Blinded with science for too long22 December 2009
Modelling human interactions does present serious stumbling blocks
The many Olde-Worlde Newtonians surviving among us collaborate happily with habitués of the space-time continuum to deal with practical engineering systems and the forces that can usefully be supposed to rule them. A force, these collaborators agree, can still for most operational and calculational purposes be regarded as ‘any cause which tends to change a body’s state of rest or uniform motion in a straight line’. Any technically schooled person accustomed to that definition can be forgiven for wondering what politicians, businessmen and economists mean when they talk of ‘market forces’.
A wag once told me that a market force is any cause that tends to change the state of rest or motion of human prey evading a vendor: but the system of forces that an economist studies, and calls a market, is not truly as crude as that. In fairness I should add that my wag also tried to put an old-fashioned economist’s actual
definition into post-Newtonian terms by describing a market as a portion of continuum allocated to buyers and sellers who are in such close communication that, directly or through dealers, the prices obtainable in one part of the allocation affect the prices paid in other parts.
The market turmoil prevailing in the financial world’s troubled present times has led many politicians, businessmen and economists to question their fundamentals deeply, and some engineers and scientists have looked on in bewilderment as controversy has enlivened the media and debating forums. Much discussed has been a particular bit of economics theory known as the Efficient Market Hypothesis. This has emerged as a result of a movement to remould ‘the dismal science’* on a sound mathematical basis so that calculations can be carried out and predictions made with confidence. This enterprise has apparently been booming along, on the whole with growing profits for proficient practitioners.
According to the E M H a ‘market price’ is the product of a confluence of all available information in the market – which is an efficient one, remember, in which all the players are rational – and therefore the price takes on a truth of its own. The ‘market price’ is the ‘right price’ in what many a believer in the theory has taken to be an absolute sense.
On such ideas have mathematical models been built and financial business based. This sort of thing has been going on for decades and what are now perceived as its errors have become topics of heart searching. Long-time sceptics have not been sparing in their recriminations. The financial press has printed such headlines as ‘Credit crunch causes analysts to rethink rational market theory’, ‘How maths theories can destroy markets’ and ‘The price is not always right and markets can be wrong’.
An academic who has made mollifying comments at a conference of financial analysts (ie specialists employed to use E M H and comparable theory to guide the business of banks, stockbrokers and other financial houses) is Professor Andrew Lo of the Massachusetts Institute of Technology, USA, who is reported to have explained economists’ long allegiance to the theory as a product of ‘physics envy’. He apparently traced this phenomenon back to Professor Paul Samuelson, who won a Nobel Prize for his efforts to recreate economics theory in the image of physics (Newtonian rather than Einsteinian or even, say, CERNian**).
As quoted in The Financial Times, the international business newspaper, Lo said: ‘Economic systems involve human interactions, which almost by definition are more complex than interactions of inanimate objects governed by fixed and known laws of motion’. Which implies that economists have been under-ambitious in seeking to emulate physicists of the Newton era. It does not follow that they would be better off following physics the way it has gone since that era. Modelling human interactions does seem to my lay mind to present serious stumbling blocks.
I am tempted to reflect that humankind’s first major models of the world were not mathematical. They were grammatical systems using verbal symbols. Language has been a wonderful but sometimes misleading model of life and nature. We should not belittle it as merely descriptive, but maybe its models should involve a symbolism and structure beyond the verbal and mathematical ones that we have.
The classical Greeks were perhaps aware of such a forthcoming problem when they contemplated metaphysics, or ‘what comes after physics’, but some practical people shudder away from the kind of thinking that comes under that heading. I guess that most engineers, scientists and economists would consider it a dead end. But I do suspect that economics will have to raise its game somehow. And if it is to succeed it will need help from other sciences: and after that it can look back on physics as a stepping stone.
A colleague has expressed amused doubts about such thoughts and has advised me, courteously, to mind my own business. And we know what our business is, don't we? To keep the lights on, of course. I just hope that whatever economic system we are left with will let us.