Paul Mason wants to be an economist


This is ever so slightly strange. Paul Mason has decided that he'd like to be an economist, figure out how the economy works in detail.

If I could rub an empty lager can, and get a genie to appear and grant me one wish for 2015, it would be for something apparently banal but revolutionary: an accurate simulation of the economy.

It would be multilayered: it would model the microeconomics of my home area, allowing me to test the lurid worries of my neighbours about the opening of a second tattoo shop. It would model the real Britain – including the sex work, the cybercrime and the drug deals. And at a macro-level it would model the whole world – from the effects of a factory collapsing on its workers in Bangladesh to those of fast fashion on the consumption habits of teenage girls here.

The reason we need such models is that the ones provided by economics are pretty useless – particularly when modelling instability, complexity and change. Mainstream economic models rely on the 150-year-old assumption that capitalism’s tendency is towards equilibrium, and that everybody acts rationally; they struggle to accommodate sudden shocks. About 20 years ago economists decided to abandon data and go for an ever more abstract series of models that are logically consistent but not tested against facts, and unable to predict real crises.

There's really only one slight problem with that idea. Hayek showed that, in theory, such detailed planning of an economy simply isn't possible. Knowledge is local, the centre cannot possible gather enough of it in apposite time scales to be able to produce such models. We end up with the end result that we can only use the economy itself as the model of the economy.

It's entirely true that not everyone actually believes Hayek on this point. There's always those who think that just a little more computing power, just a little more scientific socialism, will enable us to overcome this wastefulness of capitalism and markets.

But as the socialists themselves found out, this just isn't true. Or, sa this great essay points out, In Soviet Union Optimisation Problem Solves You.

I said before that increasing the number of variables by a factor of 1000 increases the time needed by a factor of about 30 billion. To cancel this out would need a computer about 30 billion times faster, which would need about 35 doublings of computing speed, taking, if Moore's rule-of-thumb continues to hold, another half century. But my factor of 1000 for prices was quite arbitrary; if it's really more like a million, then we're talking about increasing the computation by a factor of 1021 (a more-than-astronomical, rather a chemical, increase), which is just under 70 doublings, or just over a century of Moore's Law.

Now that is talking about planning the economy. But the same is true of modeling it. For if we can model an economy then, as Mason desires, we would be able to plan it. And the reason that we can't plan it is because we can't model it. It is just one of these things that we cannot do.

Anyway, full marks to Mason for wanting to understand more about the economy, limited points for wanting to know more about the effects of policy but really, shouldn't he have known that his desires are impossible before he became an economics editor?