Payment by Results: Hopes, Fears and Evidence

Abstract

AbstractThe idea that government should move away from paying for services to be delivered to paying in proportion to the level of reduction in reoffending achieved has obvious attractions, particularly in the current economic climate. But will adopting such a 'Payment by Results' model actually deliver the benefits claimed by its proponents? This article reviews some of the potential obstacles to success. As Payment by Results is a relatively untested concept in the field of criminal justice, the article draws on the experience of this approach in employment and in health where it has a longer history. The results suggest that the chances of Payment by Results leading to a reduction in reoffending in crime or in costs are slim. Adopting such an approach may also carry insidious, as well as obvious, dangers.KeywordsPayment by Results; re-offending; evidence-based policy; cost effectiveness; Transforming RehabilitationIntroductionThe Ministry of Justice (2011:2) makes a bold claim in its response to the Breaking the Cycle consultation and on its website:We will change our whole approach to the management of offenders and their rehabilitation, so we only pay for what works in delivering reduced levels of crime.3On the face of it this 'Payment by Results' (PbR) idea is hugely attractive, particularly in the current economic climate and under a government which is so keen, not only to reduce public expenditure, but also to reduce the scale of the public sector. Why should we pay for the delivery of a service which may or may not bring about the desired change, when it is possible to transfer the cost of failing onto the supplier. Paying only when the promised change has been achieved, and in proportion to the degree of change affected, is obviously preferable. Moreover, as Fox and Albertson (2011) explain, there are other potential benefits to PbR. The sheer possibility of making profits is expected to bring new providers into the field. This, combined with a financial incentive to achieve outcomes, is then expected to increase competition, sweep away unnecessary bureaucracy, and increase the desire to innovate. In turn, these developments should lead to a better understanding of what is effective. Also, if the anticipated gains in efficiency and effectiveness are realised, this will lead to lower unit costs (even allowing for an element of profit to be paid). The remaining savings can be ploughed back to pay for additional activities in this, or another, area of public policy. Additionally, under the Social Impact Bond variant, the possibility of making a profit will not only bring in new suppliers but also financial intermediaries who are expected to finance the service provision in advance of impacts being achieved, in return for a subsequent slice of the profits. This is expected to facilitate the involvement of smaller providers who may not have the financial resources to sustain a service over several years while waiting for their results to be evident and their payments to be triggered.The idea of payment by results is new in the field of criminal justice, so it is important to consider what evidence there is to support the Ministry of Justice's (2013a) hope that the putative benefits of PbR will be realised. In fact, the only empirical evidence which seems to support this radical policy shift is the positive findings of one small piece of research (Frontier Economics, 2009). This made some rather large claims about success from a weakly designed reconviction analysis in which a sample of prisoners serving sentences of at least one year at one prison, who received additional support in relation to some resettlements needs, was compared with a poorly matched national sample. However, payment by results has a much longer history in other fields such as health, so it is worth reflecting on that experience too. Unfortunately, a recent Audit Commission (2012a:3) review of this sort of material does not inspire confidence as it concluded that:'Our review of UK and international research evidence found few rigorous evaluations of PbR and no complete, systematic analysis of its effectiveness. …

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 28 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.