Understanding risk in the context of the Youth Criminal Justice Act.

• Author: Maurutto, Paula
• Position: Canada - Sentencing and Risk Assessment

Introduction

Over the past decade, the youth criminal justice system has undergone a complex set of changes, including the introduction of the 2003 Youth Criminal Justice Act (YCJA), the revitalization of rehabilitation and support services within the community, and the integration of restorative approaches. An equally important, yet lesser-known, development has been the increased use of empirically based risk/ need assessments at almost every decision point within the youth criminal justice system. Risk technologies are invoked both overtly and indirectly within the courtroom. While historically developed for post-sentencing placement and treatment purposes, risk assessments are being deployed in the legal arena to inform a range of decisions, including judicial interim release, extrajudicial measures, and sentencing.

This article explores and analyses the entry of empirically based risk assessments into the legal arena, with a particular focus on youth pre-sentence reports (PSRs) in Canada. The PSR, previously referred to as a pre-disposition report in youth court and as a social-inquiry report or pre-sentence investigation in the international literature, is a report prepared by a probation officer, typically at the request of a judge. The report is used by judges to assist in determining sentencing outcomes. Within the past five years, 9 of Canada's 13 provincial and territorial jurisdictions have adopted some form of risk assessment in the preparation of youth PSRs; a number of others are following this trend (Hannah-Moffat and Maurutto 2003). Arguably, this integration of risk tools alters the type and presentation of information contained in the PSR; hence, there is a need to consider the effects of risk frameworks on the judicial system. Preliminary work, for example by Silver and Miller (2002), suggests that the use of actuarial tools in sentencing could amount to "statistical justice," wherein dispositions are determined on the basis of how closely an offender matches an actuarial profile, with less significance being given to other relevant legal criteria. To date, little systematic research has explored the extent or impact of the use of risk/need information for sentencing purposes. Increasingly, the judiciary, prosecutors, and defence lawyers are raising ethical, scientific, and evidentiary concerns about the appropriateness of risk-based sentencing and the defensibility of risk information generated through the use of actuarial risk-assessment instruments (Ostrom et al. 2002; Vigorita 2003; Kern and Farrar-Owens 2004). We build on this literature by examining how this largely uncontested development could inadvertently restructure youth-sentencing practices. More specifically, our concern is with how formal, empirically based risk tools are embedded in the PSR and with the possible implications of this practice for sentencing.

The article raises a number of legal, ethical, and empirical questions related to the incorporation of risk-assessment information in the preparation of youth PSRs. In particular, we examine whether the classification of criminogenic need and past behaviour in commonly used risk tools contradicts the YCJA's legal mandate to emphasize proportionality in sentencing. Moreover, we consider if current risk-assessment technologies produce legally defensible scientific evidence or if they introduce speculation and morally laden subjective assessments that reflect white, Western middleclass judgements. Many risk indicators are associated with socioeconomic marginalization. Consequently, certain minority groups, such as women and Aboriginal youth, may be more likely to be classified as higher risk because of their greater need. This could result in more intrusive and punitive dispositions for marginalized youth.

This article draws on a combination of case-law review, detailed analysis of risk tools and manuals, and a total of 43 interviews with probation officers, defence lawyers, and Crown prosecutors in 11 provinces and territories. Where possible, we collected and reviewed risk/need assessment tools currently used in the Canadian youth justice system, along with user manuals and/or related policy documents. A subset of interviews was drawn from a larger project conducted in 2002 and 2003, in which semi-structured, open-ended interviews were conducted in all provinces and territories, with the exception of Quebec (see Hannah-Moffat and Maurutto 2003). Key government policy informants in each of the provinces and territories were contacted and asked for an interview. These regional contacts were interviewed and/or asked to provide us with additional names in research/policy, Crown's offices, and probation. The intent of the data is to explore the compatibility of risk assessments with the sentencing principles of the YCJA and their potential for producing disproportionate sentences, particularly for high-need and marginalized populations.

Inserting risk into PSRs

Pre-sentence reports have not been the object of much local or international scholarship in recent years, despite evidence of their significant impact on judicial decisions. Existing international research focuses on the quality and effectiveness of PSRs (Gelsthorpe and Raynor 1995), on the content of the PSR (Stone 1991), and on the impact on sentencing. Canadian studies conducted in the 1970s and 1980s indicated that PSRs do influence the nature and conditions imposed during sentencing. These studies reported as high as an 80% correlation between PSR recommendations and dispositions (Hagan 1975; Boldt, Hursh, Johnson, and Taylor 1983). A more recent study by Bonta, Bourgon, Jesseman, and Yessine (2005: 16), while not specifically exploring the concordance between PSRs and sentencing, reported 87.4% satisfaction with PSRs for judges and found that 68% of judges found them useful. At the international level, several Western countries have also documented high levels of association between PSR recommendations and judicial decision making, ranging from 92% in the United States (Norman and Wadman 2000), to 78% in the United Kingdom (Thorpe and Pease 1976), and in the range of 77 to 80% in New Zealand (Gibson 1973; Deane 2000). These high correlation levels suggest that PSRs are important and do affect the sentencing process. Consequently, any changes in the format and method of preparing PSRs may have significant repercussions on dispositions.

With few exceptions, the literature has not examined the extent to which PSRs have been restructured to incorporate risk-assessment information. PSRs have typically constructed a narrative profile of offenders, including demographic data on age and gender, evidence of previous contact with the police or the courts, and information about current and previous charges and convictions. The YCJA, under section 40(2), introduced a series of guidelines for the compilation of PSRs. Section 40(2) is intended to limit the scope of information made available to judges to matters directly relevant to the new sentencing principles. When the section is properly applied, information irrelevant to YCJA sentencing objectives, such as the overall criminogenic needs of young offenders ought not to be included. Instead, PSR information should focus on the circumstances of the specific offence before the court. Accordingly, the PSR may document "the results of an interview with the young person," parents, and, where reasonable, the victim. It can include pertinent information on "age, maturity, character, behaviour and attitude," "willingness to make amends," recommendations from conferences, and "history of previous findings of delinquency" and/or participation in extrajudicial sanctions (previously referred to, under the YOA, as "alternative measures"). Where warranted, PSRs can also make reference to school or work attendance and performance and document the relationship between a youth and his or her parents. They should also identify "plans put forward by the young person to change his or her conduct or to participate in activities." In preparing PSRs, some probation officers include recommendations to the courts and/or information pertaining to the availability and appropriateness of community services and facilities. This information, however, is only to be included "to the extent that it is relevant to the purpose and principles of sentencing." Hence, only information relevant to the specific offence should be recorded. Yet, the inclusion of risk-assessment information in PSRs raises questions regarding the extent to which information that is not directly pertinent to the offence but rather is a prediction of future behaviour based on a narrowly defined set of criminogenic needs and static factors ought to be used to inform dispositions.

Risk assessments can refer to a range of different practices, including intuitive judgements of the judiciary, probation officers, or clinicians as well as actuarial statistical assessments. For the purposes of this article, we focus on the use of empirically based tools that generate a profile of the offender based on a predefined set of risk factors and criminogenic needs. Within this context, risk assessments are premised on an insurance model, wherein probabilistic calculations are used to determine the likelihood of future recidivism. Considerable variation exists throughout the country with respect to the specific risk tools used, the type of information recorded, and the manner in which risk is restructuring PSR information. Most provinces have adopted some version of the Level of Service Inventory-Revised (LSI-R) developed by Andrews and Bonta (1995; 2003); for instance, the Youth Level of Service Case Management Inventory (YLS/CMI). Others have developed their own tools; for example, British Columbia pioneered the Youth Community Assessment Risk/Need Tool; Manitoba and Nunavut are using the Offender Risk Assessment and Management System...