It was to be expected that our present economic traumas would call into question the state of economics. “Why did no one see the crisis coming?”, Queen Elizabeth reportedly asked one practitioner. A seminar at the British Academy tried to answer and the FT has taken up the discussion.

The Queen’s question is understandable, given the subject’s claims on its own behalf. Ever since modern economics started in the 18th century it has presented itself as a predictive discipline, akin to a natural science. Since the future a year ago included the present slump, it is natural that the failure of the economics profession – with a few exceptions – to foresee the coming collapse should have discredited its scientific pretensions. Economics is revealed to have no more clothes than other social science. One cannot imagine the Queen in, say, nine months’ time, asking a leading political scientist: “Why did no one tell me that Labour was going to win the election?” She would understand that this was not a prediction that any political scientist could make with conviction, however much time he had spent studying present and past opinion polls.

Nevertheless, the Queen’s question was wrong, because it accepted at face value the predictive claim of economics – a feature that has distinguished it from all other social sciences. Karl Popper produced a famous argument against the possibility of prediction in human affairs: one cannot anticipate a new invention because, if one could, one would already have invented it. However, this objection can be overcome if one assumes a stable and repetitive universe in which rational actors make efficient use of the information available to them. In this environment, uncertainty disappears to be replaced by calculable risk. Shocks and mistakes may occur but these will cancel each other out, so that, on average, people get what they expect.

An important implication of this view is that shares are always correctly priced. This is the basis of the so-called efficient market hypothesis that has dominated financial economics. It led bankers into blind faith in their mathematical forecasting models. It led governments and regulators to discount the possibility that financial markets could implode. It led to what Alan Greenspan called (after he had stepped down as chairman of the US Federal Reserve) “the underpricing of risk worldwide”.

It has also led to the discrediting of mainstream macroeconomics. The efficient market hypothesis is simply an application of the recently triumphant New Classical school, which preaches that a decentralised market system is always at full employment. In their obsession with getting government out of economic life, Chicago economists claimed that any consistent set of policies will be learnt and anticipated by a population, and will therefore be ineffective. Since people – apparently including the 10 per cent or so unemployed – are already in their preferred position because of their correct anticipations and instantaneous adjustment to change, “stimulus” policies are bound to fail and even make things worse. Recessions, in this view, are “optimal”.

Most of those unversed in New Classical economics assume that John Maynard Keynes exploded these fallacies 70 years ago. Their re-emergence is not just the result of the failure of Keynesian macroeconomic policy to anticipate or deal with “stagflation” in the 1970s. It reflects a persistent bias in economics towards an idealised account of human behaviour; what Joseph Schumpeter called the “Ricardian Vice” of excessive abstraction. It is only by imagining a mechanical world of interacting robots that economics has gained its status as a hard, predictive science. But how much do its mechanical constructions, with their roots in Newtonian physics, tell us about the springs of human behaviour?

One of the most interesting contributions to the FT.com debate was the argument that, after Keynes, economists should have aligned their discipline with other social sciences concerned with human behaviour. Keynes opened the way to political economy; but economists opted for a regressive research programme, disguised by sophisticated mathematics, that set it apart. The present crisis gives us an opportunity to try again.

The reconstruction of economics needs to start with the universities. First, degrees in the subject should be broadly based. They should take as their motto Keynes’s dictum that “economics is a moral and not a natural science”. They should contain not just the standard courses in elementary microeconomics and macroeconomics but economic and political history, the history of economic thought, moral and political philosophy, and sociology. Though some specialisation would be allowed in the final year, the mathematical component in the weighting of the degree should be sharply reduced. This is a return to the tradition of the Oxford Politics, Philosophy and Economics (PPE) degree and Cambridge Moral Sciences.

Beyond this, the postgraduate study of macroeconomics might with advantage be separated from that of microeconomics. Courses in microeconomics should concern themselves, as at present, with the building and testing of models based on a narrow set of assumptions. Their field of applicability lies in those areas where we have reliable views of the future. Macroeconomics, though, is an essential part of the art of government, and should always be taught in conjunction with subjects bearing on this.

The obvious aim of such a reconstruction is to protect macroeconomics from the encroachment of the methods and habits of the mathematician. Only through some such broadening can we hope to provide a proper education for those whose usefulness to society will lie as much in their philosophical and political literacy as in their mathematical efficiency.

How to resolve this crisis in macroeconomics? The field must be revamped fundamentally. Some of its shortcomings are obvious. Before the financial crisis, most macroeconomists were blinded by the idea that efficient markets would take care of themselves. They did not bother to put financial markets and the banking sector into their models. This is a major flaw.

There is a deeper problem, though, that will be more difficult to resolve. This is the underlying paradigm of macroeconomic models. Mainstream models take the view that economic agents are superbly informed and understand the deep complexities of the world. In the jargon, they have “rational expectations”. Not only that. Since they all understand the same “truth”, they all act in the same way. Thus modelling the behaviour of just one agent (the “representative” consumer and the “representative” producer) is all one has to do to fully describe the intricacies of the world. Rarely has such a ludicrous idea been taken so seriously by so many academics.

We need a new science of macroeconomics. A science that starts from the assumption that individuals have severe cognitive limitations; that they do not understand much about the complexities of the world in which they live. This lack of understanding creates biased beliefs and collective movements of euphoria when agents underestimate risk, followed by collective depression in which perceptions of risk are dramatically increased. These collective movements turn uncorrelated risks into highly correlated ones. What Keynes called “animal spirits” are fundamental forces driving macroeconomic fluctuations.

The basic error of modern macroeconomics is the belief that the economy is simply the sum of microeconomic decisions of rational agents. But the economy is more than that. The interactions of these decisions create collective movements that are not visible at the micro level.

It will remain difficult to model these collective movements. There is much resistance. Too many macroeconomists are attached to their models because they want to live in the comfort of what they understand – the behaviour of rational and superbly informed individuals.

To paraphrase Isaac Newton, macroeconomists can calculate the motions of a lonely rational agent but not the madness of the crowds. Yet if macroeconomics wants to become relevant again, its practitioners will have to start calculating this madness. It is going to be difficult, but that is no excuse not to try.