Betting on outcomes

The current controversy over prediction markets reminds me of how the idea of Social Policy Bonds originated. If you hope for a particular outcome and put a sufficiently large bet on it, then at some point it would be in your interests to make that outcome happen. Applying that to desirable social and environmental goals, and of course restricting any goal-achieving activities to those that are legal, is the underlying principle of the bond concept. Perhaps one way forward for, say, environmental groups is collectively to bet on something verifiable, and something that a counterparty thinks unlikely, such as a less-than-consensus projection of the proportion of greenhouse gases in the atmosphere in five years. They could then strive to achieve that goal, using their own funds and the possibility of that payoff as collateral for projects that will cut greenhouse gas emissions. In principle it could work. In practice, since most of our important social and environmental goals are long term in nature, it could be helpful only if there were a secondary market for the bet. If there were, then as with Social Policy Bonds, people could trade in and out, without having to wait till the goal is achieved before they can realise a gain: they'd make a profit if the market sees quick attainment of the goal as more likely. 

Without a secondary market, prediction markets would be as limited, and for the same reason as Social Impact Bonds, which aren't tradeable. I've written about the need for tradeability here and here. Essentially, it greatly expands the time-frames of the goals we can target, and allows for research, experimentation and channeling resources into only the most promising projects. For our most challenging goals, we need these activities, and they can take years. 

A piece looking at how markets could target sustained nuclear peace. For more about Nuclear Peace Bonds see here. (It took the Bulletin of Atomic Scientist less than 24 hours to reject this piece intended for publication in their journal.)

Eighty-five seconds to catastrophe: can markets save us?

 Our political systems are failing us. They are incapable of thinking more than a few years ahead, at a time when we need them to have at least a decades-long vision. As well, the gap between politicians and the people they are supposed to represent, is widening. The result is that some of the most important determinants of our future are happening by default. The outstanding example is the drift toward nuclear conflict.

Governments are preoccupied by the short term. They make, or don’t make decisions that collectively have led us to the 85 seconds to midnight scenario of the Bulletin of Atomic Scientists. It’s safe to say that the vast majority of the planet’s 8 billion people don’t want to see a nuclear conflict. We need to find a way of connecting our wishes with sustained nuclear peace.

Outcomes and efficiency

Our best chance might be to bypass today’s self-interested, short-sighted policymakers and instead target nuclear peace ourselves, making sure that we reward the people who achieve it most efficiently, whoever they are and however they do so. This essay  envisages a Nuclear Peace Fund into which everyone can contribute including: non-governmental organisations, ordinary people, philanthropists and corporations. The Fund would be used to reward the successful achievement of sustained nuclear peace. In economic theory, and on all the evidence, markets are the best way of allocating society’s scarce resources to achieve targeted goals, so ideally, we’d inject market incentives into the achievement of our goal.

A single company?

In a free market economy people choose goods and services from a range supplied by individual companies, competing to meet their demand. Could that model be applied model to a universally desired good, that of nuclear peace? This company would have as its sole aim the sustained absence of nuclear conflict and it would be paid only at the end of a defined period – perhaps 30 years – of peace. The Fund would be kept in escrow, and would be paid out to the company only when that goal had been achieved. The company would issue shares, the value of which would rise and fall with progress toward or away from the targeted goal. Directors could be paid bonuses linked to the share price and hence to perceived progress toward the nuclear peace goal. Shareholders would be passive investors, and all decisions as to how to allocate funds for goal achievement would be made by the directors of the single company. At the end of the targeted period of nuclear peace, shareholders would be paid out from the funds in escrow and the company wound up. (Another company could be, or could have been, set up to ensure continuing nuclear peace into the future.)

The weakness is that the directors would be making all the decisions as to where to allocate funding for the various projects and their decisions would be too far removed from market incentives. As well, the identity of the directors themselves would not be immediately responsive to the dictates of the market. For a goal as vast and consequential as nuclear peace, we are going to need a huge array of projects, with a correspondingly huge need for constant oversight. We need diverse, adaptive projects, the monitoring of which would be beyond the scope of any single set of directors. So the creative destruction of failing or inefficient projects wouldn’t occur; neither, because of the necessarily large size of the company, would creative destruction operate to subject the company through the possibility of takeover. Of course, the company might fail completely– but that would do nothing to achieve the goal of nuclear peace.

On the plus side, the market prices of the shares, on flotation and beyond, would supply useful information as to how much progress is being made.

Social Impact Bonds

Social Impact Bonds (SIBs) (also known as pay-for-success bonds etc) are a contract to achieve a specified outcome with payouts linked to efficiency. They are currently deployed in about 30 countries and are a means of rewarding people who work to achieve a defined outcome.

A quick look at some of the goals SIBs target reveals their major limitation: investors must hold them from issue to redemption. So they are inherently short or medium term in nature, which is why they have been restricted to very narrow objectives, such as targeting local rates of redivision or homelessness. For such goals, monitoring costs tend to be high proportion of total outlay and, with their narrow scope, there are few opportunities to shift resources between different activities. Most bondholders will want payback within a decade or less. While there is the scope to innovate, new techniques will inevitably be drawn from an existing array of approaches, rather than a result of research or experimentation.

SIBs can allow sub‑contracting, collaboration, or replacement of providers over time, which can open space for additional or new providers. But there is limited scope to do this as, once the contract has been awarded, it is up to the lead bondholder (one body, or a consortium) to decide whether to subcontract and to whom. The identity of the lead bondholders is fixed at the outset and though they, like subcontractors, can bid for these contracts, neither would be subject to market discipline during the lifetime of the contract.

If we are going to reward the achievement of remote outcomes then, we need to reward people along the path to goal achievement, so that they can exit, with profits, before that point. It's important also not to take today's institutional structures, including that of existing public- or private-sector bodies, as a given. Our goals should shape our institutions; not the reverse.

Social Policy Bonds

This author suggests that Social Impact Bonds targeting nuclear peace be made tradeable. Tradeability would allow investors to profit from holding the bonds for periods shorter than the time between floating the bonds and redemption. Investors could buy the bonds, do what they can to raise the bond price, then sell their bonds to others who would take the next steps toward achieving nuclear peace. Ideally, they’d benefit from an increase in the value of their bonds. This means that we could realistically target the goal of nuclear peace sustained for 30 years.

Nuclear Peace Bonds be issued on the open market; investors would compete with eac other to buy them. The bonds’ redemption Fund, made up of contributions from private- and public-sector bodies, and ordinary people would be held in escrow. It could be swelled by additional contributions during the lifetime of the bonds. Each bond would entitle the holder to a share of the payout once the goal has been achieved. The bonds would be tradeable at any time until redemption. Bondholders would benefit either by being passive investors, hoping for progress toward achievement of the goal, or they would be motivated to become, or to sell their bonds to, active investors, who would work to increase the probability of early achievement of the goal, and so benefit from a rise in the value of their bonds. Some sort of tacit or explicit co-ordination between bondholders and, possibly, aggregation of the bonds into a few hands, could lead to the creation of a new sort of organisation; one of ever-changing composition and structure with the sole aim of achieving the targeted goal quickly. Anybody or any entity, public- or private-sector could hold the bonds. This body and all its decisions would be wholly devoted to minimising the risk of nuclear conflict.

Incentives would cascade down from bondholders to those working to achieve the goal; the linkage might be direct, in that bondholders themselves would be working for nuclear peace, or indirect, in that they’d subcontract achievement to other bodies, which could be rewarded according to their perceived success in their operation, or according to any change in the bond price.

It’s possible that, despite the efforts of bondholders and their subcontractors, there could be movement away from achievement of the goal, so that the operator’s bonds fall in value. From the operator’s view, that would be unfortunate, but from society’s view it’s positive, in that the bigger the gap between the bond’s price and its redemption value, the greater the incentive for other investors to buy the bonds and work toward achieving the goal. It’s possible that at the lower levels of the subcontracting pyramid, operators would be paid for their time, rather than their performance. The risk of underperformance would be borne by those higher up the pyramid, who stand to gain most when the goal is achieved.

The market for the bonds would reveal information about how likely people thought nuclear conflict were possible. If the likelihood of nuclear conflict is seen as quite high at the time of issue the bonds would fetch very little on the open market. The size of the Fund and the number of bonds issued would determine the initial price of the bonds; they should be pitched such that that price is high enough to allow for a fall in value that would motivate investors to be active, rather than passive, holders.

But what if nothing works, and the bonds fall to a very low level despite the best efforts of investors? That could happen if events outside any bondholder’s control occur: a rise in tension between two nuclear-armed powers, for example. The goal would remain unlikely; the funds would remain in escrow. Backers of the bonds could increase the motivation to achieve the goal in two possible ways:

·       Boost the funds in the escrow account and increasing the redemption value of each bond in circulation correspondingly; or

·       Boost the funds in the escrow account but issue more bonds with redemption terms identical to those currently in circulation.

Either would work, but the second option is better: if bondholders anticipated that there might be a swelling of the redemption funds, they might refrain from actively trying to achieve the goal, waiting instead for the larger payout.

Bondholders and operators can maximise their gains not only by holding bonds that appreciate in value, but by:

• Buying futures or call options in the bonds; or
• Borrowing to implement, or speed up the implementation, of their promising goal-achieving activities.

 Big bondholders would probably subcontract at least some of the work required to achieve the goal. They could do this in collusion (tacit or formal) with other large bondholders. Smaller operators could approach big bondholders with requests for funding; their rewards might be fully or partially conditional on a rising bond price. At all times, the market price of the bonds and movements in that price would provide information of great value to investors, would-be investors and policymakers.

Financial incentives aren’t everything and many working for peace are not primarily motivated by thoughts of personal gain. But even so, they’d be grateful for funds received from bondholders: a Nuclear Peace Bond regime would boost funds available to the most efficient of these organisations, which they could use to expand the scope of their activities as they see fit.

There would be many benefits arising from a Nuclear Peace Bond regime, but the most important is that it would create a coalition of interests that would oppose the tendencies that are, deliberately or not, leading us to a nuclear catastrophe. A successful Nuclear Peace Bond regime would also signal to ordinary people that they do not have to rely on politicians to bring about universally desired outcomes. The long-term effects of this could be momentous.

© Ronnie Horesh March 2026

 

Thinking about the long term

A draft piece on why Social Policy Bonds could address long-term problems 

Because our governments and their paymasters in big business have no sense of a long-term future, catastrophes such as environmental collapse or nuclear conflict are going to become more likely unless we find ways of:

• Giving people a chance to decide on our long-term goals because, I suspect, most us have a longer-term outlook than government, 
• Targeting those goals, and
• Rewarding the people who achieve these goals most efficiently, whoever they are and however they do so.

Efficiency is important. In economic theory, and on all the evidence, markets are the best way of allocating society’s scarce resources to achieve targeted goals. Ideally, we’d inject market incentives into the achievement of our goals. That’s difficult to imagine in today’s policymaking environment, dominated by lawyers, big business, lobbyists and others with their own interests to pursue – including resistance to any change. So it’s important not to take today’s institutional structures, including that of government, as a given. Our goals should shape our institutions; not the reverse.

One possibility would be to contract out the achievement of our goals to a corporation. It would be paid only when our goal had been achieved. This could work for narrow, short-term goals, where the costs of achieving the goal can be estimated. But our most important goals are long term, and for these goals, a single body would, at best, require a huge premium over its expected cost of achieving the goal. More likely, since we are talking about goals that could take decades to achieve, any single entity would not take on such a huge commitment.

(It’s a fact that people working for charitable organisations do work to achieve these long-term goals. But such efforts, though admirable, are small in scale and, because there are few financial incentives, they are likely to remain so.)

If we are going to reward outcomes then, we need:

•         to reward people along the path to goal achievement, so that they can exit, with profits, before that point; and
•         to find ways of sharing the risks inherent in long-term commitments between different operators.

A single company?

One option would be to set up a company solely to achieve this outcome and to sell shares in it. The motivation would be provided by a large sum, kept in escrow, that would be paid out to the company once the goal has been achieved. The shares would rise and fall with progress toward or away from the targeted goal. A disadvantage of this is that there would be no link between those who would profit from progress toward goal achievement and those working towards goal achievement. Shareholders would be passive investors, and all decisions as to how to allocate funds for goal achievement would be made by the directors of the single company. Directors would be promised a share of the payout or would own shares. The problem is that the directors would be making all the decisions as to where to allocate funding for the various projects. With such a huge overall goal, they would not have sufficient oversight to keep up with progress on all those projects; nor to determine which are the most promising, and which it would be more efficient to terminate. Again, because of the necessarily large size of the company, it is unlikely that it would be subject to the market discipline through the possibility of takeover. On the plus side, the market prices of the shares, on flotation and beyond, would supply useful information as to how much progress is being made.

We need to motivate all who are working to achieve the goal all the time. One big company can't do this; so we need a way that doesn't require a large, overseeing organisation.

Social Policy Bonds

The problem remains: how to inject market incentives into the achievement of society’s long-term goals. I propose that we begin, again, with a large sum in escrow that would be paid out when the goal has been achieved. I propose that, shares in a company being sold, bonds be issued on the open market. Each bond would entitle the holder to a share of the payout once the goal has been achieved. The bonds would be tradeable at any time until redemption. Bondholders would benefit either by being passive investors, hoping for progress toward achievement of the goal, or they would be motivated to become, or to sell their bonds to, active investors, who would work to increase the probability of early achievement of the goal, and so benefit from a rise in the value of their bonds. Some sort of tacit or explicit co-ordination between bondholders and, possibly, aggregation of the bonds into a few hands, would lead to the creation of a new sort of organisation; one of protean structure with the sole aim of achieving the targeted goal quickly. Anybody or any entity, public- or private-sector could hold the bonds.

One potential disadvantage of the bonds is that there will be no necessary or direct link between the progress that each operator (body working to achieve the goal) makes toward goal achievement and the consequent rise in the value of its bonds. Indeed, it’s possible that despite the efforts of an operator, there may be movement away from achievement of the goal, so that the operator’s bonds fall in value. From the operator’s view, that would be unfortunate, but from society’s view it’s positive, in that the bigger the gap between the bond’s price and its redemption value, the greater the incentive for other investors to buy the bonds and work toward achieving the goal. It’s possible that at the lower levels of the subcontracting pyramid, operators would be paid for their time, rather than their performance. The risk of underperformance would be borne by those higher up the pyramid, who stand to gain most when the goal is achieved.

But what if nothing works, and the bonds stay at a low level despite the best efforts of investors? That could happen if events outside any bondholder’s control occur: a plausible scenario, if we target such goals as reducing the impact of natural disasters. The goal would remain unachieved; the funds would remain in escrow. Backers of the bonds could increase the motivation to achieve the goal either: by boosting the funds in the escrow account and increasing the redemption value of each bond in circulation correspondingly; or by boosting the funds in the escrow account but issuing more bonds with redemption terms identical to those currently in circulation.

Either would work, but the second option is better: if bondholders anticipated that there might be a swelling of the redemption funds, they might refrain from actively trying to achieve the goal, waiting instead for the larger payout.

It’s likely that operators with a good idea will be able to maximise their gains not only by buying bonds, but by buying futures or call options in the bonds; or by borrowing from bondholders to implement, or speed up the implementation, of their promising goal-achieving activities.  

Big bondholders would probably subcontract at least some of the work required to achieve the goal. They could do this in collusion (tacit or formal) with other large bondholders. Smaller operators could approach big bondholders with requests for funding; their rewards might be fully or partially conditional on a rising bond price. At all times, the market price of the bonds and movements in that price would provide information of great value to investors, would-be investors and policymakers.

Financial incentives aren’t everything. And people working to achieve goals might not be motivated by thoughts of personal gain, but many already working for charities or NGOs and who are already working hard to achieve social and environmental goals would be grateful for funds received from bondholders that they could use for other purposes than their own financial renumeration. A Social Policy Bond regime should boost the funding the most efficient of these organisations, which they could use to expand the scope of their activities by employing more people, for example, or expanding their scope.

There would be many benefits from such an approach. Apart from the efficiency gains that would result from allowing non-public-sector bodies to compete for the work, there'd be:

•   a long-term approach to policymaking, with much greater stability of policy goals,
•  more transparency about these goals,
•  more buy-in from ordinary citizens, and
• more widespread appreciation of the notion of trade-offs inherent in policymaking.

While government would relinquish some of its powers, especially over the funding of public-sector bodies, it would still have the responsibility of articulating society's wishes coherently, and in raising the revenue for their fulfilment. These are functions that our democratic governments can do well; they are not so good at efficient achievement of long-term societal outcomes. Under a bond regime, that would be done by market forces: in economic theory - and on all the evidence - the most efficient means of deploying society's scarce resources. Social Policy Bonds would represent a radical transformation from current policymaking systems; but it's one that is now practical, has huge advantages, and would close the gap between government and the people.