The LSI-R is a testing instrument traditionally used by correctional facilities to evaluate recidivism risks. And recidivism is a huge problem in Delaware. A recent study by the Criminal Justice Council revealed that 67% of Delaware’s 2009 prison releases were reconvicted and back behind bars within three years.

Recently, Delaware court and corrections leadership have been considering the use of this assessment at the time of plea bargaining and sentencing. The question is whether the use of LSI-R in that way would be fair or effective.

The Assessment

LSI-R stands for “Level of Service Inventory – Revised.” It is an interview based assessment that attempts to predict the risk of recidivism and the appropriate classification level for sentenced inmates based on a number of criteria. It also helps to determine the appropriate service level to provide to an inmate to assist in the reentry process. This training manual gives a good idea of the nature of the assessment.

The LSI-R assessment yields a score based on point values assigned to answers to the 54 questions grouped in 10 categories. According to a user’s manual, the categories are:

•       Criminal History/prior prison conduct

•       Education/Employment history

•       Financial situation

•       Family/Marital situation

•       Accomodations (living arrangements, etc.)

•       Leisure/Recreation Activity

•       Social Companions.

•       Drug and Alcohol Problems

•       Emotional/Personal issues.

•       Attitude.

As you can see, some of the topics are factual and some are a little squishy. Some of the topics are within the offender’s control, and some, like whether they grew up in a high crime neighborhood, can be outside the control of the offender.

The scoring of the LSI-R produces a report that actually defines the probability of recidivism and gives a suggestion for supervision level. Here is a real LSI-R scoring report we found online.

Tools like the LSI-R are fairly recent concepts in United States corrections. According to Federal Probation’s 2006 article, utilizing a tool to predict recidivism was “only recently” embraced by parole boards and a few states.  The article, moreover, expressed misgivings about implementation of the LSI-R.   It cautioned that there “seems to be a trend in corrections to uncritically accept the latest ‘innovation’ and adopt it without understanding its strengths and limitations.”

The LSI-R in Delaware

According to a recent Delaware Law Weekly article, the instrument has been in used “in some capacity” in Delaware since 2003.

However, Delaware now appears to be looking to expand use of the LSI-R beyond its traditional role as a tool in the re-entry or inmate management processes.   The article states that Delaware may utilize the LSI-R to evaluate the dangers of recidivism among pretrial defendants.   A pilot program is in the works so that the LSI-R could be used “on detentioners and determine if that instrument could be of value to the court, the Public Defender and the AG’s office during plea bargaining and sentencing.” Robert M. Coupe, Delaware Department of Correction Commissioner, stated that Delaware is still “trying to identify how to start a pilot program.”

According to the article, Superior Court President Judge James T. Vaughn, Jr. also revealed that judges are trying to determine how to implement the LSI-R at sentencing. This revelation about the LSI-R’s possible new use in Delaware (along with news of the pilot program) emerged from a meeting of court and correctional facility officials in late August.

Whether the LSI-R should be used in sentencing is an interesting question. On the one hand, it is a data point for judges to consider, and more information is better than less. On the other hand, because the LSI-R produces a “score,” we are concerned that there will be overreliance on it at the cost of considering less quantifiable factors like mitigation evidence. For example, it is helpful to know how a college applicant did on the SAT test but that one data nugget is far from the whole story about that person.

Just because someone got a worse score on the LSI-R, does that mean they cannot be rehabilitated and should get a lengthier sentence? Absolutely not. But we fear that would be the tendency.

Ultimately, the LSI-R could be another tool in the toolkit for sentencing judges, but is no substitute for the consideration of well-presented mitigating (and aggravating) evidence. If used properly, it could possibly be a helpful data point. Clearly the assessment is much more suited to correctional programming and risk assessment; use in sentencing seems like a bit of a square peg in a round hole.

We would certainly support a limited pilot program which would give our leadership a better understanding of the LSI-R in sentencing. After the pilot program, it can be reassessed, specifically to see if use of the LSI-R put a dent in our extremely high recidivism rate here in Delaware.