subscribe: Posts | Comments

The physics of economics

1 comment

Put two economists in a room together and you’ll get three different opinions on the state and future direction of the economy. Surely economics, the dismal ’science’, could learn something from one of the true sciences, such as physics?

Certainly there have been efforts to do so, particularly among large investment banks and hedge funds, who have used quantitative analysis tools running on powerful computer systems to try to tease out the signals from the noise of price movements, taking into account thousands of different influences from interest rates to tax variations, asset prices to currency exchange rates and much more, all on the basis that there is some underlying predictability, some ‘law’ that governs price movement.

Which makes it all the more surprising that so many of them got it so spectacularly wrong; to the tune of $188 billion and counting. Why?

When I was younger a friend of mine worked at the Science Museum in London, looking after some of the computer-related exhibits. Apparently, buried away in a back room, was a rusting, clogged-up machine that simulated a nation’s economy using water. It had an array of pipes, valves and reservoirs which could be adjusted to mimic the effect of interest rates, taxes and other financial tools, and the flow of water would then show you which way the money might flow through the economy. A genius idea.

I believe the machine may since have been restored to working health. I hope so, because it’s a fascinating example of how economics can appear, at least at a superficial level, to mimic physics. In fact when I first became interested in economics I was convinced that I could see many parallels with the laws of physics.

Consider what happens when you increase taxes on fuel, alcohol or cigarettes, for example. In doing this you are changing the route that money takes; instead of going from the consumer to the retailer, more money will go from the consumer to the government; one valve has been closed slightly, the other opened more. If the taxes are high enough, this will reduce the flow of money into fuel, alcohol or cigarettes because people will stop buying them, and instead spend the money on other things.

And consider interest rates. Lower them below a certain point and you run the risk of creating asset bubbles due to over-cheap borrowing. Raise them too high and you clamp down on discretionary spending and slow the economy. Again, this can be thought of in terms of the flow of water through a valve.

Or so such a model might have you believe. In practice, if you look at the traffic on the UK’s roads, the media crying out about binge-drinking and the massive profits made by tobacco companies, what actually happens when taxes rise is that the money continues to flow into fuel, alcohol and cigarettes, but a larger proportion also flows to the government (while some flows into the black market). And the creation of asset bubbles depends on more than just low interest rates: lax borrowing terms and poor regulation also play a significant part.

Because what physical models of the economy fail to take into account is human nature. I don’t have any addictions, but I have known friends who would walk across town in the freezing cold to get to the all-night petrol station to buy their hit of nicotine. The cost was irrelevant, which is why cigarettes and alcohol are such easy targets for taxation. The same is true of cars; in a country with a below-par public transport system, not to mention the pride that often goes with car ownership, fuel tax is a no-brainer. So while physical models of the economy can be beautiful and elegant, they usually don’t work as predictors of price movement, because people’s behaviour is anything but beautiful and elegant.

There’s another important difference between physics and economics. In the physical world, energy is neither created nor destroyed – it is conserved. However, money can most definitely be created (by inflationary deficit spending on the part of central banks, among other things) and destroyed (through deflationary periods of the economic cycle). Once you remove that important fundamental fact, the correlation between physics and economics becomes even more fuzzy.

Whereas physics (outside of quantum mechanics, anyway) operates without subjectivity, humans are exceptionally subjective. In fact they’re greedy. Not just greedy for money, but greedy for power, political influence and success. So even if it may be in a country’s best long-term economic interest to raise taxes, for example, a politician facing a looming election may instead cut them to create a ‘feel-good’ factor, selling the future prosperity of the nation in order to buy short-term success for him/herself. Physical models cannot predict that.

Which brings me to the current, farcical situation where we have vast numbers of economists, most of them saying different things. And when they’re saying the same things, they often get it wrong anyway. Are you a monetarist? Do you think the UK will follow the US into recession? Will the deflationary effect of house price drops be offset by an inflationary response from central banks? Who knows? Nobody.

Whereas a physical experiment can be set up in such a way that the outcome is predictable and repeatable, in the world of economics there’s no real way of knowing how a particular ’stimulus package’ or ‘discount window rate’ or other increasingly desperate action will affect the rest of the market. So when an economist tells you that something will definitely happen, they are lying. They simply don’t know. Despite the appeal of the idea, economics does not have much in common with physics. It has far more to do with psychology, and that’s even less predictable.

  1. Stephen says:

    Excellent article! It is refreshing to see articles on the current thoughts about the application of physics models to economics that is not full of jingoism and political victim speak.