The Howard League for Penal Reform has been at the forefront of arguing that pressure on public spending presents an opportunity to re-frame the criminal justice debate in favour of reform. In Canada during the 1990s, for example, cuts to public spending saw the prison population reduced by 11%. During that decade, crime also fell to its lowest rate for 25 years, including drops ranging from 23% for assault and robbery to 43% for homicide. We are pleased that Ken Clarke at the Ministry of Justice is taking seriously our contention that it is perfectly possible to achieve less crime, safer communities and fewer people in prison.
We do, however, have serious reservations on the notion of ‘payment by results’, currently championed by government across numerous areas of public policy. Some of the practical difficulties were identified recently for Public Finance by the Social Market Foundation’s Ian Mulheirn. Without his understandable evangelism for the concept, however, we take an independent eye and find further areas of concern.
Firstly, as a new concept to criminal justice, payment by results has no track record of success in the sector. We must look elsewhere to see how it might work. Perhaps the closest parallel to introducing payment by results to community sentencing, with a similar range of private and voluntary sector providers, would be the ‘pathways to work’ programme in welfare. It can only cause consternation, then, to see the Public Accounts Committee verdict that the providers involved in the programme had “seriously underperformed against their contracts and their success rates [were] worse than Jobcentre Plus even though private contractors work in easier areas with fewer incapacity claimants and higher demand for labour”. These findings are not encouraging.
Secondly, for all the labelling in the terminology of ‘offender management’, we cannot – indeed should not – forget that people going through the criminal justice system are just that: people. People are complicated. Those going through the justice system, often presenting multiple and complex needs, are particularly complicated. The routes to desistance from offending may defy the simple pathways that a payment by results model would seek to impose.
For example, an individual may serve a community sentence, potentially involving a number of providers on a payment by results scheme overseen by the Ministry of Justice. That individual goes on to desist from reoffending because they have found employment. The same individual, however, has been helped into employment because they are also participating in a payment by results scheme run by the Department for Work and Pensions. Or it may be that the individual has desisted from offending because an underlying health or mental health problem has been alleviated, to which NHS treatment under payment by results has been key. Furthermore, it is possible that the individual serving the community sentence goes on to desist for reasons that have nothing to do with public services, such as entering into a relationship or simply moving to another area.
Questions arise on how the government can avoid problems such as double payment, or paying for a result that providers are not themselves responsible for.
Thirdly, we fear that payment by results will lead to cherry-picking by providers, as the inevitable focus will be on those individuals who are most likely to deliver a result. For example, we have examined Project Daedalus, a London-based initiative described by ministers as “the first payment by results programme to address resettlement and reduce reoffending”. A key part of Daedalus is a resettlement unit in Feltham Young Offenders Institute. The Heron Unit differs from other parts of the youth justice secure estate in that young people are placed there on their willingness and motivation to change, rather than on the extent of their vulnerability and needs.
This, we fear, is blatant cherry-picking. Indeed, an independent evaluation by IpsosMori concluded that the young people in Heron “indicate low levels of need in relation to areas of education, employment and accommodation, which run contrary to the aims of the programme”.
Cherry-picking can also extend to what exactly is a ‘result’. Private providers will take only limited risks, while voluntary sector providers require relatively narrow gaps between performance and payment, given their limited cash reserves.
It is likely that the lion’s share of any payment will be delivering a sentence to completion, with the supposed central aim of reducing reoffending constituting little more than a limited bonus. Similarly, measures of reoffending may be manipulated. The social impact bond pilot in Peterborough, for example, uses a reduction in reoffending over one year as its method of measuring results. While measuring reoffending over one year – rather than the traditional two years used in statistical bulletins – may help narrow the gap between payment and performance, it also ensures that a great deal of reoffending will not figure, as much of it will not be dealt with by the system within the one-year period. It is precisely because reoffending rates are actually based on reconvictions that statisticians use the two-year measure as a more accurate reflection of adult reoffending.
Ultimately, experiments in payment by results may prove irrelevant. Regardless of the progress made by the pilots, the Ministry of Justice must cut its budget by 6% year on year for the next four years. Squaring a reduction in the prison population with significant reductions to probation budgets represents a clear and present danger that may scupper the rehabilitation revolution before it has even started.
Andrew Neilson is assistant director at the Howard League for Penal Reform
