SOCIAL IMPACT BONDS AND SOCIAL VALUE

Social Impact Bonds are a financial tool being developed in the UK (by the Young Foundation and Social Finance) to provide a new way to invest money in social outcomes. Their key innovation is to link three elements:

• investments (by commercial investors or foundations);
• a programme of actions to improve the prospects of a group (for example 14-16 year olds in a particular area at risk of crime or unemployment); and
• commitments by national government to make payments linked to outcomes achieved in improving the lives of the group (for example, lower numbers in prison, and lower benefits payments).

Various versions of social impact bonds have been developed in the past by a number of people associated with the Young Foundation. More recently work has been done as part of the Innovation, Justice and Youth programme, building on concepts associated with ‘justice reinvestment'. The concept is now being developed in collaboration with national and local government as well as foundations and we hope that a version of the model will be piloted shortly.

Finance for social outcomes
Many forms of finance go towards achieving social outcomes. These include mainstream public spending of all kinds, grants, loans, equity investments by RDAs and convertible grants. There is growing interest in more innovative approaches to finance ranging from advance market commitments (AMCs) for purchasing of pharmaceuticals to local bonds (for example in Sheffield), to the burgeoning field of social investment. Some local tax tools - such as Tax Increment Financing, and Development Taxes, have also sought to capture value in novel ways.

Misaligned incentives
In the UK there has long been particular interest in designing new financial tools to address misaligned incentives in social policy, for example:

• that local authorities or NGOs responsible for providing services to young people do not share the benefits from reductions in prison numbers or benefits bills;
• that there are few incentives for agencies to invest heavily in early years support, despite strong evidence on the long-term social gains;
• that health prevention often involves action by agencies such as schools which have inadequate incentives to act

Past experiments
These misalignments prompted many of the experiments with joined-up government in the 1990s and 2000s (such as the Rough Sleepers Unit which pooled budgets), and there have been many innovative approaches designed to both map and realign the costs and benefits associated with actions. A recent example of an attempt to do this was a revised ‘Green Book' for investment developed by the UK government in the early 2000s to compare investments in programmes like Surestart, training for teenagers and higher education in terms of a Net Present Value. The aim was to capture the full range of potential costs and benefits, and to provide an equivalent guide to those which already exist for capital investment in such things as roads and railways. In practice however the range of uncertainties was too wide to make this useable. Other simpler approaches to the same problem have involved contracting, including the many examples of outcomes-based budgeting in Scandinavia and around the world. A notable UK example has been the Employment Zones which directly incentivised contractors to achieve outcomes in helping long-term unemployed people back to work, and more recently outcome based budgeting has been used in contracts around offending. For a survey of some of the issues involved in managing and capturing value see chapters 10 and 13 of Geoff Mulgan's new book ‘The Art of Public Strategy' (Oxford University Press, 2009).

The Three Models

Drawing on these lessons work has been underway to investigate the potential for new financial devices to better align incentives. This has pointed to three categories of approach which we loosely label Social Impact Bonds, in that they involve some borrowing with repayment linked to success in achieving social impacts.

Local authority Social Impact Bonds
In the first model a local authority or LSP would borrow on existing markets for a package of investment in a social impact programme (eg for teenagers at risk of NEET status) and would receive a series of payments in the future from national government if particular milestones are achieved associated with lower costs for national government. For example a city or London borough would borrow £5m for an intensive programme of work with NEETs or potential young offenders, and would be repaid according to the numbers who achieved educational qualifications relative to an agreed baseline of similar local authorities. The repayments would represent a proportion of the lifetime savings to national government (primarily through tax and benefits). Models of this kind are relatively easy to design and implement, involve relatively few players and transaction costs, though they require clear protocols on design, establishment of baselines, success measures &c.

Commissioning for social outcomes
A second model aims to directly incentivise a service provider or group of providers to take responsibility for part of an age cohort in a particular area, eg 5% of 14 year olds, with direct incentives to achieve educational and other goals by 19. This would extend the Employment Zone model, and again is relatively easy to design. Contractors would raise their own capital either through social investment sources (in which case they might be described as Social Impact Bonds, with tax treatment equivalent to CDFIs) or on the market. In the latter case there might be some risk sharing with an investor (eg a foundation). In all of these cases there are some important issues around risk transfer (and important lessons to be learned from the problems associated with PFIs, private prisons etc) as well as issues of accountability (in particular the local authority's responsibility for children).

Full Social Impact Bonds
A third alternative would share the risk for a bundle of interventions, with:

• finance raised from the market, with investors taking some of the risk for non-achievement of social outcomes

• action through a special purpose vehicle (potentially combining public sector, private and third sector) to manage a series of interventions with a target group

• and, again, payments based on results against benchmarks.

This model is somewhat more complex, with more handovers and transaction costs, but opens up a radical new avenue for bringing in new sources of finance.

Several fields have been proposed for bonds of this kind. These include: investment in early years programmes (based on the evidence from Abercedarian, HiScope etc for fairly large long-term paybacks); NEETs (focused on life time earnings) and youth or young adult offending; care leavers; and investments in health prevention and improvement. Another potential field for action is in employment creation during the downturn.

In principle the model is likely to work best in the short to medium term where:

- there is a reasonably short gap between interventions and measurable results.
- there are very tangible financial gains, for example the very high costs associated with prison places, as well as with crime
- the numbers of players are small, ie one primary national department, a local authority, finance body and other agencies working on contract.

All models bring some common challenges:

• Measurement - agreed baselines and metrics that are not vulnerable to economic downturns, national policy changes (eg new crimes being legislated), and shared analysis of lifetime costs and benefits associated with different actions and client groups.
• Action - all depend on there being a credible menu of actions to implement which significantly outperform existing ones, and they also depend on the presence of an implementation capacity. In most cases this is likely to involve a mix of public, private and voluntary organisations - in none of these fields does any one sector have a clear advantage in terms of performance.
• Risk - handling downside risks, including not only the risk of failing to achieve targets but also other risks, eg political risk (if some of the interventions are overruled by elected politicians).

Future work on the practical implementation of Social Impact Bonds will run in parallel with the Young Foundation's work on social value during 2009.