Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Understanding social impact bonds

Download PDFDownload PDF



Don Arthur

There is often a tension at the heart of government social programs: governments want to pay social services providers for results rather than inputs or process, but cash strapped providers need funding up-front.

Social impact bonds have been proposed as a way to solve the problem by bringing in private sector and philanthropic investors. Also known as social benefit or pay-for-success bonds, this innovative funding mechanism is being trialled by governments in the UK, US and in New South Wales. In its 2015-16 budget the Queensland Government announced plans to pilot three social benefit bonds.

Problems like chronic homelessness, poor educational attainment, and repeated criminal offending cost governments billions. If there was a way to intervene before these problems become entrenched, governments could save money and turn people’s lives around.

There is no shortage of social services agencies eager to deliver prevention programs for at-risk populations. The catch is that government funders can’t be sure which programs will work and which won’t. The risk is that they’ll spend money on programs that don’t work and still have to deal with the cost of entrenched social problems.

One solution is pay-for-success contracts. Popularised in the early 1990s by David Osborne and Ted Gaebler’s book Reinventing Government, pay-for-success contracts make providers meet the upfront cost of delivering the program. If the program fails, the government pays nothing. If the program succeeds, the government pays a success fee based on the outcomes delivered (for example the number of ex-prisoners who do not reoffend). If the fee is less than the government saves and more than the provider spends, then both the provider and the taxpayer come out ahead.

A problem with pay-for-success contracts is that most providers cannot afford to wait months or years to get paid. To run the program they need funding up-front. Social impact bonds attempt to solve this problem by creating a new kind of financing instrument that enables private sector investors to provide the upfront funding

Shutterstock image ID: 149921993, Copyright: ChameleonsEye


and earn a return if the program succeeds. In effect, the risk of failure is transferred from the provider to the investor in return for a right to the outcome fees if the project succeeds.

In practice it’s difficult for governments to make the social impact bond model work in its pure form. The first problem is results are often hard to measure. For example, a pay-for-success contract for a welfare-to-work program can’t simply pay a fee for every participant who moves from welfare to work because many welfare recipients will move into work whether they participate in the program or not. The government will usually need to pay for an independent program evaluation as well as for a success fee. This adds to the administrative complexity of the deal as well as to the cost. The temptation is to move away from a strict pay-for-success model to reduce complexity and cost.

Another problem is that private sector investors are often unwilling to accept the risk of losing their entire investment. In the US, New York City dealt with this problem by getting a philanthropic foundation to guarantee $7.2 million of the private investor’s $9.6 million stake. Another approach is to guarantee investors a minimum return even if the program fails.

Administratively, social impact bonds are complex. They involve a provider, an evaluator, investors and, usually, an intermediary organisation to coordinate the project. Everyone needs to get paid. The challenge for governments like Queensland is to create a model that performs better than the traditional model of directly funding pilot programs, evaluating them and then scaling up those that work.