Milton Friedman, who died on 16 November 2006, leaves a great economic legacy but a broadly failed philosophy. Ben S Bernanke (Alan Greenspan's successor as head of the United States federal reserve), the Bank of England's monetary policy committee, the eighteen-member governing council of the European Central Bank - all owe their roles and rules to Friedman.

When Greenspan reacted to the massive twin shocks of 2001 - the bursting of the technology-driven asset-price bubble and the uncertainty created by 9/11 - by reducing federal interest rates to the point where money had almost no price anymore, he was putting into practice the policy that Friedman had argued in 1946 would have averted the great depression of 1931.

But the ideas that Friedman stood for seem, if anything, more irrelevant today than they have been for a long time. Why the simultaneous success and failure?

Friedman used Abraham Lincoln in explaining his position on monetary policy: "You can fool all of the people some of the time, and some of the people all of the time. But you can't fool all of the people all of the time." A central bank can have occasional impact on the level of economic activity by controlling interest rates. But if this power is used too often, firms and households will adjust expectations of price changes and neutralise any impact on real activity. This is the core of monetarism, and it was and still is correct.

The rather dogmatic monetarisms practiced by Margaret Thatcher and Ronald Reagan in the 1980s, when the theory was thought to imply strict targeting of money supply figures, were less successful than the current version. It took fifty years for the Friedmanite truth on money to be comprehensively accepted and refined. Friedman believed, as some of his Chicago colleagues deny, that fooling some of the people some of the time is an important and useful tool of policy in crises, but that central banks must earn the trust that they put at risk when reflating the economy. This is now orthodoxy.

The European Central Bank transparently shows on its website, its view of its world: one in which changes to interest rates change prices, with various possible impacts on real events, like wages, employment and investment in between.

The independent central banker, or committee, with a politically determined inflation target, is a pretty thoroughly Friedmanite creation. It is early to tell yet, but Greenspan's big reflation in 2001 will be, if a big recession is indeed averted, a vindication of Friedman's beliefs.

Tony Curzon Price received a PhD from University College London (UCL), and worked as a consultant economist for more than ten years. Since 1997, he has lectured on economics and energy policy to postgraduates at Imperial College, London, and at the École Polytechnique Fédérale de Lausanne (EPFL)

Also by Tony Curzon Price in openDemocracy:

"Why fly?"
(12 June 2002)

"Holistic hunters' knowledge can be harmful"
(3 September 2002)

"Turning the tide: how fear will make People Flow obsolete"
(10 July 2003)

"iCommons for beginners"
(20 June 2006)

Milton Friedman, philosopher

Or maybe not. Friedman's economics, however scientific, is always also in the service of a libertarian worldview. In Capitalism and Freedom, his 1962 credo, he describes individual freedom as the ultimate value and asks for a minimal, Lockean government with power to enforce law, protect property and defend the realm. The state implied by the old Keynesian, social-democratic orthodoxy - one that saw tax and spend as a duty aimed at regulating the level of economic activity - was anathema to Friedman not because of its technical incompetence in monetary policy, but because it was just too big; it meddled too far into the private realm.

Friedman's other two big contributions to economics in the 1950s are one on the philosophy of science, and another on consumption behaviour. Both of these, like the work on inflation, are contributions to his philosophy. In A Theory of the Consumption Function (1957), he shows that a household that plans, calculates and looks after its interests, saves and consumes according to a lifetime scheme, not according to the whim of current income. The economics were impeccable by the standard of the time, but the main policy point of the article was to show that the Keynesian's favourite policy tool - of changing the amount in people's pockets through tax and spend - should usually have only muted effects on economic activity.

The assumption that individuals, firms and households are actually calculating in the way posited by Friedman's theory was obviously wrong: motivations, powers of foresight, the impact of beliefs on decisions - all these were swept aside in Friedman's model of behaviour. This is where his second contribution of the decade, The Methodology of Positive Economics (1953), comes in. He presents a defence of a "test it and see" approach to the realism of economic assumptions. It is closely related to Karl Popper's "logico-deductive" method: you posit assumptions about behaviour and environment; you derive logical conclusions from these about the behaviour of society as a whole; you look at statistics to see if, in aggregate, your conclusions hold.

What you do not have to do, goes the theory, is to test the assumptions themselves. Hence, the "rational household" may bear no relation to any behaviour that you or I have seen or imagined, but the theory produces testable predictions, especially about aggregates of behaviour. If the predictions turn out to be correct, we can argue that households behave "as if" they are as assumed. This methodological move gave more credence than deserved to the view that individuals are "really" self-regarding, atomised, interest-calculating machines. And if humans are to be treated "as if" this is their nature, then maybe libertarianism is indeed the best organisation for us.

The libertarian advantage through methodology, however, was short-lived. Over the thirty years that followed, micro-economics - the part of the subject that tries to explain small scale decisions, like the impact of a minimum wage on firms' hiring decisions - went through its own sort of quantum revolution: as the scale and detail of choice became finer, strange effects started to loom larger. Of these, effects like "moral hazard", "adverse selection", "costly signalling" added a depth and coherence to the notion of the failure of markets that allowed almost any macro-economics to be justified - at least on theoretical grounds.

George Akerlof, 2001 Nobel prize-winner, describes the withering away of the market for second-hand cars with this "parable of the lemons": the owner knows the quality of the car whereas the buyer does not; if quality is above average for the second-hand market, the owner should know that she will not be able to get value by selling it, because it will be tainted by the buyer's expectation that it is only of average value.

The potential seller of an above-average quality car will therefore tend to have good reasons to hang on to it, while the opposite argument holds for the owner of a below-average quality car (a "lemon", in American slang). The second-hand car market will race to low quality, with loss of economic potential all around.

The "lemons" example raises questions of reputation (how do you create it? what is it worth?), community (how history of repetition changes the rational calculus, how overlapping spheres of interaction make each sphere work), institutions, and, maybe most importantly, of the production of meaning: how do actions (here, offering a car for sale) lead to the creation of interpretable meanings ("this car is of low quality!") through social understanding (here, of incentives in a market).

The parable points to a world of analysis beyond markets: political institutions, committees, the minutiae of procedure. These practices have acquired significance and function in economics absent from the narrow methodological individualism of 1950s Friedman. His "as if" approach is meant to lead us to ignore these effects. But in a world dominated by them, a minimum wage can raise the level of employment; government spending can be more powerful than monetary policy; speculative bubbles (Alan Greenspan's "irrational exuberance") can seem rational; restrictive practices - like those of lawyers, taxis and accountants - can seem sensible; and even the practices of community-building and the creation of collective identity can be seen to fit. Behaviour and institutions find their logic in the painstaking mechanics of making meaning.

The remarkable conclusion from the "quantum revolution" of information economics is that the minute slivers of light it opens are sufficient to let in a whole world. And you do not even have to rely on fooling the foolish for this to happen. "As if" economics is no firm anchor in this strange world.

Also in openDemocracy on economic ideas - and an elusive Nobel prize:

Yves Gingras, "Nobel by association: beautiful mind, non-existent prize" (23 October 2002)

Tim Harford, "Thomas Schelling; games of enlightenment" (14 October 2005)

Fardia Khan, "Muhammad Yunus: an economics for peace" (25 October 2006)

Why did the economics win and the philosophy fail?

Keynesian fine-tuning is gone; Friedman's independent central banks are the rage across the political spectrum. In monetary policy, the central banks are careful to manage the trust they have earned from the financial markets, so that when really needed, we can still constructively fool a few of the foolish. But the dream of Friedman's limited Lockean state is further than ever from realisation.

Minimal taxes, limited government, foreign-policy non-intervention, free movement of goods and people - these essentially political goals that Friedman was working towards have probably receded in the last ten years. Welfare, migration control and taxes of between 35% (United States) and 50% (France) of income are the norm in the rich west. War, national integration policy, identity politics, resurgent traditionalisms - all the anathema of libertarianism - are the questions that do most preoccupy today. There must be something profoundly missing in a political philosophy that has no room or explanation for the most important questions of the day.

The issues that have passed the libertarian economists by are exactly the ones raised in the "lemons" example: how do trust and meaning emerge from social interaction, and what chaos ensues when they are destroyed? This is the study of the public realm - what we all together share (or fail to), how we do it, and what it means to us - and this has no place in the Lockean state or the libertarian economists' worldview.

Monetary policy, where Friedman succeeded, is precisely that part of the public realm where the tokens of value and meaning can be broken down into the choices of atomised individuals. Where value is captured by money, policy based on thoroughgoing individualism makes sense.

In The Methodology of Positive Economics, Friedman offers a possible evolutionary justification for the heroic assumptions of his economics: firms can be thought to behave "as if" they are perfectly calculating, selfishly motivated, because if they were not, others would take their place. The world selects for the behaviour that the theory posits.

L'arroseur arrosé, the tables are turned: the same evolutionary levelling has befallen Friedman's philosophy. The world has selected those parts that apply to it. Money works as a measure of value exactly where the relationship between anonymous buyer and seller is all that there is to a transaction. The subtle social sifting of positions has retained Friedman's for the part of the public realm that nearly fits his assumptions. Here, but not elsewhere, it is "as if" Friedman was right.