Does offender supervision reduce reoffending?

In this blog, ADR UK Research Fellow Dr Markus Gehrsitz describes how the 2015 Offender Rehabilitation Act created the opportunity for a unique natural experiment to explore whether offender supervision and license conditions reduce reoffending.

Reoffending is off the charts

Offender rehabilitation in England and Wales is in crisis. 57% of adults released from short custodial sentences reoffend within just one year of their release. This is a key driver of record-high prison overcrowding and underscores the need for effective rehabilitation strategies.

In this project I turn to administrative data to investigate whether offender supervision and license conditions can help reduce reoffending. License conditions are rules offenders must observe when they are released from prison. Among other things, offenders are supervised and must regularly check in with a probation officer. They also have to adhere to residency requirements and often need to take part in training and rehabilitation programmes.

Does this effectively reduce reoffending, or could resources be better spent? We don’t really know. I will take a hard look at the data to provide an answer.

Let’s conduct an experiment

How can we figure out whether license conditions and offender supervision reduce reoffending?

Imagine the following experiment. Over the course of a year, half of the thousands of offenders released from short prison sentences are randomly assigned license conditions and a probation officer. The other half is released unconditionally – no residency requirements, no probation officer to check in with, no strings attached. In this experimental world, all a researcher would have to do is compare reoffending rates of those who were randomly selected for license conditions with those who were released unconditionally. If the license condition group had lower rates of reoffending, we could confidently conclude that supervision works.

Does this stylised experiment sound too fanciful? Not feasible, not realistic? As it turns out, a recent piece of legislation, the 2015 Offender Rehabilitation Act, created almost exactly this kind of experimental setup. Prior to this Act, all offenders on sentences of less than 12 months were released unconditionally. After the Act went into effect, all offenders were suddenly subject to at least seven standard license conditions and had to be supervised by a probation officer. It turns out, this Act created a clean experiment for us!

Let the data speak

Even the cleanest of experiments is worthless without good data on experimental subjects. Fortunately, the Ministry of Justice has created the Data First: Cross-Justice dataset, which I can use to solve this problem. This unique data source allows researchers to track offender journeys through the criminal justice system.

Particularly valuable for my research question: the data cleanly identifies which offenders were released with Offender Rehabilitation Act license conditions, and which offenders were released around the same time unconditionally. Even better, in the data I can follow each former prisoner to see if and when they are either re-incarcerated or re-convicted by a court.

In other words, I am able to accurately calculate the reoffending rates of both groups of offenders. Once this project is completed, we will know whether offender supervision and license conditions are in fact effective in reducing reoffending.

Impact: Something that works?

My hope is that policymakers and criminal justice practitioners will watch this research closely. It is not every day that we can use a clean natural experiment to reliably assess an important policy question. We will learn whether offender supervision actually makes a dent into sky-high reoffending rates. This equips policymakers with actionable policy prescriptions.

In the criminal justice space, we often simply do not know what works and what does not – my hope is that this research will finally shed some light on the effectiveness of a policy that touches on hundreds of thousands of lives.

Find out more about Markus’ project.