It's taken me some time to work through Kaushik Basu's Beyond the Invisible Hand: Groundwork for a New Economics. The book has a lot of game theory examples, which I always find rather hard going, especially as the basic strategic insights often seem straightforward. It's quite a philosophical work as well. But the argument is clear enough.

Basu says that the methodological individualism of economics is flawed – and hence the invisible hand doesn't exist. Instead, he argues, people often behave according to social and cultural norms and beliefs, and not according to their rational self-interest. Furthermore, he argues that taking due account of the importance of custom and culture makes it more likely that inequality can be reduced than would be the case if the economy did consist of selfish individuals. He writes: “A fundamental step in broadening the scope of economics is to recognize that the feasible set of actions open to individuals is much larger than our models make it out to be.” (p27) What's more, the feasible set evolves over time.

For example, it's custom that means most of us do not steal other people's wallets when we have the opportunity to do so, or break the traffic laws when there is no policeman around. But the accepted norms can and do change. Most people most of the time do not act as a literal interpretation of rational individualism might suggest.

To emphasize the importance of custom and trust, Basu paints a picture of what an economy actually based on individual selfishness would look like – the nightmare of Kafka's The Trial. “[Kafka] concurs with Smith about the forces that can be unleashed through atomistic individual actions, with no centralized authority, but broadens our canvas of understanding by making us aware that while these can be forces of efficiency, organization and benevolence, they can equally be those of oppression and malignancy.” (p56)

It's an effective way of giving trust its due in the way we think about the economy. But I would make two observations about Basu's argument. One is that the point is by now rather familiar and many economists – maybe even most – would not dispute it. (It really came home for me on reading the 2000 paper by Ed Glaeser and Jose Scheinkman, Non-Market Interactions.) Nevertheless, models based on rational self-interest remain an extremely useful tool. Not only is it illuminating to try to understand departures from their predictions but in fact they often have strong predictive power, for in many circumstances people do behave in the way the conventional models assume.

The second is that Basu is chief economic adviser to India's Ministry of Finance. My experience of and reading about India suggests to me that it is a country very far from suffering from the excesses of deregulated markets under the influence of a conservative economic ideology. On the contrary, India still seems to have an excess of custom and convention rather than an excess of 'free' markets. So I suppose this book is aimed at Anglo-Saxon orthodox economists to try to persuade them, using their own techniques such as game theory, of the error of their ways. But in which case, the book is probably pushing at an open door, as many mainstream economists are already sympathetic to the general point even though remaining unconvinced of the need to ditch completely conventional models. You can hardly pick up a mainstream journal now without finding an article about trust or culture.

So this is another book pitched as an attack on mainstream neoclassical economics that, although far more sophisticated than many in the genre, fails to acknowledge the way the mainstream has developed or the breadth of views within it. Professor Basu obviously doesn't share the political views of some orthodox economists but that doesn't mean we need a new economics.

Rather, what economists need to do is to keep on chipping away at the huge areas of our ignorance about the economy, and especially the macro-economy and financial markets. When economists can disagree so fundamentally, about the deficit for example, we are obviously not in the realm of hard science. For me, that's a situation that cries out for more work, and more data, and, sure, some new modelling approaches. But please can we get over the obsession with overturning all of existing economics before getting on with it?