Not so outlandish

I'm aware that Social Policy Bonds can seem outlandish - at least at first reading. They imply the outsourcing of the achievement of our social goals to anybody, public- or private-sector, with no directive as to how they shall achieve our goals, except that they comply with legislation. That said, many social and environmental problems have resisted all attempts at solution. You are immediately dismissed as an idealist for suggesting, for instance, that war can be abolished, or that poverty or crime can be drastically reduced. In response, I might say Could a Social Policy Bond regime do worse than our current efforts? It's true that a bond regime could be completely ineffectual: people could simply refuse to buy the bonds on flotation, or they could buy them, then do nothing to help solve the targeted problem. If nobody buys the bonds, then that tells the issuers that they have allocated insufficient funds for redemption. If bondholders do nothing, then the value of their bonds would fall until either they have a powerful incentive to do something, or to sell their bonds to people who would do something. In all such cases, however, there would be no cost to taxpayers, even if the bonds had been issued by government, unless the targeted social problem is solved.

As important is that Social Policy Bonds can be issued in parallel with existing efforts to solve social problems. Bondholders could undertake their own initiatives, but would also have incentives to look at current approaches and boost the finances of the more promising ones.

Below is an excerpt from my book showing how this transition could work when addressing UK health problems. The excerpt is from Chapter 4. That chapter, and the complete book can be downloaded free of charge from here. Simpler, shorter explanations of the Social Policy Bond concept can be found via my homepage.

"Take health, for example. In the UK, central government provides funding for regional health authorities (for spending on doctors, hospitals and prescriptions) according mainly to population level, age and need. Government also supplies funds directly to medical research organizations and academic institutions. A transition to a Social Policy Bond-based, rather than institution- or activity- based, funding programme would see the direct funding government gradually decline, while expenditure allocated by bondholders to the outcomes that all these institutions are collectively trying to achieve — longer life spans and a better quality of life, say — would gradually rise.
On introducing such a bond regime a government could decide to reduce its funding of health authorities and research institutes by 1 percent a year, in real terms. (The government could allocate the saved funding to the future redemption of the health bonds it has issued.) So after five years, each health authority would be receiving directly from central government only 95 percent of the funding that it formerly received. But bondholders could choose to supplement the income of some of these health bodies. They may judge a particular group of health authorities to be especially effective at converting the funds they receive into measurable health benefits, as defined by their bonds’ redemption terms. Particularly effective health authorities might be working in deprived areas, where small outlays typically bring about larger improvements in health. Or bondholders might judge a particular research body to be
worthy of additional funding, because it was conducting excellent research into a condition that would be likely to respond especially effectively, in terms of health outcomes, to additional expenditure. In such cases, bondholders would supplement their selected health authorities’ or research institutes’ funding. It may well be that these favoured bodies end up receiving a large boost in income throughout the lifetime of a bond regime.

It could also happen that investors in bonds targeting health look at completely new ways of achieving health objectives; ways that currently receive no, or very little, funding. To give a plausible example, they may be convinced that one of the best ways of achieving society’s longevity objectives is to deter teenage drinkers from driving. Following this logic, they may find that one of the most efficient ways of doing so would be to lay on subsidised taxis for teenagers attending parties on Friday and Saturday nights – but only in certain parts of the country. It is difficult to imagine how our current centralised government fund allocation mechanisms could go about implementing such a programme. A Social Policy Bond regime would quickly eliminate some of the less rational distortions in other health care matters, amongst them the British National Health Service’s terminal-care budget, 95 percent of which was allocated to the 25 percent of the UK’s population who die from cancer, and just 5
percent to the 75 percent who die from all other causes. 

It is also likely that holders of bonds targeting health outcomes would greatly expand funding in areas such as health education or preventive medicine that rely on expertise outside those bodies traditionally devoted to health care. 

Could bonds targeting remote objectives, such as a large rise in longevity, or a halving of the crime rate, be compatible with a gradual transition of the type described above, where funding to existing health institutions reduces by 1 percent annually? At first sight there would seem to be an apparent mismatch between such incremental reductions in government spending and the time scale needed to reach long-range objectives. The critical point here is that bondholders would be investing not on the basis of the annual reductions in government expenditure on existing health institutions, but on the basis of the redemption value of all the bonds issued. To be more precise, it would be this total redemption value, minus the bonds’ existing market value, that would inform bondholders’ investment decisions. This sum could be many times each year’s incremental reduction in government’s institution-based spending. One of the virtues of a Social Policy Bond regime is that bondholders could expect capital gains in the short run from investments that will begin to impact on the targeted goal only in the long run. By doing the initial groundwork efficiently and speedily – not usually a very lucrative proposition in the current regime – they could see short term rises in the bond price
and early capital appreciation. The accumulated reductions in spending to existing institutions would be one, but not the only, factor influencing how much government decides to spend on achieving a specified social goal. Also important would be the financial savings (if any) that achieving the objective would bring about, and the value society would place on any nonfinancial benefits. Similarly gradual transitions would be warranted in other areas, such as education and crime, where schools and police forces, some of which are bound to be much more effective than others, are well entrenched. These institutions would receive slowly diminishing absolute levels of funding directly from government, while bondholders would again focus their spending on especially rewarding, in terms of specified education and crime outcomes, projects and institutions. As with health, it is likely that those areas that are initially most disadvantaged would again provide bondholders with the greatest return per unit outlay. In newer policy areas, particularly the environment, it may be possible to expand spending allocated via the bonds at a faster rate: expertise in the environment is still relatively mobile, and it would be easier to quickly establish new outcome-based institutions or to reorientate existing ones."

Give learned associations a chance!

Justin Gregg compares diagnostic inference, whereby animals learn to associate a cause with an effect, to causal understanding, whereby humans reason as to why something happens:

If causal understanding is such an obvious advantage over other ways of thinking, why did it take our species 200,000 years before we began using this ability to begin the spread of modern civilization? The answer is that sometimes, being a why specialist leads our species toward unexpected ludicrousness that is so bad for our species (evolutionarily speaking) that it makes you wonder if we’d actually be better off relying solely on learned associations. Justin Gregg, If Nietzsche Were a Narwhal, September 2022

I will agree that, when policymakers attempt to understand why something is happening by looking for 'root causes', this can be a waste of time or an excuse for inaction. There are too many causes of social problems, and they are too various and too variable to be identified - at least by conventional public- or private-sector organisations, with their fixed structures and limited remits. Crime, poverty, war: these have many causes that vary markedly over time and space. It is politically acceptable for governments either address the symptoms of these problems, or to spuriously reduce the number of causes to one and target that: so crime is often attributed to poverty, and climate change to greenhouse gas emissions. But these problems may have no single, fixed cause, and we need to offer incentives that encourage people continuously to explore the many, varying causes of our social and environmental problems. Often, there are suggestive associations between cause and effect, but insufficient evidence for governments to use taxpayer funds to act on those associations. So, unless effect can unambiguously attributed to causes upon which we can act, our problems remain.

What I suggest we do instead is reward the outcomes we want to achieve, regardless of how they are achieved. Provide incentives for people to investigate some or all possible causes or simply to learn by association - as in the animal kingdom - and react accordingly, even without the evidence required to draft a policy. And to bear in mind that doing nothing, in some rare circumstances, might be the optimal approach.

Social Policy Bonds would encourage this type of behaviour, rooted in a humility that recognises that we cannot always identify the causes of our problems, but that we might not need to in order to solve them. How would they do so? They would reward all approaches, however indirect or nebulous, that help solve a targeted problem. Evidence in favour of an approach could be anecdotal, associative or putatively causal: governments cannot act on all such evidence, but holders of Social Policy Bonds can; their sole criterion for backing an approach is whether it promises to be a cost effective way of achieving the goal they target. 

What do I mean when I say that doing nothing might be the optimal approach? Take climate change: the evidence that it's actually happening convinces me (currently). The evidence that we can or need to do something about it is a little more contentious. We need a framework in which, if circumstances change, incentives are in place for an appropriate, nimble response. Such circumstances might include, say, a dramatic increase in adverse climatic events (more rapid response required), or in our scientific knowledge, or a supervolcano that threatens to freeze the planet, or an unforseeable (and, admittedly unlikely) reduction in adverse climatic events. In theses latter two events a bond regime would see climate mitigation activities could be attenuated or suspended. The conventional approach would find it difficult to adapt appropriately. 

My point, ultimately, is a simple one: if we target outcomes, we do not need to identify root causes to solve our complex social and environmental problems.

Where are the incentives to take the long-term view?

Where are the incentives to take a long-term view? I mean not only for policymakers, but for corporations, philanthropists, and the rest of us. Rewards are there for turning up to work, for attracting attention, for doing things; even sometimes for solving problems, though rarely for preventing them arising in the first place. But, increasingly, we are more concerned with short-term goals. There are several possible explanations. In the private sector, industry concentration means that managers, rather than family members, control corporations. But what about the public sector? 

[Government bureaucracies non-self-evaluate. At a minimum, agencies with evaluative responsibilities are not invited to evaluate - they are kept out of the loop, their opinions unsought. At a maximum, government agencies actively suppress their own internal evaluative units and are discouraged from evaluating the beliefs and policies of other agencies. Why States Believe Foolish Ideas, Steven van Evera, 2004

There are, then, few rewards for being successful, and fewer still penalties for failure. As well, government's role is changing. The current Economist writes about its growing role in bailing out corporations:

Politicians have long sought to provide safety nets or stimulus in bad times. But over the past 15 years, they have become far more willing to shore up vast swathes of the economy. When industries, companies or people get into trouble, fiscal help is never far away. Gains are privatised, but a growing share of losses or even potential losses are socialised. To appreciate this role for the state, discard much of the conventional wisdom, which says that in the “neoliberal” era governments have let free markets run riot. Instead, this is an era of “bail-outs for everyone”. The world enters a new era: Bail-outs for everyone!, Economist, 1 October

The long-term view is the inevitable loser.

Social Policy Bonds aim to inject the market's incentives into the achievement of our social goals. But perhaps their greater benefit is that they enable the targetting of goals that can be achieved only after decades, such as reduced crime rates, a cleaner environment or world peace. Nobody knows how to achieve these goals, and existing organisations have little incentive to investigate the wide range of diverse, adaptive approaches necessary to do so. But a Social Policy Bond regime would put in place incentives for people to take the long-term view: to research, investigate, experiment, refine and implement the most promising potential solutions to our social and environmental problems, even if, as is likely, total success will take decades.