Social Impact Bonds: an explainer

Are SIBs worth another try? What are the pros and cons?

Social investment minister Nicola Willis recently announced the creation of a standalone Social Investment Agency, which will manage a fund and have an advisory board. Its aim will be to “commission outcomes for vulnerable New Zealanders, and to work with community, non-government-organisations (NGOs) and iwi providers”. The minister envisages that this new organisation will “grow in partnership with other funders”. These partnerships and investments should see the return of social impact bonds (SIBs).

What are SIBs?

SIBs are similar to green bonds but aim to finance specified social services and positive outcomes. Adapting a passage from an OECD report (2016, p. 4):

“An investor provides funding for an intervention, which is used as working capital for a service provider that is responsible for the social services delivery, the attainment of agreed outcomes and the provision of data related to them. The payment to the investor coupled with agreed interest shall be released by the government or the funding agency. Therefore, the government (or its funding agency) is the ultimate payer and determines the outcome metrics and payment terms. An evaluator may be used in some SIBs assessing the agreed outcomes and their impact. The beneficiaries from a SIB’s intervention are the population in need and recipients of the intervention.”

These services might aim at getting young people into employment, getting homeless people into accommodation, or rehabilitating people leaving prisons, for example.

If the intervention meets the agreed targets, then the private investor gets a positive return when they cash in their bonds. The investor loses if the SIB fails to meet outcomes, but this could be mitigated in the contract, or even partially insured through another party. So the contract is not necessarily a bond in the sense of a fixed-income security, but may be more like a futures contract, only much more complicated.

A lot of third-party service provision goes on in public services (for example, physiotherapy paid for by ACC) but SIBs are funded up-front by private investors. This injects new capital into social services without raising public debt. A deal may be struck when the government can estimate that its potential long-term savings outweigh the cost of repaying the investor in full at the end of the term. In theory, a secondary market could emerge in which investors freely trade the bonds.

Why is this a good idea?

The government transfers some of the risk of providing services for vulnerable people to investors and gets social outcomes met at lower cost, or even with better quality. The social services provider believes that they can get better outcomes for vulnerable people, more efficiently and effectively, and with more innovative approaches, than government departments can, and be accountable for their performance, and hence benefit from access to a steady funding stream. The investors get more closely involved in producing positive social outcomes than they would if they just donated to charities, and they contribute new expertise. Investors can “make a difference” and “make some money” at the same time. The state’s clients experience improvements in their lives without having to deal with impersonal bureaucracies.

In theory, an iwi or community could crowd-source an investment fund and then partner with the government’s social investment fund to achieve better results for the people they most care about. So the investor doesn’t have to be a rich person seeking to profit from services to the poor. It could be a good way for iwi corporations to develop by-Māori-for-Māori services.

What are the fishhooks?

There are many concerns, and the contracts get complex. First, there needs to be a worthwhile service that the government and others want done. And the outcomes need to be clearly definable and measurable. Investors should be wary of exaggerated claims about the likely social impact or outcomes.

An SIB involves a performance-based contract with an investor who puts up the capital to make it happen by means of a bond issue. Because the risks and rewards are closely tied to achievement of outcomes, the contracts need to have very clear terms and conditions, to spell out who pays whom for what, and when, and to cover contingencies if things go wrong or if unexpected events intervene. Large sums would be at stake, so the contingencies need to be understood in advance and work for the benefit of all concerned. Investors would want clauses that allow for exit or early bond sales without suffering complete loss.

Furthermore, there needs to be monitoring all along the way, as the government can’t entrust vulnerable clients to a third party without knowing that they’re safe, and investors want to see the results. If returns are contingent on performance, then all parties need to have confidence in reported data.

Easy? Not really.

Successful outcomes in public services for vulnerable people are inherently hard to specify, and it’s difficult to set clear targets and measure performance. Getting and keeping a paid job is a fairly clear goal – but, for how many hours per week, and for how many weeks, before it’s classed as a success? How is “appropriate” employment to be tested, and what if the job just retraumatises a mentally unwell person? Cases are complex, with often more than one service-provider involved, and success can be hampered by unpredictable set-backs like mass lay-offs. Putting a dollar value on a tranche of services for a cohort of clients is inevitably a “guess-timate”. An investor has to be confident that the agent can get vulnerable people into work or keep them out of the courts within, on average, a cost-bracket that will deliver positive returns.

Contracts will have many “ifs” and “buts” built in, the negotiations will be difficult and time-consuming, and there’s plenty of scope for disputes. The wellbeing of clients may get lost in litigation between principal, agent and investor.

The investor will cherry-pick providers, seeking the easier cases and leaving high-risk people to the government. Investors will choose projects with high impact that look good on the corporate citizenship page of the annual report. But their overall impact on society will be relatively small. The difficult jobs will remain with the state. If a new SIB-funded service proves to be ineffective, taxpayers’ money wasn’t wasted – but the people may still be in need of state assistance.

An unscrupulous agent under pressure to deliver results may adopt “innovative”, but expedient, methods that offend public opinion. All parties, including the government of the day, run the risk of reputational damage from any negative news about mistreatment of vulnerable people.

A change in government may lead to a change in policy goals and a loss of commitment to prior agreements. Parliament always has the power to cancel a contract – as happened to work-injury insurance contracts in 2000. And reporters may make OIA requests to expose what investors have been profiting from.

On the other hand, the government may be left to pick up the pieces, and the costs, if an agent is under-performing and investors dump their bonds.

Civil servants, NGOs and private investors talk different languages and have differing values and interests, but need to sit around a table and hammer out a complex SIB agreement for the performance of services. This is harder than getting a loan from your bank or investing in shares.

In other words, the transaction costs are high, and these bonds are still experimental.

What went wrong last time?

An SIB procurement project was launched by the Key government in late 2013 as part of its Better Public Services programme. In 2015, the New Zealand Initiative published a paper that made a cautious case in favour of SIBs. An Australian firm that was engaged in “an experiment that paid contractors up to $12,000 to help a sole parent or a person with mental health issues into paid work” (as reported in NZ Herald) also got into negotiations over a pilot SIB, according to a Treasury paper.

Another provider whose SIB negotiations fell through in July 2016 complained of “tedious bureaucratic slow processes that kill things before they’re even born”. Aside from the Treasury, departments in the social and health sectors had to be involved. On the government’s side of the table, there were, then, several entities in the mix, not just one. The public servants lacked commercial experience, and negotiations were protracted.

After four years, they didn’t deliver much. ANZ Bank wrote proudly about the Genesis Youth fund in 2019. But individual success stories on a coporate webpage don’t amount to a controlled evaluation.

So why not try again?

The delivery gurus, Luxon & Co, are preparing new ground for partnerships with investors to fund social services. Mental health, disability and offender rehab are likely targets. The main aim is to get people off benefits and into productive work.

As this develops, readers will no doubt see anecdotes from both sides of the aisle about SIBs that reportedly did or didn’t work in places like England or New South Wales. Duelling anecdotes won’t prove anything, however. Given that the taxpayer will ultimately be liable for providing services if an SIB fails, the public deserves to see robust and transparent cases for them, with cost-benefit analysis and case-controlled evaluations.

Many people may simply find it repugnant, however, to bring the profit motive into social services for our most vulnerable. The “neoliberal agenda” comment is likely to follow.

NZ is a unique environment with its cultural mix of vulnerable communities, and, on the other side, possibly a shortage of private investors and philanthropists. If I may speculate for a moment, the reserve funds managed for ACC, NZ Super and even KiwiSaver could be called upon to buy these bonds.

For the sake of any future posts on this topic, I’m keeping an open mind for now.

---

You can read more on the Goldman Sachs website. It seems to be a growing market.

The NZ Initiative’s 2015 report is online.

This image is AI-generated by Gencraft: