Some notes on Judging Risk

I just read Brandon Garrett and John Monahan’s article called Judging Risk. It’s a useful piece that considers the use of a particular risk assessment tool in sentencing in Virginia, along with a survey of judges they had published previously. They generally find uneven use of the tool and suggest a few reforms to the risk assessment process.

Most generally, I was surprised to learn that, in the US, risk assessment was quite common until the mid 1970s, when courts became more retributive in their sentencing. I wonder if this was a worldwide trend (or at least evident in similar jurisdictions). There has since been a rebirth.

Moving on to the meat of the paper, Garrett and Monahan focus on the Virginia “Non-Violent Risk Assessment” (NVRA) Instrument. This tool is significant because Virginia was the first state to formally include risk assessment in sentencing.

I was a bit surprised at how few factors the NVRA considers:

The instruments for each offense examine only the following static factors: (1) offender age at the time of the offense; (2) gender; (3) prior adult felony convictions; (4) prior adult incarcerations; (5) whether the person was legally restrained at the time of the offense; and for drug offenses, additionally (1) prior juvenile adjudication, and (2) prior arrest or confinement within past twelve months rather than legal restraint at the time of offense. If the offender’s total score on the instrument is below the cut-off, the offender is recommended for an alternative sanction. If the offender’s score on the instrument is above the cut-off, the prison or jail term recommended by the sentencing guidelines remains in effect.

Garrett and Monahan went through the data shared by the Virginia Criminal Sentencing Commission (VCSC) and found that when the NVRA classified the offender as low risk, judges still refrained from imposing alternative sentences 57.8% of the time.

In their survey of judges, the authors found that while most judges were familiar with the NVRA, only about half always or almost always considered its results. They also found that judges felt constrained by the availability of alternative sentences.

Judges were also against a requirement that they provide reasons if they diverge from the NVRA because they thought it would increase their workload.

Garrett and Monahan go on to provide a sensible roadmap for regulating risk assessment. I particularly agree with their suggestion that any tool be developed transparently and thus be open to scrutiny by all stakeholders.

I suppose my main criticism of this work was that it I am concerned the materials may not be available in the long term because they are hosted on the Commission’s website. Since they are calling for transparency reforms, it would be good if the article linked directly to a database with more permanence (like the OSF). Further, neither the data from this survey nor that from the VCSC review seem to be public.