Bypassing governments

It's unfortunate that most of us think the default agency for solving society-wide problems is government. In the west, we are lucky to have (for the most part) well-meaning people behaving ethically and in society's best interests. Even then, they often get it wrong, nor do they have powerful incentives to be efficient. But most people live under much worse governments, and bypassing them is not only more efficient: it's the only way things will ever get done.

Malaria is a global, devastating disease that adversely impacts both the health and economic productivity of numerous countries, especially those in sub-Saharan Africa.....However, [owing] to various interventions, such as insecticide-treated bed nets, residual indoor insecticide spraying, and effective medications, the toll of malaria deaths has somewhat declined in the last decade. This turnaround is in part due [to] initiatives undertaken in the private sector, since governmental programs in that region are fraught with disorganization and a lack of a public health infrastructure. Private sector helps in fight against malaria, American Council on Science and Health, 22 June

Just because we look to governments to improve health outcomes doesn't mean they will be good at doing so. In many countries, tragically, they will stand in the way. The backers of Social Policy Bonds that target the health of people in developing countries wouldn't need to form a judgement as to whether governments are likely to be helpful or otherwise: the way the market for the bonds works would ensure that only the most efficient operators, whomever they may be, will be rewarded for doing so. If bypassing inefficient or corrupt governments is the best way of improving the health, then that is what bondholders will do.

What do we want?

The Economist, in an article about planetary boundaries, discusses climate change. Do we want to see a reduction in greenhouse gas emissions, or should we focus on “radiative forcing” - the increase in energy delivered to the surface of the Earth over time, largely as a consequence of extra greenhouse gases?

Greenhouse gases are, however, only a problem because of their effect on radiative forcing. If that could be reined back inside [1 watt per square metre above pre-industrial levels] by other means, then the CO₂ limit [of 350 ppm] would no longer pertain. And that might be possible by spraying reflective particles into the upper atmosphere, to bounce sunlight back into space. Such a radical scheme would have all sorts of disturbing side effects, with political ones quite possibly outweighing environmental ones. It is by no means clearly the right thing to do. But it might be. Boundary conditions (subscription), 'the Economist', 16 June

What does 'right thing to do' mean in this context? What do want to achieve? The current policymaking environment allows us to avoid or muddle answers to this question. The result is ad hoc policy, decided on the fly, or not decided at all: just the aggregated effect of private- and public-sector interest groups working towards their own ends.

We need to do better than this. A Climate Stability Bond regime, at the outset, would compel us to clarify what we want to achieve. My suggestion is that we target simultaneously an array of physical, social, biological and financial variables. All targeted variables would have to fall into a specified range, for a sustained period, before the bonds would be redeemed. The variables could include the obvious ones, such as temperature, rate of change of temperature, sea level etc, but also such indicators as numbers of people killed or made homeless by adverse climatic events, species migration, species extinction, ocean acidity, insurance costs and insurance payouts. No longer would we focus (or pretend to focus) on one particular variable ; a bond regime would be versatile enough to target any number of quantifiable variables at the same time. Clarity over exactly what we want to achieve is essential if we are serious about actually achieving it. It's a tragedy that in this, and other, policymaking areas, we end up adopting de facto targets and subordinate all our policymaking to them. And that applies whether it's greenhouse gas levels - or gross domestic product per capita.

Incentives to offer incentives

Social Policy Bonds inject incentives into every stage of every process needed to achieve a social goal: the bonds are tradeable, which means they are worth most to those who believe they can do the best job of achieving social goals most efficiently. If someone else thinks they can be more cost-effective, they will bid more for the bonds, and buy them from less efficient investors. The incentives under a bond regime cascade down from the bondholders to everybody whom they contract to help achieve the targeted goal. To put it simply: the bondholders have incentives to offer incentives to everybody who works for them.

This is the opposite of the current system where, if we have a large organisation within which some people experiment and find an improved approach, the system itself doesn't supply sufficient incentive to propagate that approach. So, the current Economist, discussing a successful innovation shown to work in one of the many hospitals run by UK's National Health Service:

[T]he main reason innovations do not spread is that the NHS has no mechanism for ensuring they do, or for rewarding the inventive. The service is centrally funded and emphasises the universality of its care rather than its results. Such a system is likely to prove better at controlling costs than at encouraging good ideas to thrive. From petrol to prescriptions (subscription) the 'Economist', dated 16 June

Why make the bonds tradable?

A critical feature of Social Policy Bonds is that they are tradeable. In this they differ from Social Impact Bonds. There are several advantages in the bonds' being tradeable:

1. Tradeability encourages the targeting of broad social goals that might require a long time to be achieved. How so? People would buy the bonds only if they expect to make a profit on them. If they're not tradeable, that means they'd have to hold them to redemption to make a profit. That in turn means that would-be investors would want any targeted goal to have a realistic chance of being achieved within their time horizon, which might be quite short, and certainly well within a single lifetime. And that means no targeting of things like ending global conflict, or slashing crime rates, or making meaningful inroads into climate change etc. Of course, incremental targets (cutting crime by 2 percent) might be targeted by such bonds, but that's messy and that sort of reduction could be achieved by chance. Most likely, if the bonds weren't tradeable, you'd get very narrow objectives, such as those being targeted by Social Impact Bonds (in the UK), like reducing the recidivism rates of people leaving one of the UK's prisons. That's fine and laudable, but it is limited. And because it's so limited, with limited opportunities for shifting resources to and from different projects, different approaches, and varying them as circumstances change, so too are the advantages of a bond regime, which are most marked when we do not specify how certain goals shall be achieved, but leave that up to bondholders to explore and investigate all the options. If we think of it from the bondholder's point of view: we want payback within, say, five years. We're not going to undertake much research or try any different approaches unless they have a very short lead time and are almost guaranteed to work. There is little scope for experimentation and trying totally new approaches, or for long-term research in that sort of policy environment.


2. When we have such small objectives, the costs of monitoring progress toward or away from their achievement is going to be a higher proportion of the total administrative costs than they would under a regime that could target broader goals. It's almost as easy (or not much more difficult) to monitor national crime indicators, say, as to look at the behaviour of group of a few hundred specific ex-prisoners in one part of the country over several years.


3. Another crucial reason why the bonds should be tradeable is because the people best placed to achieve a targeted objective will change over time. Most social and environmental goals will require multiple steps before they are reached. The people who are best at step one will not necessarily be those who are best at step two and all subsequent steps. We cannot even specify in advance what step one, or indeed any step, will entail; still less can we identify those best placed to take these steps. If the bonds are tradeable, that's fine; the market for Social Policy Bonds will ensure that the bonds are always in the hands of the most efficient operators. If the bonds are not tradeable, then we have something not sufficiently dissimilar from the way social policy is currently implemented: government identifies some organisation, often one of its myriad own agencies, and pumps money into it. If this agency is paid for performance (as in Social Impact Bonds), it has an incentive to perform well - which is an improvement on the way things are usually done. But if, as so often, one or all of the steps necessary to resolve the targeted problem optimally lie beyond the imagination or competence of that agency, then we are going to be stuck with current (woeful) levels of under-achievement in social and environmental policy. Social Policy Bonds have the advantage in that they not only do not stipulate how society's goals are to be achieved, but they also leave the selection of those who shall achieve these goals to the market, rather than high-up people in government or elsewhere. These people might be competent in their field, hard working and well intentioned, but nevertheless incapable of or unwilling to contract the best people for the job. In contrast, the market for Social Policy Bonds would favour the most cost-effective operators at every stage on the way to achieving social goals. 


4. If the bonds weren't tradeable, no would-be investors would monitor progress toward the achievement of the targeted goal. All the monitoring would have to be done by the backers of the bonds. There would be no market for non-tradeable bonds, so there would be no bond price nor changes in bond price to indicate how close the objective is to being achieved. That backers of the bond would have to be more diligent about monitoring than under a tradeable regime, where the market price of the bonds would do much of the monitoring work on behalf of the backers until the objective were close to being achieved. My book goes into lengthy detail about the information conveyed by market prices for Social Policy Bonds and their fluctuations, all of which would be lost if the bonds weren't tradeable. Such information would be extremely valuable to policy makers, and to society as a whole.