The risk-need-responsivity model of assessment and human service in prevention and corrections: crime-prevention jurisprudence.

• Author: Andrews, D.A.
• Position: Canada - Sentencing and Risk Assessment

This article addresses issues having to do with risk assessment and justice processing in the context of the Risk-Need-Responsivity (RNR) model of crime prevention and correctional rehabilitation (Andrews, Bonta, and Hoge 1990; Andrews, Zinger, Hoge, Bonta, Gendreau, and Cullen 1990; Gendreau 1996; Andrews 2001; McGuire 2004). RNR is closely associated with a general-personality and cognitive-social-learning theory of criminal behaviour. One version of this theoretical approach is known as the Personal, Interpersonal, and Community Reinforcement (PIC-R) perspective (Andrews and Bonta 2003; Andrews 2006).

The holistic and truly interdisciplinary general-personality and social psychology of human behaviour that is associated with RNR is developing at quickly in all three major spheres: research, theory, and application (generally, see Baumeister and Vohs 2004; in regard to crime, see McGuire 2004: chs. 3, 4). This general-personality and social psychology is impressive, but "as is true of all of psychology, [it] remains a work in progress" (Carver 2004: 33). Incomplete without a doubt, but RNR and PIC-R provide a relatively powerful model of crime and crime prevention that some believe to be the pre-eminent evidence-based model for rehabilitative programming in youth and adult corrections (McGuire 2004; Ogloff and Davis 2004).

The National Institute of Corrections (2004) and the International Community Corrections Association (2) are also strong proponents of RNR. Even clinical psychologists who call for increased attention to the personal fulfilment of offenders (Ward, Melser, and Yates 2007: 3) say that RNR "has constituted a revolution in the way criminal conduct is managed in Canada, Britain, Europe, Australia and New Zealand." Revolution or not, in two decades, the field has moved from "nothing works" through "what works" to now "making what works work" (Andrews 2006).

The Canadian Parliament has made it clear that rehabilitation is a primary concern at sentencing, albeit not necessarily the paramount concern (Department of Justice 2003; 2005). Because we are talking about a model of crime and corrections, this article is not just another summary of the conditions under which risk assessments predict recidivism. Rather, we consider RNR and cognitive-social-learning theory as a guide to enhancing the contributions of the legal and court systems to crime prevention.

This might be called an exercise in therapeutic jurisprudence (Birgden 2004; McGuire 2004), but we are not comfortable with the clinical language of forensic mental health. We are not against the law and justice systems' being used to "enhance well-being" or to promote "therapeutic effects." However, confusion of process (therapy) and outcome (enhanced well-being) is problematic. Indeed, in many personal, interpersonal, and social domains, therapy is perhaps one of the least common and least powerful of routes to enhanced well-being relative to achievement and fulfilment. Additionally, reduced criminal victimization (i.e., crime prevention) is itself a more obvious source of personal well-being for both offenders and victims than is therapy. Moreover, successful enhancement of the well-being of criminals may well increase crime if other major causal variables for crime remain untouched (recall Stephen Wormith's [1984] careful analysis of the conditions under which self-control and self-esteem may interact with antisocial cognition in relation to crime). Most readers can appreciate the image of a well-adjusted, happy, and personally fulfilled criminal (the image of a psychopath comes easily to mind). Overall, the phrase therapeutic jurisprudence does not sufficiently underscore the importance of the objective of reduced offending in the context of law and justice, on the one hand, relative to the health and mental-health objective of enhanced well-being through therapy, on the other. Reducing victimization is a worthy primary goal that shouldn't be understood as incidental or secondary to personal fulfilment. (Astrid Birgden [2004] provides an interesting discussion of RNR and human well-being.)

We prefer the language of personality and social psychology over that of clinical forensic mental health because it readily encompasses therapeutic effect without missing the full range of biological, personal, interpersonal, and social processes involved in influencing human behaviour (be it criminal or non-criminal). Thus, we prefer the phrase rehabilitative jurisprudence (RJ) when the focus is the ethical, legal, decent, humane, efficient, and just pursuit of reduced offending. However, restorative justice has already claimed the "RJ" designation. (3)

We will call our approach crime-prevention jurisprudence (CPJ). According to CPJ, a purpose of the courts is to apply sanctions in a manner consistent with legislation governing sentencing and to do so with attention to reducing offending through applications of RNR and an evidence-based, interdisciplinary understanding of the psychology of the criminal behaviour of individuals. We explore the idea that judges and the prosecuting and defence attorneys (like parents, teachers, friends, correctional workers) may reduce, increase, or have no effect on criminal behaviour, depending upon their actions with respect to those accused and/or found guilty. The actions of judges and lawyers, the court process, and the sanction itself may affect causally significant factors and thereby increase or decrease re-offending. What do RNR and PIC-R have to say about enhancing crime-prevention potential?

To answer this, we proceed in four steps. First, the underlying general-personality and cognitive-social-learning model in the form of PIC-R is briefly reviewed. We do this because--in the fields of forensic mental health, criminology, and criminal justice--psychology is sometimes thought to refer only to personality and mental illness. Second, the RNR model of prevention and rehabilitation is described. Third, the research evidence on the validity of RNR is reviewed. Finally, applications at the court level and at time of sentencing are briefly discussed. We do not emphasize here the implications for primary prevention and policing, although applications are readily evident.

The PIC-R perspective

Human beings are active, conscious, and wilful, and they are goal-oriented. They are also creatures with biological predispositions and with habits and conditioning histories whereby repeated associations among stimuli, responses, and behavioural outcomes can produce automatic, non-conscious cognitive regulation of motivation, perception, and behaviour. Their behaviour is under personal control, interpersonal control, and automatic control.

Some aspects of personally mediated control envision active and conscious persons who monitor their own behaviour, compare that behaviour with standards of conduct, and self-regulate through active and deliberate self-instructions, through the exercise of coping strategies, and through self-delivery of rewarding and/or costly consequences. Predicting or influencing the impact of this type of self-regulation on criminal activity requires two key sets of knowledge.

First, we must understand the self-control skills of individuals, as they provide a direct indication of their ability to monitor, evaluate, and deliver self-instructions, cope with temptations, and self-deliver meaningful consequences. Second, the exercise of self-control will be favourable, unfavourable, or neutral with reference to criminal activity, depending on the standards of the individual. These standards are assessed through an examination of personal attitudes, values, beliefs, rationalizations, cognitive-emotional states, and identities supportive or not supportive of criminal activity. The anti-criminal or pro-criminal nature of standards is also influenced by the anti-criminal, pro-criminal, or neutral position of significant others in regard to antisocial behaviour. People are influenced by how individuals or groups who are important to them, such as mother, brother, sister, employer, or friends, are perceived to think and feel about criminal behaviour (would they approve, disapprove, or not care?).

Antisocial associates contribute not only to personally mediated control, as noted earlier, but also to interpersonally mediated control. The direct actions of others may signal the appropriateness of particular actions, enable some actions, and function directly as rewards or costs. Social support for criminal activity is most directly measured through assessments of the quality of the interpersonal relationships with pro-criminal others and the degree of isolation from anti-criminal others.

Through temperament and through reinforcement history, some of the rewards and costs from crime become relatively automatic or habitual. Non-mediated control is best assessed through examination of a history of antisocial behaviour; in particular, involvement from an early age in a number and variety of antisocial activities. Temperamentally, assess for a taste for the risky and exciting, for a certain readiness to upset others, a readiness to feel mistreated, and for weak constraint (going for the relatively quick and easy route to the rewards).

The general-personality and cognitive-social-learning perspective recognizes that the behaviour of human beings has multiple sources of influence and that motivations (potential reinforcements) and controls (potential costs) operate simultaneously. In order to understand the sources of cognitions, self-control skills, and association patterns favourable to crime, it is best to consider the main behavioural settings of family, marital relationship, school/work, and leisure/recreation. The key issues within those settings are what behaviour is being modelled (pro-criminal, neutral, anti-criminal) and what behaviour is being reinforced, punished, or ignored. Social location, age...