Social Impact Bonds

PDF version

by John Loxley

Social Impact Bonds (SIBs) are a relatively new method of funding and delivering social services. Under this approach the private or social sector finances and delivers services under contract to the public sector, against a bond issued by the public sector, promising to accomplish specified delivery criteria. If the criteria are, in fact, met over a specified period of time, then the private or social sector agencies cash in the bond, receiving reimbursement of their costs plus a rate of return based on performance. SIBs are not, therefore, bonds as traditionally defined, a title ascribed to financial instruments with a fixed lifetime and a fixed interest return, but are, instead, better seen as a form of public-private partnership (P3), in which finance, service delivery and, supposedly, risk, are devolved from the public to the private sector (see Loxley and Loxley, 2010). The financial instrument itself has more in common with venture capital than with bonds as, if performance targets are not met, the financing is not repaid.

The UK Experience as Model

SIBs were first introduced in September 2010 at a prison for short-term, male, offenders, in Peterborough, UK. An intermediary group of, in this case, a consortium of charitable organizations and philanthropists, as opposed to private companies, acted as a financing and organizational intermediary and selected a group of four social sector agencies to work with inmates and ex-inmates to reduce recidivism (Deloitte, 2012, p.6). If the interventions are successful in reducing recidivism over a six year period to 7.5% or more below that in the equivalent prison population used as a comparator, then the £5 million bond will be paid out with a return of between 7.5% and 13.0% depending on performance. Prime Minister David Cameron has expressed the view that “By the end of 2015, I want to see payment by results spread right across rehabilitation” (Eyre, 2012).

Other areas of social service intervention in the UK for which SIBs are reportedly being considered, are ‘cutting school truancy and exclusion; increasing youth employment; reducing acute hospital care by improving community support; improving provision of fostering to cut the cost of residential placements for children in care’ (Easton, 2011). The City Council of Cardiff, for instance, is considering SIBs to deliver foster services to high risk children with complex needs (Eyre, 2012).

Interest in using SIBs has been expressed in the USA at federal, state and city levels (Gross, 2012).  In November, 2012, the Minister of Human Resources and Skills Development, Canada, announced that the Conservative government would also be promoting SIBs (Whittington, 2012). Given that delivery of social services is mainly through provincial governments, one can expect pressures for SIBs at that level too.

How SIBs Are Structured

Chart 1 below explains the operation of SIBs. The government department or agency initially decides which social services might be appropriate for delivery through SIBs. It enters into a contract with an intermediary specifying the social outcomes that are expected from the arrangement, the cost of the service delivery or bond, and the rate of return to be paid if expectations are met.

13.02.06 chart

With contract in hand, the intermediary then raises capital upfront to finance the contract. The financiers could be charitable foundations or the private sector but it could even be the public sector itself that raises money (Mulgan, et al., 2011, pp 8-9). The intermediary contracts with appropriately qualified social service agencies to deliver the services and provides the capital to finance the services for the duration of the contract. If targets are met, the government then pays the investors through the intermediary to repay the bond and the additional rate of return specified in the contract.

The Underlying Rationale

Underlying the promotion of SIBs is the notion that early intervention and reduction of social problems can yield significant long-term savings to government so that an initial outlay of funds, if successful, can yield a very high social return on investment or SROI (SROI Network, 2012).  For example, Action for Children, a UK charity has estimated that for every £1 spent on early intervention, ‘society benefits by between £7.6 and £9.2’ (Easton, 2011). Such returns would be very attractive to government and would easily outweigh financial returns paid on SIBs. In other words the savings flowing from achieving outcomes would exceed the costs of achieving those outcomes. Other attractions of SIBs are said to be (Mulgan et al.2011, Deloitte,2012, Easton, 2011, Kramer, 2012) :

  1. SIBs promise to improve the outcome of social programs without increasing taxes, as it is the savings resulting from reduced demand on social services that finances that improvement.
  2. SIBs are premised on evidence-based results rather than just on activities. If the results are not there or cannot be verified, no payment is made.
  3. This approach to financing is said, therefore, to transfer risk from the government to the social or private sector.
  4. It also offers stable and predictable amounts of working capital to fund social service interventions, which is often not the case now.
  5. Participating social groups have a common interest in their interventions succeeding, so their interests are aligned.
  6. This arrangement allows for creativity and innovation in addressing social problems.
  7. Many social problems require several types of social intervention and the structure of SIBs can allow for this in the selection of cooperating service delivery agencies.
  8. SIBs are said to bring to the table the specialist expertise of governments (in facilitating, policy making and designing financial incentives), social financiers (financial analysis and funding) and service providers (technical knowledge).

The appeal of SIBs is obviously related closely to widespread fiscal austerity in the UK and elsewhere. Successful SIBs would have a twofold impact on the budget. Initially they would reduce government spending as services become funded by outside sources. By the end of the bond period, improved social service outcomes would have reduced the demand for government services and hence, even after paying off the bond, would save the government money.

What is Needed for Successful SIBs

For the SIB approach to be considered the public sector has to satisfy itself that some important prerequisites are met. To begin with, services to be delivered through this mechanism must have clearly measurable impacts, to which cost savings can be attached. Secondly, the overall savings in social spending must exceed the cost of service delivery plus the social rate of return. Thirdly, there must be a clear and objective mechanism in place for evaluating outcomes. Fourth, improvement in social outcomes must be clearly and unambiguously related to the service offered through the SIB and not due to other factors. Fifth, this may require that an independent agency be appointed to both set the target outcomes and assess the actual outcomes. Performance may need to be assessed against a control group of the population not receiving services through the SIB mechanism. All of this may need considerable work and a departure from current ways of doing business. The result is that transactions costs of introducing SIBs may be quite significant (Kramer, 2012) and these will also need to be covered ultimately through service cost reductions and savings.

Challenges and Potential Problems

SIBs imply a cost of borrowing to the public sector which is well above the cost it would incur if it borrowed directly. In this respect, SIBs are no different from P3s. Ascertaining and measuring risk transfer and service cost reductions that justify this are also problems which P3s have in common.  The case has to be established that direct public funding and service delivery cannot be equally if not more effective in delivering specific outcomes; that government direct spending cannot be made results oriented and open to innovation and creativity.

Measuring outcomes is a major area of concern in SIBs as the tendency will be for funders and service delivery agencies to pick off low-lying fruit, or more easily achieved targets, and leave more difficult problems for other service providers to deal with. Furthermore, social problems are often multi-faceted so that unemployment may be a complex function of education, housing, child care, addictions etc, and may not be susceptible, therefore, to a simple one dimensional service approach. Assessing performance could consequently be quite difficult, and any cost savings achieved might also be multi-dimensional and difficult to measure.

Savings may take the form of both additional revenues (if the unemployed return to work and pay taxes) or reduced costs (lower EI or welfare payments, reduced costs to the justice system etc.). Cost savings may be in the form of recurrent costs or the more difficult to handle reduced fixed costs. All of this adds to the complexity of measurement.

Given these complexities, a case could be made that financiers will tend to be risk averse, taking on only the easiest of challenges and foregoing both risk and innovation, contrary to claims made by SIB proponents. Competition may even arise between social service agencies for access to stable SIB funding of projects that seem to offer a high probability of success. At the same time, if funding is from commercial sources, it will only be forthcoming with quite large rates of return, possibly in the range of 20-30% p.a. (Mulgan et al. 2011, p.37).

SIBs can also be seen to be a way of reducing the public sector and possibly also public sector wages and worker protections. Replacing tax funding by charitable or private funding may seem appealing but it may come at the cost of public sector workers and of consistency in service delivery across the board.

Budgeting for SIBs will raise new problems. First of all, will they be booked as future contingent liabilities and, if so, how does this affect debt and credit ratings? Governments are unlikely to put aside the money they save when services become funded by outsiders, so what happens if and when pay-outs are needed for SIBs that meet their targets? Will they be paid out of new borrowing? Will SIBs be renewed? There are no answers to these questions at this time but these issues will need to be resolved if SIBs are to become fashionable and if claims that they will raise total funding for social programs are to be verified.

The possibly high risks attached to SIBs may, however, mean a low uptake in terms of financing. As well, charitable foundations are unlikely to be able to finance large, expensive initiatives, even though successful outcomes may require a minimal scale of operation for them to be statistically significant (Mulgan et al., 2011).  The temptation would then be for the government to make funds available to encourage a larger uptake of SIBs and larger SIB projects, as the federal government has done with P3 financing and as the UK government appears already to have done with SIBs by introducing the Social Outcome Fund (Cabinet Office, UK, 2012).


SIBs are a relatively new way of delivering social programs and are being promoted in the UK, where they originated, the USA and Canada. They promise to reduce funding costs and improve service delivery, and reduce immediate demands of programs on scarce government funding. This promise is, however, based on meeting quite stringent assumptions which might be difficult to meet in practice and which might possibly have harmful side effects. The result is that, at best, SIBs are likely to offer only a limited amount of outside funding and perhaps even then only aimed at meeting service delivery targets which are low risk and easily met.

Before embarking on SIBs, the government of Canada would be well advised to concentrate its efforts on improving the funding and delivery of services in and by the public sector.

For a copy of the bibliography, please see the PDF version, or contact the CCPA office at ccpamb@policyalternatives.

The comments of Shauna MacKinnon and Jesse Hajer on an earlier draft are gratefully acknowledged.

John Loxley is a professor in the Department of Economics at the University of Manitoba.