Steve Keen is the author of Debunking Economics: The Naked Emperor of the Social Sciences

Also in this section

March 2002

**Is There Anything Worth Keeping in Standard Microeconomics?**

May 2002

**Comment on Bernard Guerrien’s Essay**

May 2002

**In Defense of Basic Economic Reasoning**

May 2002

**Response to Guerrien’s Essay**

May 2002

**For Guerrien… and beyond**

June 2002

**High priests and run-of-the-mill practitioners**

June 2002

**Theoretical substance should take priority over technique**

June 2002

**Superior Analysis Requires Recognition of Complexity**

June 2002

**What should be retained from standard microeconomics?**

June 2002

**An American Undergraduate Point of View on Economics Education**

June 2002

**Towards a New Economics**

September 2002

**Yes, There is Something Worth Keeping in Microeconomics**

September 2002

**Can we please move on? A note on the Guerrien debate**

October 2002

**Once Again on Microeconomics**

Last modified

30 November 2006

30 November 2006

info visites 1577

{id_article}

There are many perspectives from which Guerrien’s question "Is There Anything Worth Keeping in Standard Microeconomics?" can be answered. I will consider two: the empirical and the mathematical.The Empirical PerspectiveThere is little doubt that the conventional theory of the firm and of consumer behaviour are bad empirics.

On the former front, there are now numerous surveys in economic literature which establish that what neoclassical economics teaches as "the" behaviour of the firm — profit maximisation by equating rising marginal cost to falling marginal revenue — applies to at best less than 5 per cent of firms and 5 per cent of products. Research into the actual cost structure facing most products, have routinely found that in 95 per cent or more of cases, marginal costs remain constant or fall across the relevant range of output. Research into the actual behaviour of firms shows similarly that, in 95 or more per cent of cases, firms chase the maximum possible level of sales without any consideration of declining marginal revenue (Fred Lee is the modern chronicler of this literature, with his book Post Keynesian Price Theory being the ultimate reference).

Similarly, the vision of consumers deciding what to purchase by working out the point of tangency between the budget hyperplace and indifference hypercurve has failed miserably in experimental work. The latest and best reference on this, Sippel 1997, concludes very honestly with the observation that the theory failed to predict students behaviour in a very well designed controlled experiment, and that as a consequence economists "should therefore pay closer attention to the limits of this theory as a description of how people actually behave".

The fact that despite these empirical failures, economists continue to teach the standard theory of the firm and consumer theory is perhaps the strongest indictment one can give against standard microeconomics. It has acted as a barrier to an honest confrontation with the real world, and deserves to be dropped on that ground alone.But of course, it won’t be: because ever since Friedman’s defence of the "as if" approach to economic methodology, economists have felt justified in ignoring the real world since whatever firms and consumers think they are doing, they must be behaving "as if" they were doing what economists say they do, otherwise they wouldn’t be profit maximisers or rational consumers.

**The Mathematical Perspective**

Well, if economists can’t be persuaded to consider reality because it conflicts with their mathematics, we’re going to have to turn our attention to the mathematics itself. And it turns out that there are good mathematical reasons to reject standard microeconomics.

Let’s take first of all the theory of consumer behaviour. The standard presentation-of a consumer making a choice between two commodities-makes the exercise appear simple. But whoever heard of a consumer living on just two commodities? Yet each additional commodity considered involves an additional set of axes-three for three commodities, 4 for four-and each axis increases by an order of magnitude the number of choices facing the consumer.Once we get anywhere near the number of commodities and number of units per commodity

that a typical Western consumer buys on a monthly basis, the number of choices explodes to such a level that simple "rational" utility maximisation is inconceivable. For example, if we simply consider a purchase of less than ten items each from a set of 30 commodities, the number of combinations to be considered is 1030 (to put this number in perspective, the age of the universe is under 1018 seconds).

This "curse of dimensionality" is a well-known phenomenon in computer science, and it is well-known that an exhaustive maximisation approach is simply impossible with such problems.

Instead, consumers have to be following algorithms that drastically reduce the choice space: letting their choices be guided by custom, convention, habit, income-constrained tastes, etc. The conventional, simplistic vision of consumers as rational utility maximisers is a positive hindrance to serious study of the interesting question of how consumers manage to make consumption decisions in the face of overwhelming choice. By not honestly considering the

mathematical implications of their theory of consumer behaviour, economists are practicing bad mathematics.The same applies in the theory of the firm, where students start off being taught bad mathematics. All economists know the "perfect competition" assumption that the derivative of the market demand curve with respect to the output of a single firm is zero. Not enough know that George Stigler-hardly a radical there-pointed out in 1957 that this is mathematically invalid: if the market demand curve is negatively sloped with respect to market output, it is negatively sloped with respect to the output of a single firm. This is a simple application of the chain rule for a continuous function (and also a product of the assumption of atomism).

A minority of economists appear to know Stigler’s attempt to get around this by redefining marginal revenue for the individual firm as market price plus market price divided by the number of firms times the market elasticity of demand, coupled with the assertion that "this last term goes to zero as the number of sellers increases indefinitely" (Stigler 1957: 8). Too few realise that this was a sleight of hand: the term is a constant if there is a minimum firm size, and therefore the firm’s marginal revenue is always less than price.So-called mathematical economists have attempted to evade this by assuming that each

industry consists of an infinite number of firms each producing an infinitesimal output (not an infinitesimal fraction of total industry output, but an infinitesimal output!). But this makes a mockery of the theory of exchange-how can consumers buy a single unit of anything if they have to go to an infinite number of producers to buy a single unit?-and of the theory of production itself-how can diminishing marginal productivity apply (the foundation of the proposition that marginal cost rises) if the minimum firm size is zero?

All these nonsense propositions have been put forward to defend the indefensible concept of perfect competition, and they are propositions that any decent applied mathematician would reject outright.

Yet this kind of behaviour-proposing decision processes that are empirically impossible, making assumptions that are absurd in order to preserve an initial proposition that has been shown to be fallacious-is a direct consequence of adherence to the conventional theory of microeconomics. It is bad, unscientific behaviour, that should have no place in a serious discipline.

*References:Lee, F., (1998). Post Keynesian Price Theory, Cambridge University Press, Cambridge.Sippel, R, (1997). ’Experiment on the pure theory of consumer’s behaviour’, Economic Journal 107: 1431-1444.Stigler, G.J. (1957). "Perfect competition, historically considered", Journal of Political Economy, vol. 65: 1-17.*