Do impact bonds live up to their promise as an efficient financier of social services? Probably – but there is no conclusive proof, says a leading research economist into the instrument.
Emily Gustafsson-Wright, senior fellow in the Global Economy and Development Program at the Washington-based Brookings Institution evaluated with her team the first ten years of impact bonds.
This resulted in a recently published series of five policy briefs, analysing various dimensions of the success of this special partnership between private investors, governments and social service providers.
So what exactly are impact bonds? For starters, they aren’t bonds. Actually an impact bond is not a financial instrument like regular bonds. Rather, it’s a public-private partnership.
Government and private investors enter into a contract to tackle a social problem and then hire a service provider, e.g. a social enterprise or a not-for-profit organisation for the practical implementation. It is a pay-for-success contract with an easily measurable ultimate goal, where the impact investor takes on the financial risk.
“The thing that interests me most about impact bonds is their focus on outcomes’’
“Figuring out what works and what doesn’t, all along the process. That’s what makes impact bonds special: it’s all about tailoring services to the needs of the beneficiaries.”
Two examples: Prevention and rehab centres
In the county of Essex, UK, a social impact bond succeeded in creating more stable and supportive environments to prevent children from entering out-of-home care. The programme was commissioned by the Essex County Council and executed by the charity Action for Children. Multiple investors – including Big Society Capital, Bridges Ventures and Ananda Impact Ventures – contributed £3.1mn.
The International Committee of the Red Cross launched a humanitarian impact bond to build and run three new physical rehabilitation centres in war-torn African countries, providing services for thousands of people. Nine private investors coordinated by the Swiss bank Lombard Odier financed the total cost of €17mn.
Outcomes achieved
These are but two examples out of 214 impact bonds in 35 countries that have been registered thus far, according to the Brookings Institution Global Impact Bonds database. Private investors covered over $437mn of upfront capital, required for providers to set up and deliver services. They are reimbursed – usually by governments – only if the projects meet the targets.
An analysis by Brookings of the nearly 50 completed impact bonds shows that outcomes have in fact been achieved and investors have been repaid in all cases but two. Investor returns range from 1% to 20% of the original investment.
Most of these impact bonds fall in the social welfare and employment categories.
The growth and development of impact bonds are being tracked by two international institutions, the Government Outcomes Lab (GoLab) at the University of Oxford in the UK, and the Brookings Institution.
Impacting the system
The impact bond market has grown steadily over the past decade but remains very small, relative to the size of the global investing market. The average upfront investment is just $3.16mn.
A few projects with much higher levels of upfront capital inflate the average, such as the Red Cross project mentioned above, a $25mn environmental impact bond in Washington DC, or the more recent €16.7mn Joining Forces impact bond in the Netherlands, commissioned by the Ministry of Defence.
“The relatively small size of the market however doesn’t mean that its impact is insignificant,” says Gustafsson-Wright. “Either within each impact bond or within the context it operates in.”
“Our analysis of completed impact bonds,” she says, “shows that they tend to impact the system of social service delivery and financing by shifting mindsets and building stakeholder capacity.”
No rigorous evidence
Still, there is no rigorous evidence thus far that the benefits of impact bonds outweigh the costs. Gustafsson-Wright, who has a background in conducting impact evaluations, emphasises that “to do the perfect research you need to compare a project that is managed through an impact bond with one that is not. That’s very hard to organise.”
Another complicating factor is that most impact bonds focus on interventions that prevent negative and costly outcomes in the future, such as the ones attempting to reduce homelessness or children placed in out-of-home care. This makes it hard to put an accurate figure on the economic returns.
“We don’t know the counterfactual – what would have happened in the absence of the impact bond – and we don’t know if there is a cheaper way to get those outcomes and ecosystem effects,” she says. More research is needed.
Gustafsson-Wright is however still fascinated by the potential of impact bonds: “They bring actors to the table that otherwise would not be working together.”
