Paul Collier, the director of the Centre for the Study of African Economies at Oxford university, has devoted three decades to the study of African economics. In this splendid book he has used that work to answer the most important question in development: why are so many countries now failing?

About 80 per cent of the population of developing countries lives in countries whose populations are becoming better off. Billions live in countries that are developing very swiftly. But almost a billion people – 70 per cent of whom live in sub-Saharan Africa – are in economically stagnant or declining countries. In all, 58 countries are in this desperate condition. Yet, as Collier remarks: “An impoverished ghetto of 1bn people will be increasingly impossible for a comfortable world to tolerate.”

Collier argues that these countries have fallen into one, or more, of four traps from which it is virtually impossible to escape. These are the “conflict trap”, the “natural resources trap”, the trap of being “landlocked with bad neighbours” and the trap of “bad governance in a small country”.

Seventy-three per cent of people in the bottom billion have been through civil war, 29 per cent are in countries dominated by the malign politics of natural resources, 30 per cent are in landlocked, resource-poor countries with bad neighbours and 76 per cent are in countries that have suffered long periods of bad governance and poor economic policies. Many have fallen into more than one of these traps.

What is to be done? Collier argues that trade, for all its potential benefits, will not help the bottom billion. These countries are uncompetitive exporters of labour-intensive goods and services, given the low costs and established positions of Asian producers. They cannot compete with China or Vietnam. Similarly private capital does not flow to these countries, except to exploit their natural resources. The problem is the reverse: huge capital flight. Collier estimates that almost 40 per cent of Africa's private wealth was held abroad in 1990.

Collier is also skeptical of the ability of aid to make much of a difference, at least on its own. He believes aid can help – and has helped – the bottom billion. But it has been a holding operation, rather than the start of sustained growth. He is particularly skeptical of the view that unconditional budget support will work. We have, after all, already had an experiment with the consequences of unconditional finance: oil revenues. Debt relief – the darling of the aid lobbies – is the closest thing to oil revenues that the aid industry can provide, a point its proponents ignore.

Aid will not get countries out of the traps. It cannot stop conflict, though it can help after one is over. It can do nothing about the natural resources trap: indeed, it is similar to possessing just another natural resource. It may help landlocked countries with improved transport infrastructure, but cannot eliminate the catastrophe of having bad neighbours.

So what else is needed to help countries in the bottom billion? Collier makes three suggestions: first, military intervention; second, laws, statutes and charters for improved governance; and, third, trade preferences.

The case for military intervention is most obvious, if controversial. Civil wars are so costly that well-timed military actions are quite likely (though not certain) to be cost-effective.

The second area demands changes in high-income countries: ceasing to take money looted from the poorest countries is one such change; elimination of bribery by their companies is another. It also needs charters of better governance for countries in the bottom billion: transparent management of natural resources is among the most important, the UK's extractive industries transparency initiative being a good start. The book also suggests charters for democracy, budget transparency, post-conflict situations and investment.

This idea sounds very naive. But the European Union has shown that external standards can make a big difference. Why should countries not sign up to charters of better governance in return for large quantities of aid? This is not imperialism. It is a bargain made in the interests of their own people.

The third suggestion is unrestricted access to the markets of high-income countries for labour-intensive exports from the bottom billion. Only thus, suggests Collier, are the resource-poor countries ever likely to break into world markets for manufactures.