Faye Taxman, Ph.D.
Virginia Commonwealth University
The Standardized Risk and Needs Assessment Tool Dilemma
Clarifying the Concepts of Risk & Dynamic
Risk Factors: Actuarial
Substance Abuse
Family Dysfunctional
Peer Associates
Criminal Personality
Antisocial Cognitions (Attitudes/Orientation/Thinking)
Low Self Control
Mental Health, Self-Esteem, Low Educational Attainment, Employment & Other Factors
Applying Rule #1 in Correctional Agencies
Conclusions
RULE NUMBER ONE in EBP (evidence-based practice) is that high-risk offenders should be placed into appropriate treatment services, and that low- and moderate-risk offenders should not receive the same intensity of services. (Note: The use of the term “services” here includes both treatment and control techniques.) While this may seem like a simple concept, it encompasses the following: 1) use actual risk assessments; 2) use dynamic criminogenic needs; 3) adopt responsivity or matching strategies to link offenders to services and controls; and 4) administer heterogeneous programs that address the myriad of offender issues. The goal is to combine all of these together as a supervision plan that identifies the goals and specifies expectations for the offender. These expectancies become the binding agreements that define the criteria for being successful on supervision. Stated simply, assessment is not just a stand-alone process; instead, it is a process that should lead to the goal of a supervision plan that is designed to change the behavior of the offender.
Assessment should be the beginning of the correctional process. Of course in contemporary criminal justice practice, it can occur at a number of points, including arrest and pretrial detention, sentencing, intake to probation/parole or prison, and so on. In other words, it can occur in numerous places, all with slightly different goals—at pretrial to determine risk of flight or danger to society, at sentencing to determine the appropriate punishment and/or placement, at prison to determine security levels, and at probation/parole to determine risk to the community. In all of these calculations, the goal of the assessment is to inform decisions about the degree of restrictions that an offender should be given based on the offender’s history and seriousness of the current offense. The assessment can also contribute to what is traditionally referred to as a “treatment plan,” or more specifically the corrective action plan to help the offender become a productive citizen and contributing member of society. As noted recently by Ed Latessa and his colleagues (2002), corrections practice today seldom ties the assessment to a plan for the offender. Instead the plan for the offender is generally made based on judicial or parole board decisions.
The most frequent stumbling block is an understanding of the core elements that are embedded in EBP Rule #1, and how to apply these elements in practice. That is, with the tools that are available, often there is a misunderstanding of the concepts of risk and needs. Often the terms “static” and “dynamic” are inappropriately interchanged with risk and needs. Risk refers to the actuarial (or statistical) likelihood that an offender will have further criminal behavior. Dynamic refers to the dimensions of the person’s functionality that, if improved, can affect their involvement in criminal behavior. A clarification of these concepts is the main goal of this paper, with an eye on trying to clarify how best to use these concepts in correctional practice.
The Standardized Risk and Needs Assessment Tool Dilemma
Don Andrews and his colleagues (2006) recently provided a historical review of the concept of objective assessment tools for the criminal justice system. The review detailed the generational development of assessment in the various phases of the criminal justice system over the last 80 years: clinical assessments of offenders’ risk to the community with some emphasis on treatment planning, actuarial risk assessment to assess the likelihood of further criminal behavior, actuarial risk assessment combined with dynamic variables to better guide treatment planning, and actuarial risk assessment tools supplemented by problem-specific tools. The development of standardized tools for the field has accompanied various needs in the criminal justice system, including classification, treatment planning, release decisions (from prison, jail, or parole), and sentencing. Essentially, assessment tools have been developed and used for various purposes, which adds to the complication of how to use the tool(s). Some tools are designed merely to identify risk factors related to certain decisions, while others are designed to identify the factors or needs that, if altered, an improve offender outcomes.
As is true for other fields, and as other articles in this issue of Federal Probation note (see papers by Austin and Harris), a major point of discussion in the criminal justice field has been the value of standardized assessment tools compared to that of subjective assessments by counselors and other correctional staff. The preference for subjective assessment is a long-standing issue in the field (as well as in psychology, education, and other disciplines), since professionals feel confident in their decision-making skills, and do not want to succumb to a paper-pencil test. But as discussed by Harris (2006) in another article of this edition, research persists in demonstrating that standardized objective tools enhance decision making, besides providing institutional safeguards against discretionary, biased, or inappropriate decisions. The use of standardized tools minimizes the potential for bias to be introduced into the decision making process by such human factors as the staff person being influenced by the dress, mannerisms, and/or attitude of the offender, in addition to such obvious factors as gender, class, and race.
Of course, the clinical vs. objective tool debate is an overstatement of the relative advantage of an interview with the offender. Third and fourth generations of assessment tools are accompanied by an interview (clinical in nature with “clinical” referring to an interview to collect information from the offender in a manner conducive to assessing the offender’s risk and needs). The purpose of the interview is to gather key information on key domains and then use that information to evaluate the offender’s responses in comparison to the official record (e.g., arrest records, crime report, treatment history etc.). That is, risk assessment with dynamic factors or the latest generation of risk assessment accompanied by specialized tools (such as drug screeners, mental health screeners, etc.) requires a clinical interview to obtain and assess information from the subject. A good assessment process requires interviewing the offender, which allows the criminal justice professional to gather, collect, and evaluate the offender’s responses along with other information obtained in official records. And, as promulgated by Taxman and colleagues (2004) and Taxman (2004), an important part of the interview process is engaging the offender in processing his/her own responses to the interview questions as part of a process of engaging the offender in becoming more accountable for his/her behavior.
Lowenkamp, Latessa, and Holsinger (2006) found that many offenders were not screened for actuarial risk before being placed in community correctional programs in Ohio, and that reductions in recidivism were noticeable for high-risk offenders in correctional programs that tended to be multi-dimensional and primarily served high-risk offenders. The authors developed an actuarial risk tool that focused only on the offender’s static risk factors (prior arrests, prior incarceration, age at current arrest, employed at arrest, history of failure in community correctional programs, drug use history). In a series of articles and presentations, they have reported the same results for offenders placed in residential programs, intensive supervision programs, and other correctional programs in Ohio. Using a quasi-experimental design, the researchers illustrate that reductions in recidivism are possible by using standardized risk tools, which help to ensure that high-risk offenders receive the more structured services. Their research also illustrates how poor classification schemes can result in over-classifying offenders (i.e., placing low-risk offenders in inappropriate programs) and only serve to increase the recidivism of this group of offenders. Their research basically supports Rule #1 of EBP regarding the importance of actuarial risk tools. This is the recent addition to a long-standing support for this concept from individual studies and also, more importantly, from recent meta-analyses (see meta-analyses such as Andrews, Bonta, and Hoge, 1990, Andrews et al., 1990; Lipsey & Wilson, 1998; Gottfredson, Najaka, & Wilson, 2001; Wilson, Lipsey, & Derzon, 2003).
Research studies of late have shown that the field is struggling with how best to use the concepts of risk and needs in criminal justice decisions, and particularly on how best to integrate dynamic or need factors. A series of articles in the 2006 Crime and Delinquency (edited by me and Doug Marlowe) illustrate how this struggle occurs. Taxman and Thanner (2006) detail how a randomized trial to examine the efficacy of a seamless probation-treatment protocol was affected by the classification of offenders as drug-involved. Offenders were assessed using an actuarial risk tool in one stage of the experiment and then a clinical assessment was conducted to determine drug use. Using the standard DSM-IV criteria (Diagnostic Statistical Manual IV-TR), a clinician assessed the offenders to be drug abusers. (DSM-IV states the accepted criteria for abuse and dependency.) In this experiment, half of the offenders were classified as high risk and half as moderate risk. However, few of the offenders in either the high risk or the moderate risk categories could be classified as drug dependent by the DSM-IV criteria. (The intervention involved a cognitive behavior treatment that was geared for offenders with drug problems.) Study findings indicate that the seamless system had no impact overall, but analysis found that the seamless system had a positive effect on high-risk and drug-dependent (addicted or serious abuse problems) offenders. In this study, the clinician did not use a standardized tool to assess for a drug problem, which resulted in overclassifying offenders as drug users when in fact many would not have met that criteria if a standardized tool was used. Another article in this edition by DeMatteo, Festinger, and Marlowe (2006) found that in many drug courts numerous offenders are not drug dependent and had generally low-threshold drug use (they were nevertheless classified as drug offenders largely due to their involvement in the legal system, which is one of the criteria for being classified as an abuser). Yet, these offenders are asked to participate in a highly structured program and required to go to drug treatment services. Not surprisingly, the drug courts do not tend to demonstrate reductions in recidivism.Clarifying the Concepts of Risk & Dynamic
Risk and needs vs. static and dynamic? The third and fourth generation tools combine two concepts into one instrument or protocol. The two concepts are: 1) that actuarial risk factors can be used to determine the degree to which the offender’s history predicts that he/she is likely to be a risk in the community or in a prison setting (i.e., the past predicts the future notion); and 2) that needs or psycho-social factors that should be ameliorated or addressed can be identified to reduce the risk for further involvement in the criminal justice system. The combination of these two concepts into one instrument or a cascading model (using screeners to determine the need for more in-depth inquiry into a problem area based on the results from the screener, such as fourth generation instruments include) evolved from the needs of the criminal justice system for better classification and treatment placement tools. The researchers constructed the Wisconsin Risk and Needs Tool to allocate service resources accordingly (much like a triage approach, where high-risk offenders would receive the scarce resources first to prevent harm). This resource allocation tool was constructed on a management-model premise.
The field has had a difficult time learning to use these tools in a manner that would facilitate the intended purpose—Rule #1. Again, the intended purpose is to isolate the criminal drivers while keeping in mind the actuarial risk. Criminal drivers refer to the people, places, or things that affect an individual’s involvement in criminal behavior. This means that the current status of an individual in areas that may in the past have been a problem may not be as relevant as other areas. Since many of these instruments use dichotomous (yes/no) responses or three categories of responses (none, some, many), practitioners are often left wondering how to select the drivers from other precursors. And, since many behaviors are intertwined, such as co-occurring disorders of mental health and substance abuse problems, the practitioner needs to determine which factors should be addressed as part of the criminal justice system, and which factors may be important for the person to address in the greater scheme of his or her overall health and well-being, but do not necessarily need to be encompassed in the criminal justice system.
Many attitudes, values, and behaviors lie on a continuum of “no problem” to “severe problem” behavior. This is important to keep in mind, because most human beings exhibit certain negative traits, but it is the degree to which these traits influence the subject’s involvement in negative behavior such as crime or drug use that concerns the criminal justice system. In this case, the negative influence is one that predisposes an individual to engage in certain acts, or hold certain values or attitudes that they tend to hold when engaging in behaviors that are covered by the criminal laws.
Risk Factors: Actuarial
In determining actuarial risk, we measure behaviors that predict negative outcomes (increased risk for criminal behavior) (e.g. the concept of predictive validity). The actuarial risk generally refers to demographic or historical factors (past behaviors) that affect the trajectory of an individual. For example, age of first arrest (or incarceration) is a predictor of further involvement with the criminal justice system, since the earlier an individual has been involved in the criminal justice (or juvenile justice) system, the greater the likelihood of future involvement. The actuarial concept in criminal justice is similar to that used in assessing risk factors for health insurance (e.g. family history, age of onset of a disorder, number of occurrences, etc.) or car insurance (e.g. prior driving history, speeding violations, etc.). As discussed by Gottfredson and Moriarty (2006), the statistical methods and methodology for developing these tools are sound. The emphasis is placed on criminal behavior, and the historical factors that predict the likelihood that an offender will continue criminal conduct.
The key question is the criterion variable or the behavior that is being predicted. In traditional criminal justice literature, the criterion variable is new criminal behavior (as measured by new arrests or reincarceration). Yet, many proxies that may be used in a risk assessment may not be direct measures of criminal behavior. Examples are substance use (except the tautology that use of illicit substances is a criminal act) or other victimless crimes (e.g. prostitution, etc.), technical violation for failures on probation and/or parole supervision, and so on. Clarifying this concept is important to differentiate whether the behavior being predicted is actual further criminal behavior. It should be noted that heightened law enforcement activities (arrests) in some geographic areas (which increases the odds for arrest) may influence certain variables. This is why some researchers are focused on certain classes of behavior (e.g. property crimes, violent crimes, etc.) that are less susceptible to the neighborhood context that an individual resides in.
Dynamic Factors: Criminal Drivers
The third and fourth generation assessment tools include questions about dynamic factors, or psycho-social needs that, if unaddressed, tend to increase the risk that the individual will commit criminal acts. That is, while many of these factors may be present in most human beings, it is the degree to which they influence an individual’s daily functionality that determines the degree to which they affect the offender’s behavior (criterion validity). The important component is that these need factors also predict the likelihood that an individual will become involved in criminal behavior due to the impact on the offender’s current behavior. Researchers have found that certain domains are more likely to negatively impact an individual, whereas other domains that we might think, using common-sense, have the same impact (e.g., mental health status, low educational status, or underemployment) are not directly related to criminal conduct.
Substance Abuse
Frequently the statement is made that over 70 percent of offenders are drug involved. This statement derives from reports regarding how many in the offender population report some use of illicit substances during their lifetime (or the lifetime prevalence). This statement many not refer to current use or use that is associated with dysfunctional behavioral. Using a clearer definition, researchers have generally found that about 35 to 50 percent of offenders have substance abuse patterns that require drug treatment (about one third of males, about half of the females) (Belenko & Peugh, 2005; Taylor, Fitzgerald, Hunt, Reardon, & Brownstein, 2001). The drug-crime nexus literature is a complex web that does not illustrate any causality between drug use and other criminal behavior, except for heroin or crack addicts, where the literature is clearer cut. The alcohol-crime nexus also is convoluted (besides the tautology of alcohol consumption in public, etc.), and just like the drug-crime literature, the relationship between substance use and crime depends upon the nature of the use and situation.
Table 1 illustrates the criteria for abuse and dependency accepted in the field (APA, 2004). The DSM-IV criteria differentiate between use and abuse, both of which are defined by the degree to which the use (abuse) affects the person’s daily functions. The literature on drugs and crime is most clear cut about the impact of providing treatment services for drug-dependent heroin and crack addicts—providing treatment will reduce recidivism and substance abuse. Based on this literature, it is suggested that it is important to identify drug-dependent addicts and then place them in appropriate treatment services. The priority should be given to targeting high-need (i.e. dependent) substance abusers for appropriate services. It should also be noted that those involved in the career business (i.e. entrepreneurs or those that are involved in dealing, etc.) may be classified as abusers when in fact their criminal behavior is linked to the business, and not to the drug use. While many involved in the business of drug dealing are also “dabblers” or users of small quantities of substances, their overall use is generally not due to compulsive behavior but rather to opportunity structures.
Family Dysfunctional
Family disarray and histories are generally precursors to learned behaviors—some are negative such as drug use or criminal behavior. Within this context, people learn attitudes, values, and behaviors. Differences exist in how families affect the behavior of men and women based on the degree of dysfunction in the family. For men, the stress from the family is to be a contributor (financial and otherwise) or to play a major role in the family. For women, the stress from the family is to be a caretaker or to be subservient to males in their lives. To obtain the support that is needed from the family, the offender is susceptible to responding to the pressure through criminal behavior (or drug use). The issues regarding the family are complex, in that the household may allow and tolerate certain behaviors in the home, including substance use or criminal behavior. And, the family could have expectations that the offender feels unable to meet.
Peer Associates
The other (and sometimes more influential) support mechanism that many rely upon (non-familial) generally consists of peers or associates. The risk factor is that the offender associates with others in a like situation, and this reinforces the criminal behavior. Over time, the offender essentially loses contacts with prosocial or non-criminal individuals. In other words, the offender fails to maintain the social support network that supports mainstream behaviors (prosocial). This is not just an issue of whether the offender is involved in a gang but rather whether the offender has any close associates that are not connected to criminal behavior. The question here is the degree to which the offender relies upon the peers that are involved in the criminal justice system and whether any of the associates are non-criminally involved.
Criminal Personality
Using the DSM-IV criteria, antisocial personality disorder (ASPD) and impulsive behaviors are part of the composite of personality disorders. According to the DSM-IV, approximately 3 percent of men and 1 percent of women have some form of antisocial personality. As shown in Table 2, the antisocial personality disorder is characterized by a callous unconcern for the feelings of others, gross or persistent attitude of irresponsibility and disregard for social norms, rules, or obligations, incapacity to maintain enduring relationships, low tolerance for frustration and low tolerance for use of aggression or violence, incapacity to experience guilt or to profit from experience, or marked proneness to blame others for the behavior that the offender exhibits. This personality disorder differs from psychopathy, which is a more callous version of an ASPD, and some states have developed legal or judicial definitions of what constitutes a psychopath. In terms of the medical definition (according to the DSM-IV), a psychopath is defined as having no concern for the feelings of others and a complete disregard for social obligations. The psychopath is generally considered callous and incapable of forming lasting relationships; the psychopath lacks empathy, remorse, anxiety, or guilt and tends to be devoid of conscience. Psychopaths are the extreme criminal personality. A proper diagnosis requires clinical skills as well as standardized tools (see the Hare’s Psychopath Checklist, Hare, 1990).
Antisocial Cognitions (Attitudes/Orientation/Thinking)
As distinct from the personality disorder is the attitudes and cognitions of offenders. Yochelson and Samenow (1976) in their seminal work identified 36 thinking errors that they believed are used to shun responsibility, at least as defined by society’s standards. The continuum of criminality is from responsible to irresponsible; under “irresponsible” behavior, there is a range from nonarrestable to arrestable behavior. The scholars contend that all individuals have these kinds of thinking errors; however, criminals exhibit more of them, and they tend to be on the irresponsible end of the continuum in many of the areas. The thinking error phenomenon gained further steam with the work by Walters (1990) and his colleagues, in which they developed eight subscales to measure criminal thinking; Walters’ work influenced a new tool on criminal thinking errors (Knight, Garner, Simpson, Morey, & Flynn, 2006). Most of this work has been criticized because the tools to measure criminal thinking have not been validated on a non-offender population, and therefore it is unclear whether these characteristics are concentrated in offenders or distributed among the general population. Also many of the “thinking errors” are common defense mechanisms that are used by human beings to handle situations.
The typical thinking errors include dominance, entitlement, self-justification, displacing blame, optimistic perceptions of realities, and “victim stance” (e.g. blaming society because they are considered outcasts). As noted by Mark Lipsey and Nana Landenberger (2006), such “distorted thinking may misperceive benign situations as threats (e.g., be predisposed to perceive harmless remarks as disrespectful or deliberately provocative), demand instant gratification, and confuse wants with needs” (p. 57). The issue about attitudes and orientation is that the focus is on how the offender processes and interprets information.
Low Self Control
Impulsive and risk-taking behavior is another dynamic characteristic of offenders. The general premise is that low self control does not define criminal behavior; instead, it provides a context for criminal acts depending upon opportunities and other motivating factors. A person’s decision to engage in criminal acts is affected by other factors such as natural constraints, attachments to parents, school, employment, and so on (Gottfredson and Hirschi 1990, p. 95-97). Low self control is exhibited by the offender being easily persuaded by situational and environmental factors, and without attachments there is little to constrain the individual.
Mental Health, Self-Esteem, Low Educational Attainment, Employment & Other Factors
Mental health status, self-esteem, low educational attainment, low employment options, and other factors are frequently discussed in the realm of criminogenic needs. The definition of a criminogenic need is that the factor predicts criminal behavior, and the research literature does not demonstrate that the presence of these attributes predicts recidivism or involvement in criminal behavior. Rather, low educational attainment and unemployment appear to be correlated with the offender population, which leads some to conclude that addressing these factors may also reduce recidivism. As discussed previously with other substance abuse and ASPD, the behaviors range on a continuum. The same is true with mental health disorder, where the problems range from anxieties or depression to erratic and/or risky behavior (e.g. hears voices or expresses disorganized, disoriented, or paranoid thoughts; appears lethargic and sad; unusually manic in behavior, etc.). However, it is generally recognized that in order to improve an offender’s wellbeing (which may not be related to recidivism reduction efforts), he or she would benefit from improved employment, educational, and mental health status(es). Ultimately, addressing these issues may affect the ability of the individual to be a contributing member of society and/or family; it is unclear whether addressing these factors in and of themselves will affect criminal behavior.
Applying Rule #1 in Correctional Agencies
Exhibit 1 illustrates the implementation of these principles into a model. Essentially, actuarial risk level should be determined to identify what is the offender’s likelihood of further criminal behavior. High-risk offenders should be targeted for treatment-based on the area (s) in which they score moderate or high on criminogenic needs. That is, the offender needs to be assessed also on the criminogenic needs to identify the drivers to their criminal behavior. The notion is that, similar to treatment placement models, actuarial risk should drive the priority for intensive control and appropriate services, with a focus on selecting programs that address multiple problem areas. “Appropriate” refers to attention to the criminogenic factors that have been identified.
The model presented in the exhibit illustrates how the criminogenic factors can exist regardless of risk level. That is, a substance abuser may be low risk due to the fact that he or she does not have a history in the criminal justice system. Other criminogenic factors may exist in that low-risk person, but they are more likely to be low to moderate in severity. As the offender moves along the continuum of risk (moderate to high), then it is more likely that more severe problem behaviors may occur. This is a byproduct of the offender’s inability to be a productive, contributing member of society. For example, a high-risk offender may have criminogenic needs relating to self control, peer associates, ASPD, and substance abuse. The combined treatment and control strategies should be designed to address these issues. The model also suggests that the high-risk offender is more involved in situations, settings, and individuals that are likely to further their criminal conduct. Hence, control and treatment services should be concentrated on this individual to achieve the desired goal of reducing the risk of recidivism.
Conclusions
The purpose of this article is to further elaborate on Rule #1 in Evidence-based practices to better illustrate the concepts and to define criminogenic needs in the context of risk level. This article is driven by the needs of the field to translate the principle into operational terms. An actuarial-based risk screen is important to determine the degree to which offenders should be given services and resources to ameliorate criminal behavior. The type of services is determined by how the offender “scores” or presents on several criminogenic areas. Those offenders with high criminogenic needs, particularly those that are high or moderate risk, should be given services to ameliorate the criminogenic need, which should reduce the risk for recidivism. Exhibit 1 conceptually presents the framework underscoring EBP #1; the challenge to organizations is to implement this principle.
The field faces several challenges relating to organizational stamina in implementing Rule #1 by following this core concept. The first challenge is the willingness of the organization to focus services on high-risk offenders, which generally means that moderate-or low-risk offenders should not be given such services. Minimizing the provision of services for low-risk offenders essentially results in decisions that the probation supervision should minimize the disruption from prosocial behaviors, since they are likely the glue that is preventing the offender from becoming criminally involved. Another factor is that the case plan/supervision plan should be driven by the goal to ameliorate the criminogenic drivers. This is critical, since it provides the formula for reducing the risk of offenders in the community. Exhibit 1 illustrates that when an individual is identified as having moderate or high criminogenic needs, then the plan should be to address these criminal drivers in the case plan. That is, the results of the assessment are directly relevant to the components of the case plan, because it provides an avenue to assist the offender in attending to issues that are relevant to his or her life. In short, EBP #1 challenges the organizations to redo case plans so that they address the drivers (criminogenic needs) that are more pertinent to the situational factors of the offenders. In so doing, case plans become the glue for the offender that addresses the risk factors.