Conflicts between interest groups

This page and the others under the Features heading contain posts that appeared first in the Social Policy Bonds blog

The Social Policy Bond idea materialised when I thought about  how, when betting big money on an uncertain outcome, you could use some  proportion of your expected winnings to make that outcome more likely.  Fixing horse or greyhound races that way would be illegal, but could the  same principle be used legitimately for society's good? Perhaps new  betting markets are a way in which we could effectively issue Social  Policy Bonds.

Let's take a look at Polymarket, 'the World's Largest Prediction Market'. I'm writing this on 27 February 2025, when we see that you can bet on whether Timothée Chalamet will take his mother to the Oscars  ceremony. This is a simple one-off event that's easy to verify. It's  not difficult to imagine that, having placed a sufficiently large bet, a  group of punters could persuade Mr Chalamet to bring about the outcome  they desire, if necessary with the promise of a significant proportion  of their expected winnings. One of the difficulties of the Social Policy  Bond principle is that of setting up an experiment. The bonds have  their biggest advantage over conventional policymaking when the desired  social or environmental outcome is likely to be complex, long-term in  nature, and to require the investigation of a range of diverse, adaptive  approaches for its solution. Such socially desirable outcomes could  include the slashing of crime rates, significant improvements in the  physical and mental health of a country's citizens or, at the global  level, a reduction in the adverse impacts of natural disasters. We can't  test the efficacy of Social Policy Bonds against current policymaking  aimed at achieving such outcomes.

But what some  high-minded philanthropists could conceivably do is to take out a large  bet against a readily verifiable one-off outcome like, say, the  detonation of a nuclear device that kills more than, say, 500 people  within 30 years. They could then use their influence and funds,  predicated on their winning the bet, to make such a detonation less  likely. At first sight, this sounds tempting: indirectly channelling  resources into the achievement of an unambiguously positive social goal.  In net terms: yes; nuclear peace is hugely and unambiguously positive.  But Polymarket is merely a platform that facilitates peer-to-peer  trading, so that people are betting against each other. So for every $1  million bet on achieving nuclear peace, there would be people on the  opposing side of that bet who would lose that much. If governments  collectively decided to supply the funds that would be paid out on  redemption of Nuclear Peace Bonds, then millions of taxpayers would, in  effect, be paying for that outcome: a diffuse set of people, none of  whom would benefit in any discernible way from a nuclear detonation. Polymarket is different: it's likely there would be few people on the  opposing side of the bet of our philanthropists, and they could lose  substantial amounts if sustained nuclear peace were achieved. They would  be well placed and highly motivated, then, to co-ordinate efforts to  foil any attempts to achieve that goal.

Social Policy  Bonds would be best issued to encourage goals that are not only almost  universally desired, but that also do not create a small group of  highly motivated people who would oppose those goals. The national and global goals I discuss on this site are all like that, but it's also important that they be issued in such a  way that any losses from successful achievement of the targeted goals  would be spread so thinly that nobody would be motivated to take action  to frustrate the targeted goal. So, for instance, halving crime rates  could reduce the prospects for lawyers, jailors and nurses, but I don't  think that would motivate them actively to oppose that goal: the benefits (indirect financial, and non-financial) that they would each gain would outweigh the small financial losses they would incur.