The following year corrupt officials, businessmen and politicians pinched at least $30m from the Malawian treasury. A bureaucrat investigating the thefts was shot three times (he survived, somehow). Germany said it would help pay for an investigation; later, burglars raided the home of a German official and stole documents relating to the scandal. Malawi is no longer a donor darling. It now resembles a clingy lover, which would be dumped were it not so needy. It still gets a lot of foreign aid ($930m in 2014), but donors try to keep the cash out of the government’s hands. 

Foreign aid can work wonders. It set South Korea and Taiwan on the path to riches, helped extinguish smallpox in the 1970s and has almost eliminated polio. Unfortunately, as Malawi shows, it is liable to be snaffled by crooks. Aid can also burden Foreign aid can work wonders. It set South Korea and Taiwan on the path to riches, helped extinguish smallpox in the 1970s and has almost eliminated polio. Unfortunately, as Malawi shows, it is liable to be snaffled by crooks. Aid can also burden weak bureaucracies, distort markets, prop up dictators and help prolong civil wars. Taxpayers in rich countries dislike their cash being spent on Mercedes-Benzes. So donors strive to send the right sort of aid to the places where it will do the most good. How are they doing?

A decade ago governments rich and poor set out to define good aid. They declared that aid should be for improving the lot of poor people—and not, implicitly, for propping up friendly dictators or winning business for exporters. It should be co-ordinated; otherwise, says William Easterly of New York University, “the poor health minister is dealing with dozens of different donors and dozens of different forms to fill out.” It should be transparent. Where possible, it should flow through governments.

These are high-minded ideals, reflecting the time they were laid down: the cold war was over and the West had plenty of money. They are nonetheless sound. Aid-watchers, who row bitterly over whether the world needs more foreign aid or less, mostly agree with them. They tend to add that aid should go to relatively free, well-governed countries.

By almost all of these measures, foreign aid is failing. It is as co-ordinated as a demolition derby. Much goes neither to poor people nor to well-run countries, and on some measures the targeting is getting worse. Donors try to reward decent regimes and punish bad ones, but their efforts are undermined by other countries and by their own impatience. It is extraordinary that so many clever, well-intentioned people have made such a mess.

Official development aid, which includes grants, loans, technical advice and debt forgiveness, is worth about $130 billion a year. The channels originating in Berlin, London, Paris, Tokyo and Washington are deep and fast-flowing; others are rivulets, though the Nordic countries are generous for their size. More than two-fifths flows through multilateral outfits such as the World Bank, the UN and the Global Fund. Last year 9% was spent on refugees in donor countries, reflecting the surge of migrants to Europe.

As the aid river twists and braids, it inundates some places and not others. India contains some 275m people living on less than $1.90 a day. It got $4.8 billion in “country programmable aid” (the most routine kind) in 2014, which is $17 per poor person. Vietnam also got $4.8 billion; but, because it is much smaller and rather better off, that works out to $1,658 per poor person (see map). By this measure South-East Asia and South America fare especially well.