The Risk Principle for Dummies Evan Crist, Psy.D. evan@correcttech.com 303-586-6551
What is criminogenic risk? Criminogenic literally means crime creating (and, no, you will not find the word in the dictionary). Risk refers of the probability of an adverse event. Criminogenic risk is a statement of probability that a specific offender will reoffend if there is no intervention to lower risk. The second half of that sentence is often ignored but is vital to the true meaning of the principle. Appropriate intervention can lower risk, in some cases significantly so. There are many misconceptions about risk. Risk does not indicate the likelihood of any particular type of crime. A sex offense and theft of pair of sunglasses are weighted equally in terms of risk. The severity of the crime is also irrelevant. There are a variety of objective risk assessment instruments, the most common being the Level of Supervision Inventory-Revised (LSI-R). Originally developed by Don Andrews and James Bonta and revised in 2003, the LSI-R is used worldwide as an objective measure of risk. Why does it need to be objective? While the exact objective measure used is not that important, using an objective measure is very important. Objective means numbers are assigned to specific criteria and the total score is the risk score. Prior to objective measures of risk, subjective assessment was used. Subjective assessment of risk was based on review of their history and a verbal interview with the individual about their plans and attitudes. In other words, subjective assessment is a wild guess, probably more reflective of the evaluator s biases and the likeability, or effective manipulation, of the offender. Objective assessment accomplishes two things: 1) it ensures that criteria that actually leads to crime is measured and 2) it takes the evaluator s biases and tendencies out of the equation. While different states and provinces (remember Canadian researcher actually invented the risk principle) use different thresholds for the risk categories, the State of Colorado adheres to the following cutoffs: Risk Category LSI-R Score Range 0-18 Low 19-28 Medium 29-54 High
Wanna buy a life insurance policy? If you ve ever applied for a life insurance policy, you ve participated in a similar risk assessment. When you apply for a life insurance policy, the insurance company attempts to determine how likely you are to die before the age of 70. In order to predict that risk, they look at a variety of factors that make you higher risk for a premature death. Some of these factors are completely out of your control. Male gender, older age, family history, and place of birth each contribute to risk and these factors cannot be changed. Risk factors that cannot be changed are known as static risk factors. No matter what you do, you can t change your place of birth. Fortunately, some factors that contribute to risk for early death can be changed. High risk behaviors (behaviors that are more likely to lead to a early death and thus a large life insurance pay out) include sky diving, motorcycle riding, smoking and heavy alcohol use. If you stop smoking, your risk of early death will decrease (and therefore your life insurance premium would go down). Risk factors that can be changed are known as dynamic risk factors. More about the importance of this concept later. So what? Great. I have a risk score. What does that do for me? There are a variety of reasons why the concept of risk is important. It clearly has important financial and public safety concerns. Incarceration, and correctional interventions in general, are expensive in terms of financial and human capital. The research is clear that lower risk offenders are not particularly likely to reoffend, even if no intervention is provided. LSI Total Score % chance of reoffending within one yr 0-5 9% 6-10 20% 11-15 25% 16-20 30% 21-25 40% 26-30 43% 31-35 50% 36-40 53% 41-45 58% 46-50 69% 50-54 >70% Source: Andrews, D.A. and Bonta, J.L. (2003). Level of Supervision Inventory-Revised. U. S. Norms Manual Supplement. Multi Health Systems. Toronto.
Therefore, when limited funds are available (i.e., always), it makes more sense to target those funds for interventions for high risk offenders rather than lower risk offenders. In other words, spend the money on the client who is most likely to harm society without the intervention. Another reason to consider risk level is that research indicates that the risk level should dictate the intensity of treatment provided. What is treatment intensity?: High risk offenders need a strong intervention. That does not surprise anyone. However, some research indicates that if a lower risk offender receives the same intervention, they could get worse! Really? couldn t any offender benefit from treatment? The answer appears to be it depends. In short, low risk offenders require a low intensity of treatment and high risk offenders benefit most from a high intensity of treatment. If you over-treat or under-treat, the offender tends to get worse (i.e., are more likely to reoffend than before treatment). Simple enough, until you ask the next obvious question. How do you define and measure treatment intensity? What does high intensity mean?: While the principle of matching risk level to treatment intensity level is almost universally accepted, an adequate definition of intensity is rarely given. The best definition that we ve seen is offered by Ed Latessa and Christopher Lowenkamp of University of Cincinnati. In their 2004 article, entitled, Residential Community Corrections and the Risk Principle: Lessons Learned in Ohio, they offered the badly needed and perfectly reasonable definition of intensity: high intensity treatment=residential placement and low intensity treatment=non residential treatment. In summary, the authors research demonstrated that when low risk offenders are placed in residential facilities, their rate of reoffending is higher than for low risk offenders who do not receive residential treatment. While we are still left with questions like, What about medium risk level offenders? and Short of residential treatment, how do you measure intensity of treatment?, the University of Cincinnati researchers clearly demonstrated that over-treating low risk offenders often makes them worse.
Why would it make them worse? So why would over-treating make an offender worse? Isn t that counterintuitive? There are a few theories as to why this finding has been replicated so often: 1) Low risk offenders tend to learn high risk behaviors and attitudes from high risk offenders. Anecdotally, this makes sense. In practice, we ve observed that high risk offenders tend to be stronger leaders, positive or negative, in group therapy sessions and residential milieus. In other words, attempting to use a low risk offender as a good role model for high risk offenders is not only ineffective but is actually counterproductive for the role model. 2) When you place a low risk offender in residential, what impact that does have on his or her life? Well you take away their support system (that has, relatively speaking, been prosocial and apparently a good influence on the offender) and replaced that social influence with a group of high risk offenders. Similarly, placing an offender in residential means that person will lose their job. Now they have more time on their hands to watch, talk to and imitate they residential treatment cohorts, high risk offenders. 3) When you label (verbally and behaviorally) a person as someone who is abnormal enough that we must take them out of society (i.e., place them in residential), they tend to adopt that new identify. If their personal identity is not enough, when an offender is discharged from residential treatment they must return to their natural environment with the stigma associated with residential treatment, making the transition back to their community much more difficult. What factors contribute to risk?: Just like the insurance industry, the risk model for offenders has very specific factors that are important. These models were developed in the exact same way that life insurance risk levels are determined. The life insurance industry looks at factors that correlate with early death. The criminal risk model looks at factors that correlate with being arrested. Like in the life insurance industry, some of these factors can be changed ( dynamic risk factors ) and other are static, cannot be changed. No doubt, you ve heard the phrase, Past behavior is the best predictor of future behavior. That is largely true. While people can, and do, change their lifestyles and behaviors, once people set off down a particular path, barring some dramatic road block or detour, they stay on that path. That concept is incorporated into criminal risk too.
A history of criminal behavior is one factor that is measured for criminal risk. Since you cannot change history, you are stuck with it. It is a static risk factor. Some risk assessment instruments only look at static risk factors. While they have predictive value, they make no room for an offender s ability to change his risk level by changing his or her lifestyle. In other words, there is nothing an offender can do to rid themselves of a high risk label. Since, in our view, community corrections is all about people changing their lives, we need a risk assessment instrument that measures those positive changes. The LSI-R considers history, a static risk factor, but the majority of the risk factors measured are dynamic risk factors. These dynamic risk factors are known as criminogenic needs. These needs are factors that increase or decrease risk AND can be changed. We will discuss myths and realities of criminogenic needs in our next issue.