the offense of conviction.
V. Analysis, Options For Refinement, and Research Questions
There is one paramount policy question the Commission must answer in determining whether and how to substantively refine relevant conduct and related guidelines: does the Commission want to continue to move toward a real-offense sentencing system, does it want to stay with the current mixed system, or does it want to reverse direction and move toward a charge-offense system. As referred to earlier, the answer to this question depends in significant part on the Commission's view of plea bargaining, whether the Commission continues to see as its role the regulation of the plea process -- so as to avoid unwarranted disparity and satisfy the mandate of the Sentencing Reform Act -- and whether the issues of fairness raised by the guidelines' critics outweigh the concerns over the plea process. If the Commission moves closer to either a real-offense system or a charge-offense system, the repercussions on the plea process and fairness could be significant. In addition, the complexity of guideline application may be significantly affected depending on the techniques the Commission uses to implement the change.
In answering the fundamental policy question of real- versus charge-offense sentencing, the Commission will likely want to examine information and data being collected by the current Assessment Project. These data will hopefully address, for example, whether the real-offense approach of the guidelines has helped prevent or reduce unwarranted disparity caused by prosecutorial decisions, whether the real-offense approach has led to abuses in which prosecutors take advantage of the relaxed procedural safeguards in the sentencing hearing, and whether and to what degree the critics' charges of unfairness are real. In addition, the Commission will likely want to assess what the likely results would be of a more charge-oriented sentencing system.
In broad terms, the Commission has at least six options in addressing the relevant conduct guideline. First, the Commission could simply leave the guideline alone and make no changes. Obviously, this would leave in place the substantive decisions of earlier Commissions and would not address the criticisms of the guideline. Simply in terms of complexity and application, some argue that over the past eight years, judges, attorneys, and probation officers have struggled in applying relevant conduct, but now, users are becoming more familiar and soon application problems and some of the appellate review will diminish (see footnote 7). Because the relevant conduct guideline has been amended so often, and because the concepts underlying the guideline are inherently difficult to apply, amending the guideline when no substantive changes are being made may not clarify or simplify but may simply continue whatever confusion already exists and perhaps create new confusion. In other words, it may not be productive to rewrite a guideline in an attempt to clarify it.
Second, the Commission could leave in place the scope of the current relevant conduct guideline and simply try to revise the language to address some of the application problems discussed above. For example, the Commission could spell out the offenses when section (a)(2) applies rather than referencing the multiple count grouping rules. This might eliminate the confusion over the need for multiple counts before applying section (a)(2). As mentioned above, such changes could cause confusion rather than simplify.
Third, the Commission could narrow the scope of relevant conduct - moving closer to a charge-offense system - and leave in place the way relevant conduct is applied. If the Commission followed this course, it might also alter the way accomplice and conspiratorial liability are defined for sentencing purposes by the guidelines. This could also be done in a variety of ways and would similarly implicate prosecutorial power and the plea process.
As discussed above, this would likely lead to some moderate changes in the complexity of the guidelines - as the scope of facts to be considered by district judges would decrease - while at the same time addressing some of the concerns and criticisms about fairness. However, this would leave in place the way the guidelines overall calculate sentences and thus would arguably not address the fundamental complexity of the guidelines.
Fourth, the Commission could change the way relevant conduct is used in calculating sentences but leave in place the current scope of relevant conduct. As will be discussed in future briefing papers, relevant conduct is applied in a long list of case-specific aggravating and mitigating factors. Because these aggravating and mitigating factors are applied in formulas with specific numerical values given to each factor and because all aspects of relevant conduct can drive sentences, the importance of the scope of relevant conduct is greatly increased. In other words, if relevant conduct were not so pivotal in sentencing or if it were applied differently (more simply, like some of the state systems), it might not be so complicated or so feverishly litigated. Also, if the impact on sentences of uncharged, acquitted, or dismissed conduct were limited, many of the criticisms concerning fairness could be addressed.
As stated above, currently, relevant conduct beyond the offense of conviction can drive a sentence. The Commission could limit the way uncharged, acquitted, or dismissed counts could be used in the sentence calculation. This could be done in a variety of different ways, including placing a cap on the increases attributable to unconvicted conduct or implementing a single upward adjustment for uncharged misconduct. Depending on the Commission's choice, the mechanistic nature of the guidelines could be reduced.
Fifth, the Commission could narrow the scope of relevant conduct and change the way relevant conduct is used. And sixth, the Commission could move in the other direction and expand the scope and application of relevant conduct, moving even closer to a real-offense system. Depending on the mechanism used to do so, this could further complicate the guidelines or could simplify them. If relevant conduct were expanded and the current application mechanism were retained, the system would likely become more complicated. However, if there were a single adjustment for real-offense conduct, even if the real-offense component were expansive, the overall sentencing system could be much simpler.
The decision on whether to continue with the Commission's momentum toward a real-offense sentencing system is a fundamental one that will drive the decision whether and how to refine relevant conduct. The Assessment Project should provide some information with which to help make the decision. Each of the broad options outlined above have implications for fairness, complexity, prosecutorial power, and justice. Depending on the substantive policy choice the Commission chooses and the specific mechanism chosen to implement the choice, a new balance of fairness, complexity, prosecutorial power, and justice can be struck.
United States Sentencing Commission