Originally posted at CA Fwd.
By Tom Hoffman.

Research has shown that two basic assumptions must form the philosophical backbone of implementing the monumental criminal justice overhaul that is AB 109. First, that supervision must be rooted in evidence-based programming and intervention (a view widely held by academics). Second, that offender programming must be crafted through the use of a validated “risk and needs” instrument.  These instruments most commonly identify risk and needs in three categories of criminal behavior: drugs, property and violence.

At a very basic level, the discussion around crime, punishment (or “accountability”) and incarceration can be thought of more broadly as how we respond to criminals who we are “mad at” (drugs and property) versus those we are “afraid of” (violent).  Assessment instruments have repeatedly been shown to help us more accurately and quickly identify these populations, and therefore enhance public safety.

As discussed in a previous post, the current AB 109 population was not identified via the use of a validated “risk and needs” assessment and there is a well-documented shortage of evidence-based programs in virtually every community in our state. In the months and years ahead, the success of AB 109 will be largely dependent on altering this reality in communities throughout California. This will require education, training and policy shifts across many stages of criminal proceedings in most of California’s counties.

In 2006, the California Department of Corrections and Rehabilitation (CDCR), in an effort coordinated by the Division of Adult Parole Operations (DAPO), implemented the system-wide use of the Correctional Offender Management Profiling for Alternative Sanctions (COMPAS), developed by Northpoint, Inc. The COMPAS instrument is one of a dozen or so “off the shelf” risk and needs assessment instruments currently available.

COMPAS has been in use for many years and is validated in many states across the nation.  Today this important instrument is used by prison and parole staff in the development of the supervision and programming strategy for the vast majority of inmates and parolees. A copy of a blank COMPAS form can be found here for reference.

So how does a tool like COMPAS work and what is its value?

All criminal populations, or the environment in which individuals are discharged, are not the same with regard to housing, programming, community support,  and education. As a result, risk and needs instruments must be “validated” for each specific population. A “validated” instrument simply means that there has been a post-release analysis (usually three years) conducted on a population whose risk and needs score was determined by a specific instrument.

In other words, the research verifies whether, over this time period, offenders’ post-release behavior was consistent with what the instrument predicted. This includes new violations for drug, property or violent crimes. This is an important undertaking when any public agency commits to using a predictive instrument at any point in the criminal justice process. Due to the emerging nature of this discussion in California, many organizations have not yet validated the instrument they have in place. Making this is a priority in the near future is an important stepping stone.

Regarding the risk and needs assessment process: we can all agree not all criminals who commit a same or similar offense are uniformly dangerous, i.e. the “offense” does not necessarily define how dangerous an offender is to the general public.  Some criminals are simply more dangerous than others and therefore the need exists to accurately and quickly identify which ones those are. Research has demonstrated that when supervising agents (Probation, Parole, Court Officers, etc.) attempt to predict the likelihood of future violent criminal behavior by an offender under their control, they are accurate about 50 percent of the time. When this same evaluation is done using a validated risk and needs assessment instrument, the accuracy climbs to 70 percent, and in some cases can approach 80 percent.

If you’re involved in implementing AB 109 and unfamiliar with these premises surrounding evidence based programming and methodical risk and needs assessment, there is decades of research and data available that may enlighten you as to their value.  In fact, because California was home to the world’s largest prison and parole population for most of the past two decades, a great deal of the published, generic research actually focuses on California.

It is my personal experience that once armed with the facts, the logic behind embedding these important concepts in a wide range of public safety policies and programming strategies is compelling. In an ideal world, these concepts would play a meaningful role in the decisions taken by public safety personnel across the criminal justice spectrum, from prosecution to release to supervision in the community.

The absence of a shared understanding and commitment to these instruments throughout the state’s criminal justice infrastructure (local police, District Attorney, the Courts via Probation Department pre-trial and sentencing reports, Sheriff Department’s in custody programming, etc.), can create conflict and ineffective coordination. This leads to less than effective outcomes for everyone, including the public and the offender.

One of the core obligations of the criminal justice system is to mitigate the risk violent criminals represent to the public at large, to the extent possible. Boosting accuracy when predicting future criminal of those who have already committed crimes through the use of risk and needs assessment instruments is entirely consistent with this thinking.

For those looking for information and research around risk and needs assessment, a quick web search will reveal a robust library of information literally at your fingertips. Well regarded research organizations such as the Stanford Criminal Justice Center, the Washington State Institute of Public Policy, the Pew Foundation, the California Expert Panel, the Vera Institute, the Center For Evidence Based Corrections at UC Irvine and others have published many investigative and research papers around this important discussion.