Blindheim – blind home – is a strange name; but that is the name of a village in Bavaria. Presumably it had a home for the blind some time in its history. But what made it famous was a battle in the war over who was to succeed the childless Charles II, king of Spain and ruler of most of Latin America at the end of the 17th century. In a battle fought in Blindheim in 1704, General John Churchill, Duke of Marlborough and an ancestor of Winston Churchill, commanding the forces of England, Holland, Savoy, Baden and the Holy Roman Empire (as Austria was then known), defeated the armies of France, Spain and Bavaria. Queen Anne rewarded him with the ruined manor of Woodstock and a stupendous 240,000 Pounds. He rebuilt the manor and renamed it Blenheim Palace; presumably he found Blindheim too German or unattractive a name for it. Blenheim became a treasured name; a village on New Zealand’s South Island, across the straits from the capital Wellington, was named Blenheim. The New Zealand branch of Australian Agricultural Economics Society held a conference in Blenheim in 1988; Ronnie Horesh read a paper there.

Then, as now, governments promoted economic growth, both because it would make their citizens better off, and because they would get more revenue out of the citizens, which they hoped to spend on social welfare services. Leftists thought this was the way to ameliorate social injustice; rightists thought it was a way to feed government corruption. The two met in conferences and shouted at one another.

Meanwhile, governments issued bonds with a fixed maturity and bearing a fixed interest rate, and used the proceeds to achieve social objectives. There was no connection between the raising of the loans, and the achievement of the objectives. Governments’ ability to borrow depended on how far lenders trusted their promise to pay interest and repay the bonds and what inflation the lenders expected in the meanwhile; governments’ achievement of social goals depended on their efficiency and honesty. Governments could contract out their services if their outcome was measurable: but they could not do that for broad social objectives such as making the poor less poor or the sick less sick. So the conventional government model of borrowing and spending worked poorly.

Self-interest made the private sector cut costs and deliver results faster. Horesh sought a way of introducing the same incentive in government expenditure programmes. Just then – in 1989 – Ayatollah Khomeini had issued a fatwa against Salman Rushdie for writing The Satanic Verses: he promised that a Muslim who killed Rushdie would get about $6 million if he (the murderer) survived, and would go to Islamic heaven if he died.

Horesh thought, if someone could be rewarded for murder, why could people not be rewarded for doing something good? He suggested that a government should borrow to achieve measurable results – for instance, to bring down unemployment to a certain percentage – and should repay the bonds only when it had achieved the objective. Meanwhile, it should pay no interest on the bonds. Such bonds would be like bets placed on horses: investors would place bets on government efficiency. Those who thought the government would achieve the result quickly would be prepared to pay a high price for the bond; those who were skeptical of the government’s ability or willingness would pay a lower or no price. A bond that a government issued, promising to use the proceeds on a welfare programme, would sell at a discount; the discount on it would depend on bids received from optimists and pessimists. If market interest rate was 10 per cent and enough investors thought the government would achieve its objective in a year, a $100 bond would sell at $90. If investors had little trust in the government, and thought there was one chance in a hundred that it would keep its promise, they would be prepared to pay $1 for a $100 bond. Horesh’s idea was to turn governments into race-horses; if the odds were good enough, bettors would flock into government bonds as they did to horse races.

To make the bonds more attractive, Horesh suggested that the government should contract out the achievement of the objectives to the private sector. The contractor would have a profit incentive to achieve the results.

He suggested a number of areas in which social policy bonds could be tried out. For instance, law-and-order bonds could bet on the number of reported crimes, employment bonds on the unemployment rate, health bonds on life expectancy, housing bonds on the number of homeless, education bonds on literacy and numeracy, pollution bonds on particulate matter in the air, and so on.

It is now almost 30 years since Horesh first wrote about social policy bonds; some governments have woken up to their promise. The British government has set up a Government Outcomes Lab to advise non-government organizations that want to try out the bonds, and set aside 80 million Pounds to top up their experiments. The subsidy would depend on how far the experiments reduced government welfare expenditure. The government of New South Wales has similarly set up an Office of Social Impact Investment to issue bonds and use them to reward those who resolve social problems such as child criminality and ineffective parenthood. According to Rockefeller Foundation, 17 American state and local governments are using social impact bonds.

There has been much intellectual experimentation around the idea Ronnie Horesh floated a generation ago. Robert Shiller, who shared the 2013 Nobel Prize for economics, put forward a number of variants on the idea of economic funding of social objectives: crop insurance based entirely on weather variables (such as rainfall and its timely distribution) that affect farmers’ income, government bonds called trills which would pay investors a trillionth of gross domestic product and would therefore be popular with growth optimists, indexing income tax rates to income inequality so that they would go up whenever inequality rose, livelihood insurance for which successful people would pay higher premiums and from which unlucky ones would benefit, stock markets for standardized single-family homes, and so on.

Governments are far behind ideas, but some have begun experiments. The return on such experimentation would depend on the current inefficiency and corruption of governments; it should therefore be particularly high in India. But perhaps because the returns on corruption are so high, there has not even been talk of experimentation in India; billions were squandered by corrupt ministers in the past decade, but our honest prime ministers have been studiously disinterested in preventing corruption. One cannot make a horse drink water; but it should be pointed out to rulers that all that they have to do is to pick up the instruments and try them out. Paradoxically, those instruments are being tried out in rich countries which do not have much corruption, and ignored in countries where they could be most beneficial. We have a surprisingly honest and competent office of the Comptroller and Auditor General which publishes great reports on official misdeeds; maybe it should be equipped with the technology of social policy bonds and placed in charge of all social expenditure.