Social Impact Bonds: The best of all worlds?

The 'innovative finance' model of Social Impact Bonds (SIBs) is being hailed as an answer to funding critical social problems. In this post, guest contributor Helen Dickinson (@DrHDickinson) looks at the hype and potential of SIBs and if they really are the best of all worlds.


In recent months there has been a lot of talk about Social Impact Bonds (SIBs) – or Social Benefit Bonds as they have been known in New South Wales. Ostensibly buoyed by evidence of these having generated significant impacts in a range of tricky policy areas in their applications in the UK and US, Australian governments are currently clamouring to put these mechanisms in place and commentators are calling for the application of SIBs in a variety of contexts. Yet, in the rush to apply the latest management fad there is a serious risk that at best this results in a series of unintended consequences or is applied inappropriately, but at worst it may prove to be a flawed concept based on limited evidence.

What are SIBs?

For those not familiar with this idea it sits firmly within the ‘investment’ approach to social welfare that seems to be increasingly popular at present. The concept originally emerged from a post-Global Financial Crisis Britain, which saw significant cuts to spending on public services. In one sense SIBs are what you get when you cross Payment by Results mechanisms with public-private partnerships (PPPs). An investor (a bank or investment fund typically) provides financing to be used to tackle an identified social challenge. An intermediary organisation facilitates the relationship between the investor, the government and the service provider (typically a community organisation) and agrees the outcomes that should be achieved for a particular population for the project to be considered a success. The investor receives a return on their money depending on the outcomes achieved through the project. So, for example, if there is a 7.5% reduction in re-offending rates of when measured against a control group an investor may see a return of 2.5%. If re-offending rates are lower then investors can receive greater returns up to a maximum of 13.3% and may lose their capital if the reduction rates are lower than 7.5%.

The important point here is that money is only paid out where specified social outcomes have been met – which makes them less like a traditional bond product and more like an equity product in practice. Where governments pay additional returns based on higher outcomes the idea is that they pay for these from savings that will accrue from reduction of future liabilities (i.e. if you stop someone from reoffending then you don’t have to pay for court costs, jail time, etc.).

(for those who want to read in more detail about these processes, the Brookings Institute have published a detailed report on impact bonds and this sets out some really helpful detail) on roles, different types etc.).

Lack of evidence

From their initial launch to March of 2015, the Brookings Institute estimates that 44 SIBs have been used in areas such as preschool education, reduce prison recidivism, avoid foster care placement and increase youth employment across Europe, North America and Australia. This report notes that few deals have completed at least one set of payments to investors, which means that there is still a relatively small evidence base about the efficacy of SIBs. Yet this has not tempered enthusiasm for these in government, private sector or community organisations. Neither has the fact that a criminal justice SIB based in New York City failed to meet its three year targets meaning losses for investors of $1.2 million for Goldman Sachs and the loan guarantor, Bloomberg Philanthropies, of $6 million.

A lack of evidence for these mechanisms isn’t problematic on its own: after all there many things we do in public services that don’t meet the evidence threshold. Personally speaking I feel morally uneasy at the idea that investors are putting money into areas of significant disadvantage and misery in the hope of financial return. As Kyle McKay argues, ‘anytime Goldman Sachs’ is speculating on investments linked to broader societal outcomes, we should probably take notice’. Large financial institutions are providing upfront capital to projects, which they are presumably pretty certain of making a return on. Post global financial crisis we have all heard about restrictions on lending and so this seems like a strange move, unless there is some significant potential profit. While this is perhaps an unfair comment on organisations who are genuinely committed to corporate social responsibility (and it is a personal opinion), it may be naïve to think that these organisations aren’t assured of seeing a return on their investment. After all, as we will see below, it would be far cheaper to invest philanthropically than set up a SIB.

Who profits?

The idea of organisations making profit from welfare delivery is of course not a new thing. Community organisations have for years made profits on government contracts and often redirected this into other activities that create some form of public value. Or that’s the theory anyway. The reality in many areas has been that governments have underfunded social service delivery contracts knowing that community organisations will find a way to make ends meet somehow. Part of the challenge for many community organisations in costing out services at present, say in relation to the roll out of the National Disability Insurance Scheme, has been because they have not really ever assessed what the true costs of these services are.

So profit making in itself is not new, although it takes on a different dimension with these sorts of products. If we have learned anything from the last 30 years of outsourcing and increasing delivery of services by third party providers it is that this isn’t free. Cheaper and better quality it may be (and history has shown this certainly isn’t guaranteed with outsourcing) but this comes at a price.

SIBs are notoriously difficult to set up and take a huge amount of legal and financial input. The Brookings Institute reports that for one SIB, 27 contracts were written and more than 1,100 legal hours were billed. SIBs have the potential to bring a whole range of transactional costs. The history of PPPs and contracting for services more generally demonstrates that writing and overseeing contracts isn’t something that governments have always excelled at. If we are to increase the number of SIBs we may need to see a transformation in the workforce capacity of governments.

The trouble with outcomes

One of the most important parts of SIBs is setting out the outcomes metrics and associated payments. If you thought this was difficult with traditional contracting then it gets a whole lot more tricky with SIBs.

Outcomes are notoriously difficult to set out and to measure. Many SIBs have had to develop new outcome measures as they simply don’t exist for what they are trying to achieve. SIB contracts aren’t typically long in the context of making significant impacts on social outcomes timescales. The Brookings Institute found the shortest contract was 20 months and the longest 120 months with an average of 3 years and less. What this means is that it will often be difficult to measure outcomes in a true sense and so proxy indicators will be used. McHugh and colleagues note that this risks ‘encouraging an emphasis on a simplistic ‘mechanical’ model of cause and effect, resting on the notion that an intervention is a singular ‘thing’ or event which results in a clearly discernible outcome. This fails to grasp the complexity of the conditions and contexts of the social problems that SIBs are aimed at addressing’.

In previous work I have argued that complex social problems can’t simply be ‘managed’ and that if we try to focus just on the mechanical cause and effect processes then we miss the actual ingredients that make these processes actually work. In trying to isolate these mechanisms there is a potential risk that SIBs perpetuate this and make us less able to deal with real social issues.

Unintended consequences

By moving service delivery away from governments even further than occurs with current contracting or outsourcing arrangements there is a very real danger of informational asymmetries and also of reducing democratic accountability. SIBs remove a number of functions from government to the extent that it can be difficult to play an effective stewardship role, particularly in the event of a crisis or failure in delivery. Public services are typically founded on a collectivist notion, which SIBs and other recent moves around individualisation pose significant challenges to.

Again, I’m not anti these things as a simple rule, but I do think when they have the potential to pose such challenges then they are worth some careful consideration. SIBs are not the only potential mechanisms that might deliver these sorts of changes and it may be worth investing in some other approaches in the hope of delivering these outcomes. Without considering some of these potential impacts then we won’t have a full sense of impact and there is the danger that parties sign up to things that don’t necessarily fit with their aims and ethos.

Best of all worlds?

The rhetoric around SIBs is that it provides the best of all worlds in the sense that it provides private sector investment and expertise alongside community organisation insight into innovative service delivery and dealing with complex social challenges and delivering results for the public sector at minimal (financial) risk. Yet a key challenge remains in terms of whether SIBs actually solve social problems. It does not resolve an issue that is present in current contracting regimes – that of meeting programme targets but not dealing with the fundamental issue. If we can’t see the wood for the trees then it may undermine the overall intention of the investment approach.

If we were really serious about the reduction of inequalities then we would need to give thought to their structural determinants and it is likely that SIBs would not be the obvious solution. Australia doesn’t currently face the same levels of cuts to public service spending as other jurisdictions and it may therefore be worth thinking carefully about SIBs and their appropriateness to specific contexts before jumping to this solution too quickly. SIBs have their place in the rich tapestry of modes and means of welfare delivery but they aren’t the answer to all the dilemmas governments currently face.

Posted by @corr_lara