Social impact bonds: A good new idea

Here’s a cross-post of my latest piece at The Conversation, about why social impact bonds are a good idea.


What’s new in public finance you ask? Social impact bonds, that’s what. This is the leading edge of the social finance revolution. It just might make a big difference to Indigenous and remote Australia, and, curiously, arts and cultural funding is mixed up in this.

I recently had the honour of speaking at a meeting in Broome that was hosted by two local non-profits – the Kimberley Institute and Goolarri Media Enterprises – on the subject of Social Impact Bonds (SIBs), which are a new and still experimental public-private model to finance the delivery of social services.

The motivation for this interest in SIBs is crisis and opportunity. Paul Lane, Director of the Kimberley Institute, said during the meeting that the Western Australian government is facing spending something like 20 percent of its health budget in the Kimberley, where just two percent of its people (including the Yawuru) live. Similar statistics run across public housing and prisons. In the face of sustained effort and vast spending on inputs, the far North-West is home to chronically failed social service outcomes.

It is not the people that have failed, for hard-working and dedicated service providers are doing their best within the constraints of the current funding and delivery model. The problem is not the people; it’s the model. Social impact bonds (known as social benefit bonds in Australia) may represent a community and private investor driven chance to break into a new model.

Because SIBs focus on outcomes, rather than inputs, they enable more innovative or experimental approaches to service delivery. Howard Pedersen of the Kimberley Institute said the language and culture programs, including the Jandamarra productions, were a significant component of building community capacity to take on social change around alcohol problems in remote Fitzroy Crossing.

The standard model of government grants to service providers inevitably pits providers against each other over scarce funds. Government connections become highly valued, turf warfare breaks out and providers have little incentive to work together. Spending flows to crisis situations, leaving little available for early (and much lower cost) intervention.

Worse, when a program does succeed – as happened to the Women’s Collective in Fitzroy Crossing – it is rewarded by having its funding cut. The incentives embedded in this system are perverse from top to bottom. Kevin Fong of Goolarri thinks that SIBs will enable the community and providers ‘to gang up on the problem, not on each other.’

Social impact bonds have the potential to revolutionise public finance and public service delivery. The basic idea of a social impact bond is to use private sector investment to finance a social project, delivered through a service provider, whereby if the project achieves certain pre-agreed results, which are measured by an independent evaluator, the government then pays the private investors (the service providers are paid by the investors through an intermediary, and receive the funding regardless of outcome). If the project fails to achieve its contracted outcomes, investors lose and the government (and the taxpayer) pays nothing.

The first and most well documented SIB was at Peterborough Prison in the UK to reduce rates of recidivism in short-stay prisoners, where the social benefit comes from the prospect of reduced spending on prisons and the costs of crime to the community. The project started in 2009, and is about half way through now. About 20 further SIBs appear to be underway in the UK, many focusing on disadvantaged youth.

More than 40 SIBs are currently in development worldwide. The Obama administration’s 2012 budget committed US$300 million to develop SIBs. SIBs are underway already in Australia. In part the burgeoning of SIBs is because of mounting government indebtedness, but there is also a sense that this is a good new idea whose time has come.

SIBs are not for everything, and they are unlikely to replace core funding of mainstream social services. But they do have a place at the frontiers in dealing with underfunded ‘wicked’ problems that often require local knowledge and initiative, and problems that may benefit from new approaches but that are too difficult because of entrenched interests, bureaucratic hurdles (usually, ironically, about transparency and accountability), and the constipating effects of public funding silos.

SIBs are a better way because they get the incentives right. They focus entirely on the outcomes, and leave the way in which that is achieved up to the service providers, who are therefore free to engage in experimentation to discover what works (in public funding models this is usually far too politically risky in case of failure). SIBs thus promote fast and effective learning about what works. Service providers can intervene early rather than waiting for crisis to develop.

SIBs get the incentives right because they shift risk from public sector to private investors, which is to say that they shift risk of experimentation in service delivery to where it is best carried.

Government, under media scrutiny and with taxpayer money, only becomes involved at the end, when, if the performance outcomes specified in the social impact bond are achieved, then investors get paid. The taxpayer only gets value for money here.

These investors are initially likely to be philanthropic foundations or socially motivated wealthy individuals, with investment coming from corporate social responsibility funds next, but it is entirely conceivable that such bonds can be sold to for-profit investors, such as pension funds.

Now I want to ask what else SIBs might be used for, beyond that which they have been proposed in these early days in addressing socially costly problems such as disadvantaged youth, recidivism, and homelessness.

These SIBs have proceeded because they are easy to count, and the interventions are neatly set up as a controlled trial. But might SIBs be a source of arts and cultural funding?

Here is how that might look. A group of “stakeholders” would identify a costly social problem that could be addressed through an arts and cultural intervention. They would undertake initial cost-benefit analysis to estimate the savings to government spending down the line, which let’s say is $100.

They might estimate that these savings come about by delivering a service now costing $50. There is a risk this will fail, as the service is unconventional and untried. So it is politically risky, and thus unlikely to get public funding.

The stakeholders would operate through an intermediary (such as Social Ventures Australia) to approach private investors, who are in effect “social venture capitalists”. If they agree with the assessment of value creation, and like the management team, then this is pitched back to government.

The intermediary puts all this together, writes up a SIB contract that is agreeable to both investors and government, which will be somewhere between $51-$99, so say $70. Crucially, they also identify and agree to some measurable outcome that will trigger the bond payment. A further independent party is contracted to run the evaluation.

When the bond is issued, investors transfer $50 to the intermediary, which contracts the service providers. They now have their funding. Five years later an evaluation is conducted of the outcomes (not of the inputs or methods, just of the outcomes).

If they meet the agreed threshold, the bond triggers and government pays the investors $70. Investors are now up $20, and government has saved $30. If the results are not achieved, investors lose $50, and government pays nothing (government also avoided the risk of that failure).

It is silly to think of funding say Opera Australia this way. They do produce something of intrinsic value, and there is probably a positive externality, but it is a much bigger stretch to argue that Opera Australia does something that in the long run saves public money.

Yet the examples from the Kimberley do illustrate cases where this may well happen, but these arts and cultural interventions are chronically underfunded and the first to be cut. They do not stand up well under media and political scrutiny, even when local providers have good reason to believe they work.

I was involved in evaluating one of Goolarri’s projects a few years ago called Kimberley Girl, which fits this bill. It’s nominally a beauty pageant for Indigenous young women in the Kimberley region. But really it’s an intensive workshop program in developing confidence, self-respect and presentation skills among disadvantaged youth: workplace-readiness in other words.

It can be life-changing for some of the participants, and that can make a huge difference to the public budget down the line. We estimated that one dollar invested in Kimberley Girl returns seven dollars in future public savings.

These types of unconventional intervention, delivered by an Indigenous media company, of all things, are not priorities for government funding. But they work.

A social impact bond can enable private investment to carry the risk of testing such programs, with the taxpayer only involved to compensate those programs that succeed and might then be scaled up and delivered more conventionally.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s