Are social impact bonds still the grants of the future?

Posted on 18 Jun 2019

By Matthew Schulz, journalist, Our Community

A few years ago, it seemed that everyone was talking about social impact bonds, a "pay for success" model for welfare-related programs, as the name suggests.

In reality, it's a complex and often costly business, because as grantmakers know, outcomes can be very difficult to prove.

Social impact bonds are a way of getting private investors to fund for-purpose programs, with their ethically minded investing repaid if sought-after outcomes are achieved.

It's usually governments that issue the bonds, and their pay-off is that social improvements - such as less crime - will also reduce the financial hit on the government budget in the longer term.

The main players in a social impact bond often include:

Government: In Australia this has tended to be state governments, which seek to spread some risks to the private sector, save on program costs, and access new funds.

Private investors: Many philanthropic funds, foundations and corporates have social development goals they want to reach. Often, they're happy to accept a reduced return for that social good.

Not-for-profits: Only larger-scale not-for-profits can handle the complexity of the arrangements currently required for these bonds.

Organisations seeking longer-term funding deals have been keeping a close eye on these financial instruments as an alternative funding source.

At the Australian Institute of Grants Management, we didn't have to go far to find someone with experience with these bonds. Cathy Truong, the executive director of our sister enterprise and giving platform GiveNow, has some great insights as a result of her past roles. We asked her for her thoughts about the state of play of these bonds.

Tell us a bit about your background in social impact bonds.

My first serious interaction with the social impact bond market was as the executive officer of the Trawalla Foundation. We invested in the Benevolent Society Benevolent Bond released in 2013. I also worked for six months at Impact Investing Australia, which was an organisation designed to drive the rollout of these sort of social investment instruments.

Are social impact bonds still a hot topic?

Cathy Truong
Our Community staffer Cathy Truong has a background working with social impact instruments

Social impact bonds came to prominence a few years ago. They were really "hot" in Australia back in 2014 when optimism was high. We are now in the "suck-it-and-see" phase. A number of social impact bonds were released between 2013 and 2017, and all are "running through their paces" as most commonly they have a five-year horizon.

Incidentally, social impact bonds also get called social benefit bonds or simply social bonds.

What's good about them?

They are perceived as a great addition for the sector, as they open the door for private sector funding. With most of the social issues the community deals with, we know that we could have significant impact with large-scale funding. And it's not realistic to expect this from government budgets or private donor drives.

A business corporate knows that when it needs to fund expansion into an activity it can't self-fund, it goes to the market asking for support in the form of a bond. Social impact bonds bring this "tool" and thinking to the social sector.

Another thing that is attractive about a social impact bond is that it forces participants to engage in quality data capture and rigorous evaluation frameworks. While not everything can be measured, many key indicators can be.

Sustained, quality data capture can be a valuable contribution.

That combination of access to large sums of funding, a strong evaluation framework, and a medium-term horizon allows for smart and innovative interventions to support the community.

Is this a live option for grantmakers in state, federal and local government agencies in Australia and New Zealand?

Social impact bonds work best for state governments - local governments are too small to absorb the start-up resources required to get this sort of instrument going.

At the federal level, the government in most cases, doesn't usually have control of the policies required to ensure smooth operation of the bond. That's partly because state governments are usually the existing funders for areas where these sort of interventions work best, such as in detention centres, homelessness and family services.

Given the number of trials, can we expect them to become more common?

Possibly. The current challenge is whether the early bonds will achieve what was expected; that is, positive outcomes for the community and investors getting an above-market return. Success here will obviously spark future interest.

The big challenge is that there is so much infrastructure required to allow the impact investment market to thrive: reliable data sets, social banking processes, effective and standardised evaluation frameworks, trusted advisory services and intermediaries, and government expertise in the process. And that's just for starters. Until these things exist, the market is not likely to develop rapidly. And at this stage, the market is still young.

So is this just an expensive way of getting things done that could be done using traditional funding methods?

In some ways, yes. With sufficient funding, community-sector organisations could just get on and develop the interventions that are highly likely to succeed. Fully funded organisations would implement their own evaluation frameworks, because they are as interested as any other agent in knowing what works, and what the key influencers are. The reality is that not-for-profits don't have access to this scale of resources.

Social impact bonds have the potential to bring in much greater resources than governments or not-for-profits would normally be able to access.

So there are some big barriers to adoption?

I believe there are three big downsides that inhibit the widespread adoption of social impact bonds.

Firstly, they're expensive and resource-intensive to establish.

Secondly, they require data sets that most sections of the not-for-profit sector can only dream about. To think we could have proper evaluation frameworks - with appropriate data - for the mainstream part of the sector is ambitious.

Finally, the system depends on the presence of a control group in order to prove the efficacy of the intervention; you need an unaffected population to compare to. In the real world this is hard to find. (Ed: For more about counterfactuals, read our report on the benefits of randomised control trials featuring the Behavioural Insights Team's Rory Gallagher.) It also means that you can't have more than a few social impact bonds operating in any one area of the sector at once - if you're running multiple interventions at once you can't be sure which of those is contributing (or contributing most) to your impact goals.

So how can investors know that the proof being asked for - to measure the interventions and therefore pay out - really is up to scratch?

There are checks and balances in the system. First, the metrics are agreed to by the government and the participating not-for-profit before it is taken to the market.

This means that the not-for-profit must opt in to the social impact bond. We have to trust that the not-for-profit believes in the intervention, accepts that the metrics are appropriate, and is capable of efficiently managing the complexity. Otherwise, they wouldn't - or shouldn't - progress with a proposal.

Secondly, it is usual practice for the not-for-profit to also be an investor, which further aligns their involvement with the market.