Why should we believe Jeffrey Sachs has anything significant to say on how to save the world? His formidable publicity machine - which cites Al Gore as a supporter - gives several reasons. He is a well-known economist - but then he would be, with that publicity machine. He has advised the United Nations, the World Bank and many governments - but so have I, and so have hundreds of my colleagues. More impressive is the statement that he has actually done effective economics.

In 1985, when Sachs was a 30-year-old Harvard professor, he went to Bolivia, which had an inflation rate of 20,000%. The first evening he was there he wrote a “shock therapy” economic action plan that brought inflation under control. His detractors say that he destroyed the economy and killed a lot of people. Only the cocaine industry thrived.

He claims responsibility for designing the USSR's post-communist strategy to revitalise a stagnating economy. Within three years, his “shock therapy” had cut the economy to a third of its previous size and handed half what was left to the mafia. Life expectancy plummeted.

As an economist working in the Third World and Eastern Europe over 40 years, I have to pick up the pieces after people like Sachs, who start advising as they step off the plane. I have had to try to stop a famine in Sierra Leone, to stop firms and industries collapsing in Eastern Europe, and to get a food and agricultural sector back on track everywhere. Real economists find out how the economy, the markets and the industries work; they collect the facts and they apply hard economic analysis. Only then do they advise.

Sachs's book fits what development professionals expect from the latest development guru. It is very convincing about things we know little about, and has a sprinkling of ideas seductively close to ours, for example, on wiping out malaria. As it moves to countries we have worked in, industries we know and technologies we use, it is just wrong.

Sachs starts off convincingly enough. I am no scientist, but what he says about global warming and its effects, and the immense and increasing demand for fossil fuels, is close to what I read in the newspaper. However, when he, an economist,concludes that in spite of this evidence, the problem is not too bad and global warming can be controlled at a modest cost, I am not convinced.

Much of his book is about the need to tackle global poverty, and here I do have the experience to check. He says, for instance: “It is not surprising that all 10 of the countries ranked as having the lowest human development are water-stressed countries with extensive dryland populations: Niger, Sierra Leone, Mali, Burkina Faso, Guinea-Bissau, Central African Republic, Chad, Ethiopia, Burundi and Mozambique.” This from a development guru, who is director of the Earth Institute at Columbia University and a former director of the UN Millennium Project.

One might have thought that someone in these organisations might be able to read a rainfall map. Or have visited Africa and walked through the thick bush of Mozambique, or have eaten bananas in damp Burundi (rainfall 60in), or endured the rainy season of Bissau (80in) or of Sierra Leone (300in). Or have noticed that Sierra Leone, now the poorest country in the world, was once one of the richest in the British empire - with the same climate. Or have noticed that the USA has a vast dryland area. Or have thought that other factors, including a history of civil war, might have given a better explanation.

The analysis is what Sachs calls “the science of economic development”, or what my African colleagues call “black man's economics”, a set of techniques that western economists would not dare use at home, but which they casually inflict on Africa.

This book uses the technique of plotting the growth rates of about 70 countries against possible causal factors: ranging from the average education attainment to fertility rate and social and racial cleavages. Since there are thousands of possible causal factors, the technique is poor economics. Its only advantage is that gurus can use it sitting in front of a computer in New York, and have their recommendations prepared before they board the plane, avoiding all the tiresome hassle of collecting facts and analysing them.

There are frequent inaccuracies, sometimes a whole argument, such as the one on the economics of railways, sometimes just the odd sentence or paragraph. And his presentation of statistics does not inspire confidence.

Sachs calls for vast funding for his solutions, which include a green revolution for Africa, population control, preserving the environment, control of disease, community development, and demonstration development villages. Yet all these ideas were implemented in colonial times, when the technology and money existed for them. They failed, and similar technically excellent projects have failed time and again over the past 60 years. Sachs has been told this, but fails to pass on the facts in his book. And he does not present an analysis of the reasons for their failure - which must be the key questions for development.

His description of what will happen if his good ideas are implemented is all based on “Let us assume...”. When this is combined with highly contentious and unjustified statements, it reads like a series of fairy tales. There is no excuse for suppressing key information, for getting one's facts wrong or for using bad economics. I suspect the average man on the average bus knows this: Sachs, to judge by the evidence of this book, does not.

Common Wealth: Economics for a Crowded Planet by Jeffrey Sachs
Allen Lane £22 pp386

Times Recommends

• 10 things you should avoid in hospital (unless there really is no option)
• Exclusive: Philip Pullman's essential reading list
• The White War by Mark Thompson