The calculus of the record: Criminal history in the making of US Federal Sentencing Guidelines

Abstract

The Federal Sentencing Guidelines, developed by the United States Sentencing Commission in the 1980s, appear to exemplify the turn from individualization toward aggregated, rationalized risk management that ostensibly became hegemonic in the late 20th century. In this article, we challenge that presumption by building on Harcourt’s (2007) argument that penal actuarialism emerged as part of the individualization project rather than as a repudiation of it. We trace how ‘criminal history’ came to be the primary mode for capturing defendant characteristics in the Guidelines formula, and delineate how time became the unit of quantification to transform criminal history into ordinal measures of current culpability. We draw three lessons from this case study: about the durability of old practices and logics; about how the individual penal subject lives on in sentencing regimes like the federal Guidelines system; and about the metrics of time and history as ways of knowing the juridical subject.

Keywords

Actuarialism criminal record punishment and society sentencing temporality

Introduction

The Federal Sentencing Guidelines system is perhaps the most iconic example of a ‘formal-rational’ legal system (Savelsberg, 1992) that ostensibly became hegemonic in the late 20th century. Developed during a period when numerous state systems in the USA were also moving from indeterminate sentencing schemes to ones that guided and constrained judicial sentencing discretion (Harcourt, 2003; Tonry, 1996), the federal guidelines represented an extreme version of this move in the intricate level of control they imposed upon those tasked with applying them (Frase, 1999). The annual Guidelines Manual, which delineates the rules for calculating the appropriate sentence range for all criminal defendants convicted in federal court, is comprised of eight chapters as well as six appendices and supplements that typically fill more than 2000 pages. The ultimate goal of these rules is to ‘inject transparency, consistency, and fairness into the sentencing process’ (USSC, 2004: iv). It does so by translating defendants’ conviction and ‘relevant conduct’ (USSC, 2013) into numeric ‘offense levels’ ranging from 1 to 43, and converting their past criminal records into numeric ‘criminal history scores’, which are then grouped and categorized on a scale from 1 to 6. The offense levels and criminal history categories are then set in a two-dimensional grid as the y-axis and x-axis, respectively.

By design, this new system appears to have eschewed the individualization inherent in the previous indeterminate sentencing that aimed for substantive rationality through flexible and particularized intervention (Savelsberg, 1992). The Guidelines’ formulation groups diverse individuals into ordinal categories based on the numeric translation of just two sets of information: the circumstances of the present criminal acts and the prior criminal record of the actor. As such, the Guidelines look on the surface like a case-in-point of the rise in aggregate management of offender populations (Feeley and Simon, 1992).

In this article, we trace the original development of the criminal history component of the Guidelines formula, to argue against that narrative. Specifically, following Harcourt (2007), we show how the use of criminal history in the federal sentencing scheme came to constitute the principal mode for individualizing the penal subject. As a practice that survived the transformation of both the federal and state sentencing systems, the reliance on criminal history in sentencing challenges the ascendant theoretical account of punishment as a practice that sublimates the individual to the group. By pointing to the significance of how individual characteristics are articulated in the form of criminal history, we offer a more dynamic picture of the calculative logic and the context-specific moral, emotive, and political sensibilities through which penal logic emerges and takes shape (Sparks, 2001).

Using primary documents from the initial years of the United States Sentencing Commission—transcripts of hearings, drafts of the Sentencing Guidelines, and notes and memos written by and between the Commissioners¹—we first examine how ‘criminal history’ in this context was elevated from one of many potentially relevant defendant characteristics delineated by Congress to its singular status as the x-axis in the Sentencing Table. We then document how criminal history enters into the Guidelines with considerable—nonetheless unspoken—legitimacy, including how its pre-existing uses were imported into and then reshaped through the Guidelines construction project. We unpack how those who participated in drafting the Guidelines came to focus on criminal history as a singular defendant-related characteristic; how they read criminal history as a marker of both risk and culpability; and how they ultimately operationalized the goal of knowing the criminal subject through his criminal past via a metric of time.

The developmental course of criminal history in this context reveals new insights into the particular case of the Guidelines’ formation, and speaks to the broader theoretical terrain addressing late modern criminological and penal transformations. Specifically, our examination of this history contributes to a diverse and growing body of empirically grounded, site-specific examinations that endeavor to contextualize penal/legal change in order to sharpen theoretical explanations of how and why particular logics and techniques ascend (Hannah-Moffat and Lynch, 2012). Such work challenges the predominant theoretical explanations of sentencing and punishment transformations, which adopt a top–down perspective to put forth ‘a hierarchical and determinist view of the processes of transformation’ and obscure the ‘fluid, unstable, and synergistic’ ways that different rationalities may emerge and operate (Tata, 2007: 427, 430). By scaling down, rather than up, we emphasize the durability, adaptability, and transportability of temporally bound cultural and ideological knowledge structures (see Fine, 2012).

Understanding this logic is of more than just theoretical concern; the effects of criminal history enhancements on both institutions and individuals are substantial. Under some state sentencing guidelines, the recommended length of prison sentences are two to three times longer for defendants in the highest criminal history categories, as compared to those in the lowest criminal history category (Roberts and Yalincak, 2014). In the federal system, criminal history can account for nearly half the sentence severity (Stith and Cabranes, 1998). Certain criminal history-based enhancements, like the ‘career offender’ guideline, can add decades onto prison sentences (Baron-Evans et al., 2010).

In the next two sections, we delineate the role that criminal history has played in sentencing, including justifications for its use, and then provide some brief background of the Federal Sentencing Guidelines’ development.² We proceed to our analysis, first exploring how the defendant became understandable primarily through a criminal past, which eclipsed the myriad other ways he could have been measured and constructed. We then examine in detail how criminal history was transformed into an actual metric in this system, delineating its antecedents and its evolution during the drafting stage. We conclude by drawing three lessons from this case study: about the durability of old practices and logics even in moments of dramatic change; about how the individual penal subject lives on, in a reconstituted form, in sentencing regimes like the federal Guidelines system; and about the metrics of time and history as ways of knowing the juridical subject.

Criminal history and penality

As an object of analysis, criminal history may not at first blush seem to be particularly generative of new insights about law, punishment, or criminal subjectivity. Its relevance to criminal sentencing is legally well established and broadly accepted as a matter of common sense. It is widely used as a sentencing factor by courts in most western nations (Roberts, 1997), and its historical roots stretch back to Leviticus (Durham, 1987; see King, 2014 for US historical context specifically). Whether used diagnostically as an indicator of appropriate rehabilitative interventions, or to suggest the boundaries of necessary punishment under the logic of deterrence, criminal history has long played a role in sentencing determinations across diverse penological rationales. It has been particularly integral to the emergence of risk-based penal models, such as selective incapacitation, which were developed in earnest in the 1960s and 1970s and became ideologically hegemonic in the USA in the 1980s (Von Hirsch, 1985).

As the goal of selective incapacitation rose in prominence, criminal history was institutionalized as a key (or sole) variable in predictive formulae that emerged as the solution to the failures of prevailing practices. While this same period also saw a resurgence in retribution as a penal goal, which has little place for offender-specific sentencing factors, including prior criminal record, criminal history nonetheless has become entrenched in contemporary sentencing schemes of all philosophical bents (Von Hirsch, 1985, 2014; see Rappaport, 2003 for a critique of its use under a retributive rationale).³

Consequently, criminal history became a core element of most sentencing reforms that occurred in the 1970s and 1980s, even if the rationale for its use has not always been apparent (Tonry, 1996, 2014). As numerous scholars have documented, the sentencing transformations of this period were as much reactions to what was perceived as a failure of the prevailing rehabilitative rationale as a positive vision of what sentencing should look like (see, for example, Paternoster and Bynum, 1982; Tonry, 2005). Thus, there was significant variation in the reforms that emerged in this period—a ‘crazy quilt of diverse systems’ (Tonry, 2005: 1238)—from the development of formal guidelines systems, exemplified by Minnesota’s scheme implemented in 1980, to statutory reforms specifying sentence ranges for given offense convictions, such as those enacted in the late 1970s by states like California and Arizona.

The standard theoretical account of how the criminal record became so central to contemporary sentencing schemes is that it is largely due to the rise of penal actuarialism (Feeley and Simon, 1992) as embedded in the broader shift to legal rationality, which privileges coherence, consistency and predictability across aggregated ‘cases’ over substantive, individualized justice (Tata, 2007). Yet the presumption that actuarialism in the penal field is the product of a larger ideological shift toward aggregate risk management may not be fully founded. As Harcourt (2003, 2007) provocatively suggests, the rise of actuarial methods may represent a turn to individualization rather than away from it, and any delays in their broad dissemination and implementation can be attributed largely to resource constraints. Harcourt (2007) traces actuarialism’s penal use in the USA to the 1930s, followed by several decades of robust development and even competition among scholars about appropriate models for predicting individualized risk. By the time prediction tools resurged in the 1970s, the very institution that had been key to their development and implementation—parole—was in retreat. Parole’s assessment tools, however, were portable and could be moved into the emerging sentencing regimes with ease. For Harcourt (2007: 110) the new actuarial age represented the ‘zenith in the turn to the individualization of punishment’.

Harcourt’s account foregrounds the critical importance of historical antecedents in understanding contemporary phenomena. The historical place of using prior criminal records to determine punishment raises two important considerations. First, if the actuarial use of criminal records is indeed rooted in individualization efforts, then that logic is likely to endure, in some form, even if the ‘new penology’ is focused on aggregated categories of risk. Second, and more concretely, the technologies developed around its use—particularly as a factor in parole risk-prediction tools—will potentially survive even revolutionary change to formal legal structures. More fundamentally, Harcourt (2007) challenges the prevailing logic that the current predominance of actuarialism and categorization are the counterpoint to individualization. Rather, just as clinicians use categories to diagnose and label individual patients, actuarial methods aim for more reliability in diagnostic labeling to determine interventions for individual subjects.⁴ Furthermore, while the purposes of actuarialism may have changed—say, from reforming to incapacitating the penal subject—individualization is central to the actuarial project.

Therefore, adapting Harcourt’s argument, we suggest that the actuarial strategies developed in conjunction with the Sentencing Guidelines reflect an evolving technical twist on long-established logic. That is, the underlying commitment to using the ‘criminal’ past to dictate the present sanction decision was so well established that its centrality to the Guidelines project was not up for debate. Moreover, rather than a move away from individualism, the Sentencing Guidelines incorporated criminal history as the mode of individualization: the criminal past substantially constructs the present defendant as a juridical subject. In other words, the Commission translated its mandate to provide guidelines that account for defendant characteristics in addition to offense characteristics almost solely through assessing the individual defendant through his prior criminal record.

From the Sentencing Reform Act to early implementation

The Federal Sentencing Guidelines were initially promulgated by the United States Sentencing Commission following congressional passage of the 1984 Sentencing Reform Act (SRA), which had been in the making for nearly a decade (Stith and Koh, 1993). Federal sentencing reform first emerged in the context of a broad political and legal critique of indeterminate sentencing that was generally sympathetic to the plight of criminal defendants and prisoners, and concerned with the discriminatory effects of an overreaching criminal justice system (e.g. American Friends Service Committee, 1971; Clark, 1970; Mitford, 1973). Notably, the early versions of the sentencing reform legislation contained provisions to lower maximum prison terms and encourage alternative sanctions (Stith and Koh, 1993). They also included provisions that used current prison capacity as a limit on the punitiveness of the future sentencing guidelines, and that called for serious, holistic consideration of the ‘history and characteristics of the defendant’ when determining appropriate sanctions (s.1437 §100, 1978). In the end, though, the overarching aims of the early legislative proposals were in some ways the antithesis of those embodied in both the statutory elements of the 1984 SRA and in the version of the Guidelines authorized in 1987. While the goal of consistency and controlling ‘unwarranted’ disparity remained a constant from beginning to end (Murakawa, 2014), the punishment metrics transformed as reform efforts moved into the more law-and-order 1980s (see Stith and Koh, 1993 for a detailed history).

The SRA specified that an independent sentencing commission, administratively placed in the judicial branch, would be tasked with promulgating a set of guidelines that responded to the statutory mandates of the Act (28 US Code (USC) §994(a)). The statute itself reflected a reluctance of legislators to commit to a single vision of what the purpose for punishment should be. Rather, Congress delineated four potentially contradictory sentencing goals that the courts needed to consider when determining sentences. Specifically, Congress required future sentences to reflect the seriousness of the offense and ‘promote respect for the law’, afford adequate deterrence, protect the public, and provide the defendant with necessary education or training, care, or treatment (18 USC §3553(a)(2)(A–D)).

The SRA suggested two axes around which sentence guidelines should be built to meet these penal purposes: offense characteristics and defendant characteristics (28 USC §994(c–d)). Seven potential offense characteristics and 11 potential defendant characteristics, including traditional sentencing factors such as age, family ties and responsibilities, mental and emotional condition, and employment history, were delineated in the statute. Criminal history was number 10 on the defendant characteristics list, with no indication that it should be elevated above any other potential factor for consideration.⁵ Nonetheless, by the time the Guidelines were authorized in 1987, the traditional mitigating defendant characteristics specified in the statute had been deemed either categorically irrelevant or not ordinarily relevant to the sentencing decision (Harcourt, 2003). Conversely, criminal history had been promoted to a one-dimensional representation on the x-axis of the sentencing grid—a placement that ensured it would drive all sentences under the Guidelines.

While there was some intellectual overlap with the state guidelines systems that preceded the federal one,⁶ original members and staff of the US Sentencing Commission have strenuously argued that those state systems—for better or worse—did not serve as models for the federal guidelines (e.g. Breyer, 1988; Knapp and Hauptly, 1992; Nagel, 1990). Thus Kay Knapp, who held high-level staff positions at both the inaugural Minnesota Sentencing Commission and the inaugural US Sentencing Commission, reported that the US Sentencing Commission ‘essentially ignored the state efforts’ due to its policy choice to develop a ‘highly mechanistic’ system (Knapp and Hauptly, 1992: 681–682). The federal system did eventually adopt a grid system comprised of offense characteristics and criminal history as the two axes, which resembled most state systems’ tools. But as Frase (1999) has suggested, the state systems are distinguishable from the federal system on their relative flexibility and in their simplicity and user-friendliness. The system that emerged from the original Commission’s efforts was neither of these things.

Defendant characteristics and an emerging calculus of criminal history

The Commission went into the Guidelines construction project open to a broad conception of which defendant characteristics might play a role in the sentencing calculus. In late 1985 and early 1986, a sub-committee of the Commission was tasked with creating a list of potentially relevant offender characteristics. Commissioner Paul Robinson (nd) suggested to the group that it make a comprehensive list of offender characteristics, explicitly linking each to one of the purposes of punishment. In response, the committee produced a 12-page document titled ‘List without narrative: Offender characteristics’ (Offender Characteristics Committee, 1986), which included an exceptionally wide array of factors, broken into three general categories: factors related to the commission of the offense; factors related to prediction of future criminal conduct; and ‘factors related to other humanitarian or societal interests’ (1986: 2). The list included items as diverse and unusual as ‘low intelligence’, ‘chromosomal abnormality’, ‘falsely claimed poverty’, and ‘choice between competing evils’ (1986: 3). About a month later, Commissioner Block, the Chairman of the sub-committee, requested feedback from select outsiders about what they deemed to be important offender characteristics.

One of the responses Commissioner Block received seemed to foreshadow what was to come. The letter writer was a law professor and commissioner in Pennsylvania’s Sentencing Commission who called the list ‘absurd’ (singling out ‘chromosomal abnormality’ for derision) for its focus on anything other than measures of criminality (Jones, 1986: 1–2). He advised the Commission to follow Pennsylvania’s lead and focus on dangerousness, operationalized by ‘two paramount criteria: seriousness of the (current) offense, and the offender’s criminal history’ (1986: 1). These words were underlined in pencil, likely an indication of their salience for the letter recipient.

By May 1986, the Commission had indeed shed the many possibilities for assessing relevant individual defendant characteristics. Following a public hearing that dealt with the offense characteristics comprehensively, the Commission held its second hearing that month on ‘Offender Characteristics: Prior Record’, in order to ‘discuss the availability and reliability of prior record information and how that prior record information should affect sentencing’ (Hearing Transcript, 1986: 1). Thus, rather than holding a general ‘offender characteristics’ hearing, the Commission had by this point isolated the criminal record as the key factor by which to account for individual defendant differences, and sought input from a range of witnesses on how to account just for that in the sentencing calculus.

The commentary at the hearing reflected a general acceptance of criminal history as a measure of defendant characteristics, beginning with the opening statement by Chairman Wilkins. He opined that most would agree that ‘the offender’s prior record’ is relevant to sentencing guidelines, but that there was room for debate about how it should be weighted and what should be included. With that, the Commissioners and witnesses grappled in their testimony with criminal history’s place in the forthcoming Guidelines in practical, rather than moral or ethical, terms. There was no pushback expressed in the hearing against the narrowing of defendant characteristics to criminal history alone. Rather, the focus was on the assigned task of determining how to include criminal history in the sentencing calculus. Witnesses provided disparate responses to this question, focusing especially on the availability of records that would help determine, as one of the Commissioners put it, ‘what actually happened’ in the past (Hearing Transcript, 1986: 37).

Even those most concerned with the risks to justice for defendants in the new sentencing system melded their pragmatic and justice-related critiques of criminal history, thereby blunting any blows to the carving out of such a significant role for criminal history. North Dakota Chief Probation Officer William Broome offered among the most pointed justice-related criticisms of the proposed use of criminal history, particularly about its potential racially biased effects. He also raised more pragmatic concerns about consistency and reliability, characterizing the criminal record collection process as ‘a shotgun approach’ that involved sending multiple requests and determining prior criminal history from the bits and pieces that were returned (Broome, 1986: 1). Ultimately, though, he concluded by validating the underlying criminal history logic, acknowledging the need for a ‘systematic approach to sentencing … to incapacitate those whose criminal records indicate they are serious offenders’ (Broome, 1986: 5).

Criminal history as a ‘salient factor’

This early public conversation set the course for the Commission’s operationalization of criminal history into calculable measures. That process relied heavily on past practices. Even before the SRA was imagined, criminal history had been institutionalized in the federal system at both the front and back ends of the sentencing process. At the front end, an assessment of defendants’ criminal record constituted a standard element of probation’s pre-sentence investigation and report, upon which many judges relied in determining sentences. At the back end, US Parole used criminal history as a factor in determining appropriate parole dates, which was the standard mode by which prisoners gained release from federal custody.

The future valuation system was premised upon those past uses. Specifically, the technology devised by the Commission to translate criminal history into the x-axis scale can be traced back to US Parole, and especially its adoption in 1972 of a recidivism prediction instrument known as the Salient Factor Score (SFS). The SFS instrument was a substantially pared-down measure from the broader, more inclusive assessment tools previously relied upon in parole release decision making (see Harcourt, 2007 for a detailed history). Laid out in the form of a checklist, the SFS included specific categories—such as number of prior convictions, age at first commitment, and history of drug use—which were added together to produce a score between 0 and 10.

As widespread support for indeterminate sentencing began to collapse, the Parole Commission responded to critics by revising the SFS in 1973, 1976, and 1981. The primary effect of each revision was to reduce the number of items considered as ‘salient factors’, and to privilege those factors that were considered to be most indicative of an individual’s criminal history (Hoffman, 1994). The final revision of the SFS in 1981 excluded from consideration individuals’ education level, family circumstances, employment status, and age at first offense. The rationale was to ensure that the factors relied upon were ‘objective, easily scored, not large in number, and not subject to manipulation (falsification)’ (Hoffman, 1994: 480). Another important motivation underlying the revisions was a ‘concern for fairness’, such that predictions would not be based on factors that may have been beyond the individual’s control (Hoffman, 1994: 480–481).

Federal parole, as an entity, seemed an ironic source for such a critical component of the Guidelines architecture since the SRA threatened that agency with extinction by requiring that the new sentencing system phase out parole entirely (Hoffman, 2003). Nonetheless, the SFS in essence migrated to the front-end reform project underway in the newly formed US Sentencing Commission. The first entry of the SFS came from its primary architect Peter Hoffman, then-research director at US Parole, who co-wrote a series of papers in the 1970s that sketched out a vision of a rationalized guidelines sentencing system based in part on the SFS model (Hoffman and DeGostin, 1975; Hoffman and Stone-Meierhoefer, 1977; Hoffman and Stover, 1978). Hoffman and DeGostin (1975) modeled a potential sentencing grid for the proposed system, reimagining the soon-to-be extinct parole release instruments as front-end sentencing criteria.

The SFS then made its way into the actual architecture of the Sentencing Guidelines in the wake of the contentious, floundering organizational process of drafting the Guidelines (Tonry, 1996). As the deadline for finalizing the Guidelines loomed, Hoffman moved from US Parole to become a ‘Principle Technical Advisor’ at the US Sentencing Commission in 1986 (Hoffman, 2000). Through Hoffman, the SFS became the blueprint for the Commission’s emerging instrument, the Criminal History Score (CHS) (USSC, 1986, 1987a, 1987b), which in turn became the basis of the x-axis of the Guidelines (see Hoffman and Beck, 1997 for more details).⁷

Drafting the Guidelines and finalizing the criminal history calculus

In the first half of 1986, the Guidelines could have gone in two very different directions, both in form and substance. Commissioner Paul Robinson drafted a highly detailed scheme that required ‘complex mathematical formulae involving multiplication of quartic roots’ (Commission Dissent Response, 1987: 2). This draft was circulated among select criminal law experts in July 1986, and was soundly rejected as impractical and potentially unconstitutional (Commission Dissent Response, 1987). Over the objection of Robinson, it was dropped in favor of a simpler version, ‘Preliminary Draft of the Sentencing Guidelines’, which was released for public comment and critique in September 1986.

The first observation to be made about this draft is the primacy of offense characteristics over defendant characteristics. Of the 170 pages of text, 100 pages were devoted to ‘Chapter 2: Offense Conduct’ and only 17 pages were devoted to ‘Chapter 3: Offender Characteristics’. Moreover, in the ‘Offender Characteristics’ chapter, the first five pages were devoted to individualized characteristics of the defendant’s behavior directly tied to the offense, including ‘post-offense conduct’. These included the role in the offense, whether the defendant accepted responsibility, whether his livelihood was tied to the offense, whether he cooperated as an informant, and whether he perjured himself or obstructed justice during the adjudication process. As such, these considerations were ‘offender characteristics’ only insofar as the individual is solely understood as the present criminal defendant.⁸

The ‘Offender Characteristics’ chapter also made explicit the Commission’s commitment to criminal history as the sole durable defendant characteristic relevant to sentencing in the new system. The section entitled ‘Criminal History’, comprised nearly 10 pages of the chapter and provided the only consideration of the defendant characteristics outside of the scope of the present offense. The section opened with, ‘[a] sentence adjustment for an offender’s criminal history can be justified on both just punishment and utilitarian grounds’ (USSC, 1986: 126). From a just punishment perspective, the criminal defendant who had repeatedly committed crimes was ‘more blameworthy’ than first offenders. Still, ‘crime control arguments provide a stronger justification for using criminal history to adjust a sentence’ because, according to the statement, increasing the sentence would both deter and incapacitate recidivists (USSC, 1986: 126).

The Commission went on to elaborate definitions and instructions for scoring criminal history. Like the antecedent SFS, the score was additive. The calculus also excluded from the CHS some factors that had counted in the SFS, such as the age of the offender. This was because, in contrast to parole’s singular concern with recidivism prediction, the Commission ‘views the CHS as a measure of both risk of recidivism and just punishment’ (Hoffman and Beck, 1997: 193).

The preliminary Guidelines dispensed with the remaining defendant characteristics, as delineated in the SRA, at the end of the ‘Offender Characteristics’ chapter. Pointing out that the three crime-related factors specified in the statute had been addressed in the chapter, it then admitted that ‘the other factors have not, however, been thoroughly addressed in this preliminary draft’ (USSC, 1986: 136). The chapter concluded with an invitation for public commentary on which of the additional factors ‘should be considered relevant to sentencing, and in what circumstances’ (USSC, 1986: 137).

This chapter unveiled a very specific scoring system for criminal history, even though the Commission still seemed to struggle with the fundamental question raised at the May 1986 ‘prior record’ hearing: how much should criminal history matter? Recognizing that ‘no formula exists’ (USSC, 1986: 130), the Commission instead set forth two alternative means of weighting criminal history. The first was a criminal history score table that resembled the two-dimensional grid laid out in the Parole Guidelines. In this grid, however, the criminal history score appeared on the y-axis while the ‘base offense value’ was listed on the x-axis. The alternative approach was a vestige of the alternative Guidelines model rejected just months earlier—a much more complicated three-step formulation that combined both addition and multiplication formulas based upon the offense level score as well as the criminal history points.

Finalizing the Guidelines: Criminal history’s outsized place

The Sentencing Commission’s invitation for input on the Preliminary Draft Guidelines resulted in hundreds of written comments and copious testimony at public hearings from federal judges, prosecutors, defense attorneys, probation officers, and other stakeholders. Despite the Commission’s explicit request for feedback on offender characteristics, relatively few respondents focused on that section. Those who did so offered comments primarily about technical definitions related to criminal history, such as whether to count tribal convictions as priors, and legal questions, such as the constitutionality of counting unadjudicated prior acts. Even fewer respondents commented on the unfinished business of determining whether the other non-criminally related defendant characteristics had a place in the Guidelines.⁹

Additionally, most respondents did not question the inclusion of a criminal history category, or engage in debate over the mode of quantification. For example, in 10 single-spaced pages of comments, District of Colorado Chief Probation Officer William Graves included only a passing reference to the criminal history formulation:

I do think it important to factor criminal history into sentencing calculations. Either alternative provided by the Commission would work. The use of the table format appears the easiest to use and that approach that would result in less calculation error.

(Graves, 1986: 4)

The lack of debate may be because, as United States Attorney Robert L Barr noted, criminal history was considered to be a critical factor in assessing ‘reality’ of a defendant’s situation. For Barr, the whole of the defendant’s ‘reality’ was his past and present criminal acts. ‘The underpinning of the Sentencing Guidelines appears to me very simple’, Barr wrote in his 1986 testimony to the United States Sentencing Commission.

The authorities must look at what the defendant actually did in the commission of his crime and at his or her criminal history generally. The court must base its sentencing decision on reality, not perception. The Guidelines will require the criminal justice system, composed of its various components as I have stated earlier, to look at all factors in the fresh light of reality. This is good.

(Barr, 1986: 4, emphasis in original)

By relying on an ordinal system of calculation that found its ultimate expression in a grid, the Commission was able to present a visual calculus as a seemingly straightforward starting point for taming all of the conflations and conflicts that led to sentencing reform in the first place. From there, the role of criminal history in the sentencing calculus—as the key defendant-related factor, represented by a dedicated axis of a sentencing grid—was established. Thus, even though there were a number of significant changes between the preliminary draft and the final, implemented version, the form in which criminal history entered sentencing, and the system of valuation, was set.

In the next Draft Guidelines, which circulated for public comment in January 1987, the Commission adopted the sentencing grid, rejecting the three-step process it had proposed as an alternative. It also moved criminal history to the x-axis, and recommended a range of sentences, rather than a set term. Defense attorneys at the public hearings raised concerns about the criminal history score, as they had in earlier forums, criticizing especially what they perceived as mechanisms that too quickly moved defendants into higher criminal history categories, penalized them too frequently for misdemeanors (Hearing Transcript, 1987: 181–183), and ‘buil[t] race and class bias into the Guidelines’ (Long and Stone, 1987: 15). Overall, however, the tenor of the comments and hearings in reaction to the revised Guidelines was much less critical compared to those made in response to the Preliminary Draft that was circulated the prior year. The Commission made several more changes to the proposed Guidelines, before submitting them to Congress in April 1987. Congress approved that version of the Guidelines, making them effective in November 1987 (see Table 1).

Table 1.

Evolution of the Criminal History (CH) section (1986–1987).

General features	1986 Preliminary Draft Guidelines	Jan. 1987 Draft Guidelines	Promulgated Guidelines (effective Nov. 1987)
Justification for criminal history	Just punishment and utilitarian grounds	None	References all four goals in Comprehensive Crime Control Act, and notes that repeated criminal behavior indicates rehabilitation is less likely
Sentence calculation	Grid, with CH as the y-axis. Proposes alternative sentencing table	Grid, with CH as the x-axis	Grid, with CH as the x-axis
‘Prior sentence’/ length of ‘prior sentence’	Pegged to time served, with a rebuttable presumption that one-third of the time was served	Pegged to maximum time imposed	No change
Juvenile convictions	Counted for offenses against persons	Counted based on nature and recency of offense	No change
Military sentences	Counted if there is a counterpart in civilian law	Excludes only sentences resulting from summary court martial and Art. 15 proceedings	No change
Foreign convictions	Counted if conduct would be criminal in USA	Considered only to justify departures from Guidelines	No change
Tribal convictions	Counted	Considered only to justify departures from Guidelines	No change

The promulgated Guidelines included a newly updated Criminal History section. Most notably, the Commission deleted the ‘Offender Characteristics’ chapter in its entirety, and created a separate chapter devoted to ‘Criminal History and Criminal Livelihood’. In doing so, it pronounced that no individualizing defendant characteristics, other than those tied to past or present criminality, would be integral to the new sentencing regime. In Part H of ‘Chapter 5: Determining Sentence’ the 11 defendant characteristics delineated in the SRA were each addressed in brief policy statements. It deemed eight of those—age, education, vocational skills, mental and emotional condition, physical condition including substance abuse, previous employment record, family responsibilities and community ties—’not ordinarily relevant’ to mitigate prescribed sentences. The final three—criminal history, role in the offense, and criminal livelihood—were declared relevant and the reader was directed to the appropriate chapters and sections for calculating their value.

In the new Criminal History chapter’s preface, the Commission explained that judges must consider the prior record in sentencing because ‘repeated crime behavior is an indicator of a limited likelihood of successful rehabilitation’ (USSC, 1987b: 4.1). In a supplementary report, the Commission elaborated on this justification, indicating that the CHS could serve the ‘crime control goals of general and special deterrence, and incapacitation’ and would be consistent with a ‘just deserts philosophy’ (USSC, 1987c: 41).

While the Commission drew directly from the predictive instruments created by the United States Parole Commission to develop the final criminal history score, it rejected prediction as the sole explanation for considering criminal history. Ultimately, it left out some predictive factors that were elements of the SFS, and included others that had no predictive value (Hoffman and Beck, 1997). Recognizing an inherent difficulty in simultaneously looking both backwards and forwards, the Commission proclaimed its hope that the criminal history measure, as structured, would ‘limit[] the tension’ between the predictive and punitive philosophies (USSC, 1987c: 43).

The Commission also appended an entirely new section to the Criminal History chapter, to account for the so-called ‘career offender’. The Commission justified this section—although not its late insertion—with reference to the SRA’s mandate to specially treat career offenders. In this way, enhancement characteristics that had been deleted from the criminal history formula in its transformation from the SFS were reinserted back into the Guidelines. And because ‘career offenders’ were automatically assigned to the highest criminal history category, the carefully planned x-axis of the sentencing grid did not even apply. After going to such lengths to create and justify the CHS, the Commission created an entire category—one that perhaps would contain the defendants most amenable to risk prediction—to which such calculus was almost entirely inapplicable.¹⁰

The measurement of the (criminal) man: Quantifying criminal history

While not a point of discussion in the opening chapter of the Guidelines’ final version, criminal history was, as we have shown, a central element to the overall punitive calculus from the beginning of the construction project to its completion. Its role in that regard grew exponentially as the Commission shed other ‘defendant characteristics’ as irrelevant to the sentencing endeavor. Ultimately, it came to be represented as one full axis on the sentencing grid.

The Guidelines quantified the prior record through two logics of time: prior sanction lengths as indicators of history severity, and the temporal relation of past acts to the current conviction. Time, though, was a practical constraint underlying the formulation of criminal history during Guidelines construction. The technical record-keeping capabilities of the period itself (the mid-1980s) dictated to some degree how long, and what types of records were reliably accessible, and these constraints were then codified into the Guidelines as penal principles. The 1986 ‘Prior Record’ hearing provided an early indication that time was both a problem of, and solution to, quantifying criminal history. Concerns over how far back the Commission should reach in counting the criminal history seemed to meld the practical limits in obtaining records and the purposes, for just sentencing, of relying upon such records.

There was considerable discussion in this early hearing of the appropriate ‘decay period’ in regard to criminal history. Although the witnesses all agreed that past conduct was relevant in assessing punishment for the current offense, they could not agree on what period of time must elapse before a record should no longer count as part of a defendant’s criminal history or why a particular period served the goal of sentencing under this new system. US Attorney William Weld settled on 15 years as ‘fairly reasonable to suppose that a person may have straightened up their act’ (Hearing Transcript, 1986: 8), while Federal Defender Thomas Hillier noted that the Parole Guidelines and many state jurisdictions used 10 years as the cut-off for criminal history. Despite significant efforts to develop a justifiable rationale for the components of the criminal history axis, it was, as one Commissioner remarked during a hearing, as if they were simply ‘pulling time out of the air’ (Hearing Transcript, 1986: 61).

By the final iteration of the Guidelines, these issues were settled. Several sub-sections of the Criminal History chapter provide the specific formula for determining how prior records will count in the calculation of a present sentence. The prior record is first converted into points on the basis of prior sentence lengths (maximum sentenced imposed, not time actually served). The points are then grouped into one of the six ‘criminal history categories’ that comprise the x-axis of the sentencing grid:

(a) Add 3 points for each prior sentence of imprisonment exceeding one year and one month.

(b) Add 2 points for each prior sentence of imprisonment of at least sixty days not counted in (a).

(c) Add 1 point for each prior sentence not included in (a) or (b) up to a total of 4 points for this item. [These include sentences of probation, and/or short jail sentences for misdemeanors and other petty offenses.]

(USSC, 1987b: 4A1.1)

Under this formula, qualitatively different prior convictions—which may well have very divergent values for assessing culpability or predicting risk—get homogenized in an ordinal system. For instance, under these provisions, a defendant with a prior homicide conviction with a sentence of more than 13 months will have an equivalent criminal history score (three points) as someone who has three prior driving-on-a-suspended-license convictions that resulted in either probation terms of a year or more or maximum jail sentences of 30 days or more (also three points). Such equivalence between seemingly incommensurate crimes reflects a particular judgment about the relation between culpability and time (see Lee, 2014), wherein failure to change over time was to be treated as harshly as a crime of violence that occurs only once. Indeed, the Commission appears to have adopted precisely this argument in revising the criminal history formula for the soon-to-be enacted Sentencing Guidelines.

The second use of time in the criminal history formulation adds to the sentence based largely on temporal relation of the instant offense to prior convictions. So §4A1.1(c) specified an additional two points for defendants who committed the instant offense while under any previous criminal justice sentence, and §4A1.1(d) added two points if the instant offense occurred within two years of any sentences counted under §4A1.1(a) or (b) (this second ‘recency’ provision was eliminated in 2010). The chapter goes on to define the ‘decay’ periods as 15 years for priors resulting in 13 months or longer maximum sentences (with the clock to begin after release from incarceration) and 10 years for all other ‘countable’ offenses. As such, it implicitly adopted a kind of culpability distinction based on the 13-month line that it devised as meaningful in this calculus.

While it might be easy to conclude that criminal history rose to such significance in the Guidelines formulation because it can be converted into measurable units of time (length and lag) that can be standardized across persons and jurisdictions, this does not explain why other, easily convertible sentencing factors were left behind. Age was the first factor listed in the SRA for consideration, which would need no conversion to have been be included; defendant’s educational attainment, also specified in the statute (28 USC §994(d)(1–2)), is a frequently quantified measure in many empirical projects. Employment history, a third factor suggested in the SRA and rejected by the Commission, could be quantified much along the same lines as criminal history, using time employed or stretches of unemployment as measures of meaning (28 USC §994(d)(6)). Like criminal history, each of these are fraught with interpretive problems as to how and why they should matter in sentencing a criminal defendant, but the Commission hardly grappled with these issues as it propelled criminal history to the center of the sentencing calculus.¹¹

Conclusion

The record we have unearthed here suggests that the primacy of prior criminal history in the Guidelines sentencing calculus reflects the continuity of the past in shaping a new future. There was a widely acknowledged and accepted place for criminal history as a window into the individual defendant, which may be why there was so little debate about whether it belonged at all. The time conversion formula thus became a technocratic mode of making a fitting place for criminal history within a system that was to provide ‘honesty…, uniformity…, [and] proportionality’ in sentencing (USSC, 1987b: 1.2).

Our examination therefore calls into question just how revolutionary the Guidelines were, at least in regard to both the conceptual and technical aspects of the criminal history component. We have shown the deeply rooted logics underpinning the use of criminal history in punishment determinations in this system, and highlighted the technical harbingers of the CHS in instruments long developed and refined at the back-end of the punishment process (see also Harcourt, 2003, 2007). This does not mean that the effects of the Guidelines were not dramatic; they were. The distinction we make is that the tools used to catalyze those dramatic effects are neither as new nor as untethered to historical techniques and practices as is imagined in more macro-level accounts of penal change.

Moreover, the Commission did not reject the SRA’s mandate to incorporate a consideration of the individual defendant in drafting the Guidelines. It constructed an elaborate and individualized assessment structure, but one that is narrowly drawn. In the authorized version of the Guidelines, relevant defendant characteristics are calculated through provisions delineated both in Chapter 4, devoted to criminal history and criminal livelihood, and Chapter 3, parts B, C, and E, which provide for adjustments based on role in the offense, obstruction, and acceptance of responsibility. Each defendant is thus individualized across multiple measures of past and present bad acts. As such, our observations concur with Harcourt’s assessment that the late 20th-century shift in penal rationales is more the effect of actuarialism than the cause. In the case of the federal Guidelines, the ‘will to know the criminal’ (Harcourt, 2007: 173) has driven the technology. Indeed, the defendant as criminal, past and present, is known with precision, and that knowledge plays a major determinative factor in his treatment at sentencing.

The brief account of criminal history that we have presented here also makes plain that the valuation of time and history is not merely a mechanical—or even logical—computational exercise. To be sure, in determining the appropriate sentence, the law purports to punish a particular kind of offender, one who is no more than a sum of past and present acts. However, history is not simply the cumulative result of past and present, but, as we have shown, it also can embody an imagined future that exerts an outsized influence aimed at averting an anticipated risk or expected failing of the to-be-punished defendant. This concept of time—a specter of the past along with an anticipated future that together casts a long shadow over the present—was itself haunted by an ostensibly outdated parole system whose logic and techniques suffuse what now serves as the all-important x-axis in the Sentencing Guidelines grid.

By positing time as an unnatural feature of the new sentencing system we can consider the particular ways that criminal history came to be incorporated into this legal scheme—cloaked by seemingly technical, but hardly mechanistic, assessments of people and things. In this sense, then, there exists a ‘political economy of time’ (Cohen, 2014), through which time is a currency in making and marking democratic citizenship. In the Guidelines formulation, though, time is negative currency—what we might call ‘bad time’. For the legal subject in sentencing proceedings under this scheme, the more time spent involved with the criminal justice system in the past triggers a diminution of citizenship through lengthier periods of removal from the polity in the future.

We conclude by urging a skeptical view of the element of time as a static, autonomous force in theoretical accounts of punishment. Rather than proposing grand theories neatly packaged in linear temporal narrative, we should instead consider new ways to open up examinations of how ‘knowledges and techniques are put to use, and given meaning in context’ (Garland, 1997: 199), including how criminal history—and its imposition of time as a mode of calculability—constitutes a technique by which penal subjects are known. Doing so requires not only greater attention to the historical processes that have produced modern punishment but also more critical attention to the manipulation of time by those historical actors.

Footnotes

Notes

Author biographies

Mona Lynch is Professor of Criminology, Law & Society and Co-director of the Center in Law, Society & Culture at the University of California, Irvine. Her current research focuses on criminal adjudication and punishment, with a focus on how race intersects with criminal justice processes in the USA.

Alyse Bertenthal is a PhD Candidate in Criminology, Law & Society, University of California, Irvine.

References

American Friends Service Committee (1971) Struggle for Justice. New York: Hill & Wang.

Baron-Evans

Coffin

Silva

(2010) Deconstructing the career offender guideline. Charlotte Law Review 2: 39–99.

Barr

(1986) Remarks before the US Sentencing Commission Regional Hearing, Atlanta, GA, 29 October.

Bownes

(1986) Letter to William W Wilkins, Jr, US Sentencing Commission, 4 November.

Breyer

(1988) The Federal Sentencing Guidelines and the key compromises upon which they rest. Hofstra Law Review 17(1): 1–50.

Broome

(1986) Letter to Judge Patrick A Conmy, US Sentencing Commission, 24 November.

Clark

(1970) Crime in America: Observations on Its Nature, Causes, Prevention and Control. New York: Simon and Schuster.

Cohen

(2014) The political economy of time: Political time, citizenship and rights. Paper on file with authors.

Commission Dissent Response (1987) Preliminary observations of the Commission on Commissioner Robinson’s dissent, 1 May.

10.

Durham

(1987) Justice in sentencing: The role of prior record of criminal involvement. Journal of Criminal Law & Criminology 78(3): 614–643.

11.

Feeley

Simon

(1992) The new penology: Notes on the merging strategy of corrections and its implications. Criminology 30(4): 449–474.

12.

Fine

(2012) Tiny Publics: A Theory of Group Action and Culture. New York: Russell Sage Foundation.

13.

Frase

(1999) Sentencing guidelines in Minnesota, other states, and the federal courts: A twenty-year retrospective. Federal Sentencing Reporter 12(2): 69–82.

14.

Frase

(2005) State sentencing guidelines: Diversity, consensus, and unresolved policy issues. Columbia Law Review 105(4): 1190–1232.

15.

Garland

(1997) Governmentality and the problem of crime: Foucault, criminology, sociology. Theoretical Criminology 1(2): 173–214.

16.

Graves

(1986) Letter to William W Wilkins, Jr, US Sentencing Commission, 9 May.

17.

Hannah-Moffat

Lynch

(2012) Theorizing punishment’s boundaries: An introduction. Theoretical Criminology 16(2): 119–121.

18.

Harcourt

(2003) From the ne’er-do-well to the criminal history category: The refinement of the actuarial model in criminal law. Law & Contemporary Problems 66(3): 99–151.

19.

Harcourt

(2007) Against Prediction: Profiling, Policing, and Punishing in an Actuarial Age. Chicago, IL: University of Chicago Press.

20.

Hearing Transcript (1986) US Sentencing Commission Public Hearing on the Treatment of Prior Criminal Record, 22 May.

21.

Hearing Transcript (1987) US Sentencing Commission Public Hearing, 11 March.

22.

Hoffman

(1994) Twenty years of operational use of a risk prediction instrument: The United States Parole Commission’s salient factor score. Journal of Criminal Justice 22(6): 477–494.

23.

Hoffman

(2000) Simplifying the US Sentencing Commission’s offense scale. St Louis University Law Journal 44(1): 365–376.

24.

Hoffman

(2003) History of the federal parole system. Report for the US Parole Commission.

25.

Hoffman

Beck

(1997) The origin of the federal criminal history score. Federal Sentencing Reporter 9(4): 192–197.

26.

Hoffman

DeGostin

(1975) An argument for self-imposed explicit judicial sentencing standards. Journal of Criminal Justice 3(3): 195–206.

27.

Hoffman

Stone-Meierhoefer

(1979) Post release arrest experiences of federal prisoners: A six-year follow-up. Journal of Criminal Justice 7(3): 193–216.

28.

Hoffman

Stover

(1978) Reform in the determination of prison terms: Equity, determinacy, and the parole release function. Hofstra Law Review 7(1): 89–121.

29.

Jones

(1986) Letter to Michael K. Block, US Sentencing Commission, 6 March. George E MacKinnon Papers, Minnesota Historical Society.

30.

King

(2014) Sentencing and prior convictions: The past, the future, and the end of the prior conviction exception to Apprendi. Marquette Law Review 97(3): 523–598.

31.

Knapp

Hauptly

(1992) State and Federal Sentencing Guidelines: Apples and oranges. UC Davis Law Review 25(2): 679–694.

32.

Lee

(2014) Repeat offenders and the question of desert. In: Roberts

Von Hirsch

(eds) Previous Convictions at Sentencing: Theoretical and Applied Perspectives. Oxford: Hart, 73–90.

33.

Long

Stone

(1987) Comments of the Public Defender Service for the District of Columbia on the Revised Draft Sentencing Guidelines, 16 March.

34.

Mitford

(1973) Kind and Usual Punishment: The Prison Business. New York: Random House.

35.

Murakawa

(2014) The First Civil Right: How Liberals Built Prison America. Oxford: Oxford University Press.

36.

Nagel

(1990) Structuring sentencing discretion: The new Federal Sentencing Guidelines. Journal of Criminal Law and Criminology 80(4): 883–943.

37.

Offender Characteristics Committee (1986) List without narrative: Offender characteristics. George E MacKinnon Papers, Minnesota Historical Society.

38.

Paternoster

Bynum

(1982) The justice model as ideology: A critical look at the impetus for sentencing reform. Crime, Law and Social Change 6(1): 7–24.

39.

Rappaport

(2003) Rationalizing the Commission: The philosophical premises of the US Sentencing Guidelines. Emory Law Journal 52(2): 557–644.

40.

Roberts

(1997) The role of criminal record in the sentencing process. Crime & Justice 22: 303–362.

41.

Roberts

Yalincak

(2014) Revisiting prior record enhancement provisions in state sentencing guidelines. Federal Sentencing Reporter 26(3): 177–190.

42.

Robinson

(nd) Memo to Offender Characteristics Committee. George E. MacKinnon Papers, Minnesota Historical Society.

43.

Robinson

(1987) Memorandum to all Commissioners: Draft dissenting view of Commissioner Paul H Robinson, 23 April.

44.

Savelsberg

(1992) Law that does not fit society: Sentencing guidelines as a neoclassical reaction to the dilemmas of substantivized law. American Journal of Sociology 97(5): 1346–1381.

45.

Sparks

(2001) Degrees of estrangement: The cultural theory of risk and comparative penology. Theoretical Criminology 5(2): 159–176.

46.

Stith

Cabranes

(1998) Fear of Judging: Sentencing Guidelines in the Federal Courts. Chicago, IL: The University of Chicago Press.

47.

Stith

Koh

(1993) The politics of sentencing reform: The legislative history of the Federal Sentencing Guidelines. Wake Forest Law Review 28(2): 223–290.

48.

Tata

(2007) Sentencing as craftwork and the binary epistemologies of the discretionary decision process. Social & Legal Studies 16(3): 425–447.

49.

Tonry

(1996) Sentencing Matters. New York: Oxford University Press.

50.

Tonry

(2005) Obsolescence and immanence in penal theory and policy. Columbia Law Review 105(4): 1233–1275.

51.

Tonry

(2014) The questionable relevance of previous convictions to punishments for later crimes. In: Roberts

Von Hirsch

(eds) Previous Convictions at Sentencing. Oxford: Hart.

52.

US Sentencing Commission (USSC) (1986) Preliminary Draft. Washington, DC: US Sentencing Commission.

53.

US Sentencing Commission (USSC) (1987a) Revised Draft Sentencing Guidelines. Washington, DC: US Sentencing Commission.

54.

US Sentencing Commission (USSC) (1987b) Guidelines Manual. Washington, DC: US Sentencing Commission.

55.

US Sentencing Commission (USSC) (1987c) Supplementary report on the initial sentencing guidelines and sentencing statements. Washington, DC: US Sentencing Commission.

56.

US Sentencing Commission (USSC) (2004) Fifteen years of Guidelines sentencing: An assessment of how well the federal criminal justice system is achieving the goals of sentencing reform (November). Washington, DC: US Sentencing Commission.

57.

US Sentencing Commission (USSC) (2013) Federal Guidelines Manual. Washington, DC: US Sentencing Commission.

58.

Von Hirsch

(1985) Past or Future Crimes: Deservedness and Dangerousness in the Sentencing of Criminals. Manchester: Manchester University Press.

59.

Von Hirsch

(2104) Proportionality and the progressive loss of mitigation: Some further reflections. In: Roberts

Von Hirsch

(eds) Previous Convictions at Sentencing. Oxford: Hart.