Size Matters

A lot of economics is about predicting people’s behaviour using axioms akin to laws in physics. The assumption that economics is like the physical sciences really bothers me. I side with Friedrich Hayek: the system is too complex for us to be able to provide a precise model of economic behaviour.

That scientific tools should not be directly applied in economics has been said many times before, but it’s not a point that usually receives the attention that it should, probably because most (but not all) of the people who say so are those who suck at maths. It’s hard to say why people persist with the use of complex calculus in behavioural modelling even though it is impossible for a majority of the population to even understand the reasoning process being ascribed to them, and the minority that does understand it would have to be in need of therapy if they really stood around at supermarkets trying to figure out how many units they should buy to equate the marginal utility to their marginal utility from money.

Russ Roberts argues that economics is more like biology than physics, and draws an interesting analogy between the two:

“We do not expect a biologist to forecast how many squirrels will be alive in ten years if we increase the number of trees in the United States by 20%. A biologist would laugh at you. But that is what people ask of economists all the time.”

Economists are rather cocky and don’t acknowledge the limitations of the field. We can look for the causes, try to understand the relationships between variables, figure out what incentives work, which ones don’t and in what settings or the channels through which a policy or action can take effect – but to believe we can predict the exact outcome and mathematically calculate the magnitude of the change in different variables is, well, cute – we can’t, and we’ve been wrong nearly every time we tried.

But perhaps there is something we can learn from physics. In his talk, “There’s Plenty of Room at the Bottom”, Richard Feynman says,

“Atoms on a small scale behave like nothing on a large scale, for they satisfy the laws of quantum mechanics... we are working with different laws, and we can expect to do different things.”

Early models of the atom had subatomic particles arranged like the solar system: a nucleus at the centre with electrons orbiting it. There was no reason to assume such perfect symmetry between the two. It later turned out that the atom looked quite different.

In economics, we assume identical, rational micro agents and predict macro results by aggregating their behaviour. Microeconomics is more intuitive, in a sense more reliable because it is reasonable to assume that individuals try to maximise over certain parameters. But expecting perfect conformity between micro and macro outcomes may not be quite so reasonable. Perhaps we should consider the possibility that on the macro level, the presence of some non-rational agents or collective decision-making leads to thoroughly counterintuitive results.

The results might be quite different if we focus on studying macro behaviour as a whole instead of specifying individual motives for each micro agent beforehand. All we can really say is that when a lot of people’s behaviour is aggregated, on average it tends to produce certain types of results.