BOOK REVIEW:Basic Instincts: Human Nature and the New Economics by Peter Lunn; Marshall Cavendish Business. £19.99 (€25)

We know that economists make mistakes a lot of the time. But what if they're not just mistaken but wrong, if there's a basic flaw in the way classical economics understands human behaviour?

In his lucid, radical and entertaining book, Pete Lunn, who works with the ESRI, makes a compelling argument that the most basic, and therefore least examined, assumptions of economists are utterly at odds with the evidence of how we actually make choices.

Bringing to bear his original training as a neuroscientist and his long experience as a journalist (he worked with BBC's Newsnight and as founding editor of Newstalk radio in Dublin), he shows that we are not really the creatures economists assume us to be.

In one way, Lunn's basic argument ought to be reasonably obvious. Economics is rooted in mathematics and maths deals in abstractions.

Mathematical models work best if you assume that the systems they try to represent are stable and rational. It is simply easier to make these models if you take it for granted that people will make the choices that are in their own best interests.

Those interests, in classical economics, are defined in self-seeking terms. Humans are understood as independent, selfish and rational materialists.

So consumers will always try to get as much as possible as cheaply as possible. Companies will always try to maximise their immediate profits. Workers will do as little as they can for the highest wages they can get.

But if all of this were true, economic models really would be crystal balls and economists would be vastly better than they have proved to be at understanding and predicting the real world.

None of this is as obvious as it should be, however. Economics is much more an ideology than a science. With the triumph of so-called free-market economics (what Lunn calls "marketopian" thinking), its orthodoxies have come to be seen, not as theoretical abstractions, but as the very definition of reality.

"Ideas," writes Lunn, "can become so embedded and accepted that where they come from is largely forgotten." And these core ideas come from simple convenience: "orthodox economics assumes that people are independent, rational, selfish materialists because it makes it easier to do the sums . . . "

Lunn begins his questioning of these assumptions with what might be called the Leonard's Corner test, after the area of Dublin where he now lives. He uses it to introduce his notion of "economic life", rather than "economics" - the texture of real human behaviour in the marketplace as opposed to the dubious abstractions.

He notes, for example, that some of its small businesses thrive even though they charge higher prices than larger competitors. This is, he says, because local human "relationships prove more powerful than market forces".

His case is that those relationships are much more subtle, sophisticated and collective than anything that is dreamt of in the dominant economic philosophy.

The same, he argues, is true of businesses. Lunn takes as his case study one of the 20th century's great success stories, the razor blade manufacturer Gillette. It did not, he says, succeed by doing what classical economics holds it must do - "providing razors people wanted, as efficiently as possible, at competitive prices in search of profit".

Instead it invented a product men didn't know they wanted, has continually spent vast sums improving the product and making it more expensive to buy, and spends $800 million (€510 million) a year that could have been used to reduce prices on marketing to add magic to its brand.

According to orthodox theories, it ought to have gone out of business years ago. It thrives because those theories tell us so little about economic life.

If the accepted notions of basic human instincts don't really work, how can we build a surer foundation? Lunn turns to the emerging discipline of behavioural economics, which, unlike the orthodox model, is based on actual observation of real people. It draws on the systematic study of the ways in which we make choices.

The Ultimatum Game, which Lunn calls "the equivalent of the atom in economic analysis", is a good example. In the experiment, a person is given a sum of money on condition that they share it with a stranger, and has to decide the appropriate proportion to offer. Both sides know that if the offer is accepted, both get to keep their share. If it is declined, all the money is taken back.

According to classical economics, the stranger should accept whatever is offered, since the alternative is to get nothing. Knowing this, the first person should make a low offer.

But actually, most people will not accept low offers at all and most people will not actually make such offers. Across different societies, and even when the money involved is real and substantial, people prefer notions of fairness and honour to simple selfish desires.

With both rigorous logic and demotic charm, Lunn builds on this and other findings to make a persuasive and highly thought-provoking case that conventional economics is actually far too pessimistic in its assumptions about human instincts. We balance our greed with notions of fairness, with a liking for familiarity and an aversion to risk, with a valuing of relationships, and with a strong instinct for co-operation.

For all its supposed devotion to a realistic view of human behaviour, free market economics often runs counter to these genuine instincts and fails accordingly.

This is an important book that both popularises and teases out the implications of an emerging revolution in thought.

Delightfully, Lunn thinks it important enough to write in a way that is never patronising but that is nonetheless entirely accessible to the general reader. In that, he has given his fellow economists another reason to gnash their teeth.