The Public Accounts Committee (PAC) has just published a damning report about Transforming Rehabilitation (TR). It argued that the Ministry of Justice “not only failed to deliver its ‘rehabilitation revolution’ but left probation services underfunded, fragile, and lacking the confidence of the courts.” It is unfortunate that the large-scale failure of TR will be seen as another negative episode in the government's increasingly chequered history of efforts to improve public services markets.
When it was launched, TR was billed as the successor to the Peterborough Social Impact Bond. This was largely seen as successful, as shown through evaluations of its collaborative delivery model and in the measurement of the results it achieved. The fact that TR shares relatively few features with Peterborough indicates the way government designs and implements its contracts with external providers really matters.
Start with the implementation failures. Transforming Rehabilitation (TR) was developed under Chris Grayling, a politician with serious pedigree in bombastic Payment by Results experiments: the large-scale and now defunct Work Programme was also his baby. TR was a particularly extreme form of experiment, where major responsibilities were to be passed on to private sector providers. When the contracts went out to tender, they already sought to reduce the cost to government of delivering on these responsibilities. This downward cost pressure was made worse because the providers who eventually won had 'underbid' to assure themselves of winning tenders. Alongside this issue of funding insufficiency, the government lacked assurances around minimum services standards and didn't subject providers to enough scrutiny. Amongst all this, the people who the services were to be for - ex-offenders - had no voice whatsoever.
These failures of implementation happened despite TR adopting design features which we already knew made it particularly innovative, and therefore risky for government. Like the Work Programme, it used payment-by-results, where some payment (in the case of TR) or all of it (in the case the work programme) was withheld until the provider showed evidence that it had achieved the objectives desired (like less re-offending or more welfare users in work). Also like the Work Programme, it used so-called 'black box' commissioning, where the private provider has total flexibility on how they run services.
Research into both TR and the Work Programme has identified that the combination of risky innovations with failures of implementation exposed both the government and providers to a constellation of risks. The contracts were not able to balance tensions between the government's desire for efficiency, service users' need for quality, and private providers' profit-seeking.
Individual service users were particularly vulnerable in these situations as the quality of services declined in response to providers cutting costs. These people were not able to influence service quality by 'voting with their feet', nor by influencing what happens at the ballot box.
We are keen to share our perspective and see what lessons can be learned from policy challenges and failures. Feel free to get in touch with us at golab@bsg.ox.ac.uk or interact on twitter. We would like to know what your thoughts are.