Why economists are almost always wrong

That verbose title is almost the reverse of a quintessentially arrogant statement of economic supremacy published in the UK’s Daily Telegraph - on the editorial page of the business section - by Andrew Lilico. Entitled “Economists are nearly always right about things, despite what you may think” in the print edition, its content and tone encapsulated everything about economic theory, and economists’ blind belief in it, that led me to write Debunking Economics over a decade ago.

Figure 1: Lilico's article in the print edition of the UK Daily Telegraph

The two main factors that made that book necessary were the capacity of economists to intimidate opponents with their apparent depth of knowledge of a difficult subject, and the reality that economists knowledge of their own subject was, to coin a phrase, not even shallow: it was frequently outright wrong.

Lilico’s article contained numerous examples of this, but I’ll focus on an apparently esoteric one: his defence of the assumption of rational behaviour, which is an integral part of conventional economic theory. As a piece of prose, his defence of it is a gem of “logic” and linguistics that should have fans of Noam Chomsky rolling in the aisles, but that’s not why I’m quoting it here:

“But rationality is not an assumption of orthodox economic theory in that sense. Instead, it is what is called an “axiom”. No behaviour can prove that people aren’t, in fact, rational, because for an orthodox economist the only kind of explanation of any behaviour that counts as an economic explanation is an explanation that makes sense of that behaviour - that shows why the behaviour is rational. (Lilico, 2014/01/22)”

OK, now stop giggling (or put the brandy bottle away). Apart from the fact that this would show both committing suicide as rational (and yes of course there are economic theories that say it is) and deciding not to commit suicide also as rational, it’s wrong about economic theory itself. This is because Paul Samuelson, the prime developer of the underlying theory of rational choice, reacted to precisely this criticism of economics as unscientific: if a theory can explain everything, it explains nothing - a point Lilico himself makes in taking a swipe at professional critics like me:

“Irrationality and other heterodoxy is usually little better than an all-encompassing conspiracy theory, explaining everything and thus nothing -- for while many behaviours may not be rational, there is no behaviour that is not irrational. In the process, heterodoxy misses all that is fruitful and important.”

OK, he shot himself in the foot by blasting heterodoxy for “explaining everything and thus nothing” when precisely the same critique applies to his unfalsifiable definition of rationality, the treble negative in the penultimate sentence may necessitate another swig of brandy -- and Chomskyites, get up off the floor. But let’s get back to Paul Samuelson and why Lilico doesn’t know what he’s trying to talk about.

In 1938, Samuelson (Samuelson 1938) developed what became known as “the axioms of revealed preference”, in response to criticisms that a key component of the economic concept of rationality was unobservable, and therefore more metaphysical than scientific -- according to Popper’s definition of science.

This was the idea of subjective utility, and its representation in so-called “indifference curves” that John Hicks recently devised. Samuelson, in other words, wasn’t happy that the theory rested merely on unchallengeable beliefs -- yet Lilico believes he’s defending economic theory by claiming that rationality is an uncontestable “axiom”. This is typical: mainstream economists like Lilico don’t know the history of their own discipline, let alone know much about its flaws.

The economic literature also contains a disproof of Lilico’s dismissive “No behaviour can prove that people aren’t, in fact, rational”. In fact, Samuelson designed his  “axioms of revealed preference” so that it would be possible to test the behaviour of people, and conclude whether they were or were not rational. These axioms of rational behaviour were:

• Completeness: shown any two shopping trolleys full of goods, a consumer could say which one she preferred (or if she was indifferent between them).
• Transitivity: if the consumer preferred trolley A to trolley B, and trolley B to trolley C, then she necessarily had to prefer trolley A to trolley C.
• Non-satiation: if she was indifferent between two trollies A and B, throw one Mars Bar into trolley A and she must prefer A to B.
• Convexity: this is a bit trickier to explain. If she was indifferent between trolleys A and B, then any trolley containing a mixture of the goods in A and B would have to be preferred to A or B.

Samuelson’s argument was that, if a consumer behaved rationally as defined by these axioms, then you could present them with a whole range of shopping trolleys, and derive what their “indifference curves” were -- so it wasn’t true that they were “unobservable”, you could actually derive them from empirical research.

In the late 1990s, the German economist Reinhard Sippel decided to test this in a meticulous experiment: he presented his students with sets of prices that they could use to “buy” a handful of goods, where the choices were set up to test how rational these students were according to the “axioms of revealed preference”.

The substantial majority of his students turned out to be “irrational” according to the theory of rational behavior, because they violated some or all of these axioms. They’d choose a trolley A instead of B one time, and then B instead of A another; given a choice between A and B, they’d choose B, then between B and C, they’d choose C, and then between C and A, they’d choose A.

Were they “irrational”? No, of course not: it’s the mainstream economics definition of rational that is “irrational”, because it ignores the blinding complexity that exists even in what appears to be a simple choice between different combinations of a handful of goods -- say ten goods for example, where you could buy between zero and nine units of each.

How many shopping trollies could you fill with different combinations of those ten goods? One hundred million of the buggers. Try to imagine a parking lot filled with 10,000 shopping trollies in one direction, and 10,000 in the other. Do you think that (given a set of prices and an income constraint) you could easily pick the one you preferred out of all others you could afford?

Of course not. So what you rationally do is you ignore many of those combinations, using grouping of different goods together (“buy fruit” rather than “buy apples or oranges or bananas or…” according to what’s cheapest), follow habit (“Dad, why do you always give us turnips?”), custom, etc. So rational behavior isn’t what Samuelson defined it to be -- to consider all options and pick the best -- but to reduce the complexity of your decision-making process in sensible ways so that you can make a satisfactory decision in finite time.

I have great respect for Sippel because I am sure he thought, when he set the experiment up, that he would both vindicate the theory and bring it alive for his students. Instead, he looked at the experimental results and concluded that they contradicted economic theory:

“We conclude that at the evidence for the utility maximization hypothesis is at best mixed… we would like to stress the diversity of individual behaviour and call the universality of the maximizing principle into question…,” he said. “We should therefore pay closer attention to the limits of this theory as a description of how people actually behave, i.e. as a positive theory of consumer behaviour. Recognising these limits, we economists should perhaps be a little more modest in our `imperialist ambitions' of explaining non-market behaviour by economic principles.”

Sippel’s reaction to his experiment was the mark of a real scientist. Lilico’s defence of economics despite its many empirical failings is the mark of a zealot. That is the real weakness of mainstream economic theory: that it engenders in its followers a manic belief that is impervious to empirical reality.

Steve Keen is author of Debunking Economics and the blog Debtwatch and developer of the Minsky software program.