Earlier this month several Clinks staff attended a talk on desistance and commissioning by Fergus McNeill, Professor of Criminology and Social Work at the University of Glasgow.  We felt his ideas had great significance for how VCS organisations carry out and monitor their work.  If it interests you, you can hear Fergus for yourself at our conference in January.  We’d also love to hear your feedback on some of the issues raised here, so this post includes a few discussion points at the end.

Fergus’s summary of desistance theory argues that there is more to the cessation of crime than is commonly allowed by measurements such as a simple two-year reoffending rate.  The graph below represents the career of a (hypothetical) persistent and prolific offender.  The X axis represents his age, the Y axis represents the frequency of reoffending.  Fergus pointed out that there was a third dimension (not easily represented on a graph, but still important), which was the severity of offences committed.

The colours are best explained in Fergus’s own words (from his speech to the 2010 NOMS conference):

He starts at the age of eight; his offending escalates rapidly in adolescence; plateaus in the early 20s and drops in the mid-20s until he desists at the age of 30.  All of the multi-coloured areas under the curve are social cost, cost to victims, financial costs, costs in the system.  If he gets either an effective prison sentence or an effective community sentence at the age of 18 and stops immediately if some intervention works at that point, then all that’s saved for the public purse and to crime victims, potential crime victims, is represented in both the amber and green areas under the curve.  That’s the dream in offender management – instant and complete cessation of offending.

Maybe more realistically, the amber area might reflect a deceleration in offending that’s supported through processes of offender management.  This person still shows up on your re-offending stats for another six years, but they’re changing and the change is the green area.  Even where all that you’ve done is decelerate offending, take five years off a criminal career, reduce the frequency, that’s still a big saving and still an important contribution to the public good.  And there’s a third dimension here, which I’m not clever enough to represent on a Powerpoint slide, which is severity.  The missing axis here is severity.  If you move somebody down the scale of gravity of offending, you’re also contributing significantly to the public good.

A second diagram, taken from the article Steps Towards Desistance by Anthony Bottoms and Joanna Shapland explains how that might happen:

This illustrates the complexity of the desistance process, its contingency and bidirectional nature.  Fergus used the example of someone who diets; they may become more determined as the result of a good weigh-in, or relapse and go on a binge after a bad one. They may then be helped back ‘onto the wagon’ by the love or support of others, or by a health scare, or by any number of other random events.  VCS organisations will be familiar with the levels of uncertainty and fear among ‘ex-offenders’, and the levels of support that are required on this erratic journey, as well as with the rewards that lie at the end of the process. But the crucial point is that only the offender can be in full control of this process. They are the most efficient manager of their own risk (which is not to say that all of them are immediately ready to exercise that control). Without their cooperation, the justice system can only hope (but not guarantee) to control them, using systems that may be more or less efficient.

What does all this mean for how Government might design and commission services in its ‘Rehabilitation Revolution’?

Both diagrams flag up the highly problematic nature of ‘binary’ measures of reoffending.  In England and Wales, this is currently measured  by the percentage reconviction rate within two years. Payment by Results (PbR) is intended to pay contractors based on the ‘justice savings’ model. Payments are made based on the idea that there will be long-term savings from a lower reoffending rate – the area under Fergus’s graph. A two-year binary model may count a lull in offending (or undetected offending) as desistance.  It may thereby reward savings that will not be made..  Perhaps more worryingly for the VCS, it may also fail to reward what Fergus calls ‘secondary desistance’, withholding payments from organisations whose work has helped to shorten the length of a criminal career or reduced the intensity or severity of offending. Even if there has been a new offence, PbR based on binary reoffending rates struggles to accommodate savings that might be made through reductions in the intensity, frequency, or gravity of criminal actions.

One of the slides in Fergus’s presentation provides  a systems perspective on NOMS. In this model, inputs (e.g. offenders, staff, the estate, and other resources) are transformed by systems and processes (e.g. OASys, accredited programmes, the Nine Pathways, PbR) into outputs (e.g. prison security standards, compliance with programmes, services commissioned). The hope is that these result in outcomes (e.g. reduced reoffending, government savings, and public confidence).

The trouble is that the system has no guaranteed control over at least some of its inputs, most notably offenders themselves.  If you have people who have justice 'done to' rather than 'done with' them, you won't reliably stop them reoffending, because you're not engaging the best source of risk management - themselves.

As Fergus writes in his own blog post:

"If we take a desistance-based perspective seriously it means we have to recognise that neither NOMS [i.e. the National Offender Management Service] nor any provider of services can command, control or compel reductions in reoffending, for the simple reasons that (a) desistance belongs to the people involved and (b) integration of ex-offenders belongs to communities (though the state, civil society and public services have duties here too). In other words, neither the commissioners not the deliverers of services in fact ‘own’ desistance and social integration; these things are not the outcomes of their work — they are human and social processes that depend on people and communities finding ways to resolve their conflicts, tackle their issues, realise their potential, accept and support their fellow citizens."

Furthermore, Fergus asserts that ‘reducing reoffending’ is not a positive agenda around which to build a service; it’s an absence of something, is too technocratic, and  has little emotional or    visceral appeal.

If that is true, what might a model of commissioning look like in practice which takes desistance into account?

Fergus suggested that in thinking about objectives, especially ones to be specified by PbR contracts, it is important to promise what we can deliver, and be able to deliver what we promise. The very best programmes may have striking impacts on reoffending, but organisations that present this as something that they can be sure to produce are offering themselves as hostages to fortune. Reduced reoffending cannot be promised. It may be better thought of as a positive (and even a likely, but NOT a guaranteed) side-effect of a quality service.

It is this quality of service, as defined by proportionality in sentencing, compliance with human rights, and constructive reparation of harm, that can be promised, and it is on this that justice services should be judged. Work to support desistance should happen alongside the service commissioned, but that service should not be judged by reduced reoffending, because to do so is to promise too much and therefore deliver too little. Fergus summed up desistance-based commissioning as something that engages seriously with offenders and communities, and allows very local commissioning of services. The national role would be to agree deliverable frameworks to define service quality and acceptable outcomes. Fergus described this as a Copernican model of commissioning, placing the person changing at the centre of a network of holistic, tailored services, which have the dual function of working with individuals whilst also mediating their reintegration into the surrounding  community.  

If you’ve read this far, perhaps you need to go away and let Fergus’s ideas sink in. His full PowerPoint presentation is attached to this blog post, and you can hear him for yourself at our conference in January. We would also be very interested to hear your responses, so please consider leaving some feedback in the comments box below, especially around the following questions, which strike us as particularly important or relevant:

• Many organisations will be thinking about PbR contracts. What do you think the implications of this model are for how you will collect and monitor evidence for outcomes?
• How can the implications of a desistance-based approach be ‘sold’ to local and national decision-makers (and to the public)?
• Do you agree with this model? Do you accept the conclusions that Fergus draws in terms of commissioning and the VCS approach to working with prisoners?

Blog post by:

• Ben Jarman
• Katie O’Donoghue