HMP Peterborough Social Impact Bond
Social Impact Bonds as an advanced tool to solve the social problems, spend public money in more effective way and develop social entrepreneurship.
Authors and organisational affiliations: Dharmendra Kanani, Phil Messere, Joanne Bye, Matt Roche, Big Lottery Fund; Tom Symons, Social Finance
1. Context and history
The Ministry of Justice (“MOJ”), the Big Lottery Fund and Social Finance launched the world’s first Social Impact Bond (SIB) in September 2010 to finance the One* Service intervention to reduce re-offending by 3,000 short sentenced male prisoners leaving Peterborough prison.
At the time the SIB launched, there was no statutory provision for offenders who had served sentences of less than 12 months. Reoffending rates amongst short sentenced prisoners are high. Around 60% are convicted of at least one offence in the year after release. But, they leave prison with typically £46 in their pocket, often with nowhere to live, no job to go to and no family waiting for them. A high proportion of them leave prison only to return a few weeks later.
The Big Lottery Fund has made an arrangement to provide £6,255,751 by 2018, primarily in the form of outcomes payments (which are payable in 2016 and 2018) to repay those who have invested in the SIB, provided certain agreed outcomes with respect to reduced recidivism are achieved. The Big Lottery Fund decided to provide this funding as the proposal had the potential to bring real innovation to a really stubborn social issue faced by those most in need. It was part of the Big Lottery Fund’s Replication and Innovation funding stream and it was clear that the proposal would not be able to proceed without this support.
The Project has generated ministerial interest since it was launched in 2010:
2. Main characteristics of the Peterborough SIB
The aim of the Peterborough SIB is to reduce reoffending rates. The Social Impact Bond contract with the MOJ stipulates that the investors will receive a return if the number of reconviction events among Peterborough prison leavers, triggered by offences committed within a 12 month period following release, falls by 10% for any cohort or 7.5% overall. If the Social Impact Bond delivers a drop in reoffending beyond 7.5%, investors will receive an increasing return effectively capped at a maximum of 13% per year over an eight year period. The payments represent less than half our estimate of the saving in costs to the prison system as a result of the reduction in the re-offending. In the event that the interventions fail to deliver a reduction in the rate of re-offending, the investors receive no payment.
Peterborough Social Impact Bond structure:
On the back of this contract, £5 million was raised from 17 investors, primarily from the UK and two from the US, to fund the product. The Big Lottery Fund agreed to fund the outcomes payments with the MOJ.
St Giles Trust is the main delivery agency for the project:
3. Outcomes and evidence of good practice
The Peterborough SIB can be seen as good practice in relation to both the mechanics of a new way of financing and tackling stubborn social issues and the impact that it is having on reoffending rates.
Success of the SIB model
In terms of the mechanics of the SIB itself these seem to be working very well to date. Investors, commissioners, service users, delivery organisations and Social Finance themselves are content with progress to date. The investment for the SIB was raised successfully and the co-ordination of activity on the ground seems to be working well. In particular the SIB, as modelled in this instance, has Social Finance playing a co-ordination role in Peterborough, with the freedom to move where funding is channelled to support the best possible outcomes. This is demonstrating that this model provides flexibility in a way that grant or traditional contract funding can’t.
In broader terms the Peterborough SIB has been picked up as a best practice example throughout the world, leading to more SIBs being set up in the UK and overseas and more funding being channelled to support further SIBs.
Impact on reoffending rates
The SIB is not measured by a binary measure (whether the offenders reoffend) but rather by a frequency measure (the number of reconviction events). This encourages the providers from the One* Service to work with all the clients, including those who are the most prolific reoffenders. The frequency of reconvictions for each cohort (which takes on average two years to be released from prison) will be compared by an independent assessor to that of a statistically propensity scored matched group of prisoners across the UK from the Police National Computer.
Interim figures released by the MoJ suggest that there has been a 12% decline in the frequency of convictions events per 100 prisoners since 2008 in Peterborough alongside an 11% increase nationally. However we cannot infer from the figures whether the pilot has been a success or whether an outcome payment will be triggered because the measurements used are distinct from those which will be used for the Peterborough Social Impact Bond. The Peterborough cohort will be compared by an independent assessor with a statistically compiled control group.
The first results of the Peterborough Social Impact Bond are expected in April 2014.
4. Outcomes (for different stakeholders)
5. Strengths and Weaknesses
- allows government to test innovative approaches without taking financial risk
- gives sustainable funding to VCSE organisations
- works with service users in a holistic way that aims to address all of their needs
- uses an adaptive system and real-time data to provide flexibility to the service, feedback loops on what works and data-led decision making
- outcomes data is only available on an annual basis and only from Year 4 onwards. While it was necessary for outcomes data to wait until year 4, outcomes data at more regular intervals would enable better real-time decision making
- SIBs more generally have the scope to produce perverse outcomes
- it is unclear how ‘cashable’ actual savings will be (given that they would require de-commissioning to realise)
6. Comparisons with other experiences
A PbR approach for short-sentence offenders has been piloted in Doncaster prison. This is operated by Serco. For the duration of the pilot 10% of Serco’s annual revenue is contractually dependent on them achieving a five percentage point reduction in reoffending rates. If former prisoners end up back in court and are convicted - on any charge within a 12-month period - then their revenue is affected.
In alliance with long-term voluntary sector and social enterprise partners - Turning Point and Catch22 - Serco designed a scheme that can be adapted to meet the specific needs of individual offenders. To do this, the Ministry of Justice gave the flexibility to make decisions at a local level.
At a broader policy level, the Government’s Transforming Rehabilitation programme has sought to introduce statutory provision for all short-sentence offenders.
In relation to SIBs there are a number of other now in operation, which are worth reviewing in comparison to the Peterborough SIB, including:
- A SIB in London aimed at reducing the number of NEETs
- A SIB in Essex that is aimed at reducing the number of young people going into care
- A SIB in New York that is also aimed at reducing reoffending rates
7. Overall assessment and transferability
Due to early successes of the Peterborough SIB and the different approach it brings to both tackle entrenched social problems and make savings to the public purse, it has been used as a model for the development of SIBs throughout the world.
It has also led to additional monies coming on stream from the Big Lottery Fund and the Cabinet Office to support further SIBs, via their outcome funds.
The marriage of social investment with effective voluntary sector providers like St Giles has been a model emulated in other social investment programmes and projects seeking to draw in new funding. Learning has been shared through round table and stakeholder events in 2013 and 2014 as well as in publications and conferences such as the Good Deals Conference in October 2013.
In the opinion of the peer reviewers gathered during the meeting in Warsaw on 10-11 April 2014 the case of social impact bond represents the innovative approach towards solving social problems relying on:
- Promoting result oriented performance
- Involving private funds
- Executing the cooperation and involvement of different actors and stakeholders
- “allowing government to try new ways without taking any financial risk” (Czech comment)
At the same time all the commenters stress the role of preconditions which have to be fulfilled before the implementation of social bonds could happen in their countries. Cultural and legal differences appear to be the most important in many cases. The model of social impact bonds seems to be complicated and in the opinion of many participants requires long and detailed preparatory work adjusting the solution to the reality of particular country. Many questions remained without an answer like what are the underwriting costs (the costs of attorneys, consultants, program evaluators) within the total investment, how exactly the flow of finance between the ministry and social enterprises looks like, who in the government is in position to start this kind of investment and so on.
The case was found very interesting by all the participants and four of them (Finland, Sweden, Italy and Hungary) got inspired from it and started to translate it into local circumstances.
HMP Peterborough Prison:
References and further information
London NEETs SIB
Essex Edge of Care SIB
New York Recidivism SIB
For further information contact the Big Lottery Fund by emailing firstname.lastname@example.org