A very interesting article on the use of a machine by the famous economist Bill Phillips (of Phillips curve fame).

It is 2 metres (7ft) tall, 1.5 metres wide and a metre deep. It runs on water and most of the time it is screened off at the back of a lecture room in Cambridge. But when the nine members of the Bank of England's monetary policy committee announce their latest decision on interest rates today they will owe a debt of gratitude to the computer built in a garage in south Croydon by Bill Phillips - an engineer turned economist from New Zealand - almost 60 years ago.

A sensation when it was unveiled at the London School of Economics in 1949, the Phillips machine used hydraulics to model the workings of the British economy but now looks, at first glance, like the brainchild of a nutty professor. Where the Bank's team of in-house economists are equipped with state-of- the-art digital computers, the profession's first stab at modelling was very much a do-it-yourself affair with a whiff of the Heath Robinson about it.

The prototype was an odd assortment of tanks, pipes, sluices and valves, with water pumped around the machine by a motor cannibalised from the windscreen wiper of a Lancaster bomber. Bits of filed-down Perspex and fishing line were used to channel the coloured dyes that mimicked the flow of income round the economy into consumer spending, taxes, investment and exports. Phillips and Walter Newlyn, who helped piece the machine together at the end of the 1940s, experimented with treacle and methylated spirits before deciding that coloured water was the best way of displaying the way money circulates around the economy.

Irving Fisher also used hydraulics to model an economy because he didn't think the math was there to work out the general equilibrium (Gerard Debreu finally worked this out in the 1950s.) Hydraulics worked for Phillips, however, to demonstrate a coordination problem of policy:

...one early demonstration of the machine displayed the difficulties that can arise when monetary and fiscal policy are not synchronised. Phillips asked one of his students to be chancellor of the exchequer and control taxes and spending; the other to be governor of the Bank and control interest rates. Predictably, the policies were uncoordinated and the upshot was that water overflowed on to the floor.

Now if they could work out that the two people pouring the water are engaged in a game...

(h/t: Tyler Cowen)

Labels: economics