Bert hamminga Idealisation and Concretization      000213        Control questions

In economics, the supply and demand model does not describe the apple market. Nor does it describe the shoe market. It describes a market were the commodities are completely indistinguishable. There are no such markets on earth. That is why the model is called an idealisation. The economist makes a theory about a thing that does not exist, hoping it looks enough like existing things to learn something about them.

This is not something peculiar for economics. Ideal models are used in all sciences. Galileo's law of falling bodies, for instance (falling bodies accelerate according to a simple formula: s=�gt²), holds in the absence of any resistance while the body falls. When you jump from an aeroplane, you fall pretty much according to Galileo in the beginning. But after you gained some speed the wind on your body starts to count. Well before 200 km/hour, depending on your weight, size, how you hold your body and what you wear, you will even completely stop accelerating and assume constant speed.
Some falls look more like Galileo's model than others. On Venus and Saturn Galileo is far off the mark, but falls on the Moon conform Galileo pretty well. Strictly speaking, one cannot expect ever to have a "pure" Galilean fall anywhere, but some falls are so close to Galileo that the error will be hard to measure.

Scientists typically start with a highly idealised model (idealisation), assuming lots of forces to be absent. Once they feel they control that idealised model, they set out to improve it by adding the many forces not yet taken into account thus making the model more "realistic". This is called concretisation.

In Galileo's case, you need to add aerodynamic assumptions to calculate how the resistance of something falling in air depends on the form of the body (parachutes, for instance, differ drastically from wings), the weight relative to the air in which it is falling, and the speed. You learn these concretisations on aerotechnic schools and it is very difficult. Many newly designed  aeroplanes and parachutes still turn out not to behave according to the predictions of established theory. There are always surprises once they are tested! That is science, you can never be sure. That is why designers finally always need test pilots and test parachutists.

In many sciences, the ideal conditions supposed by the ideal models can be approximated by experimental set ups. Galileo predicts that in vacuum, a feather falls just like a cannon ball. If you pump the air out of a long vertical airtight tube, it is not completely void of air, but almost. And yes, you will see that in that tube, a feather falls almost like a cannon ball.

Economists have little opportunities like that. So, their fantasies about what would happen on markets where

• every single good is exactly like the other
• all suppliers know what all consumers are would buy at every price
• all suppliers know what all their fellow suppliers would offer at every price
• all consumers know all suppliers and the price they offer for that product
• suppliers can instantaneously adapt supply if they experience a price change
• consumers can instantaneously change their amounts purchased if they experience a price change

...there is not something like an airtight tube in which these circumstances can be created, to see whether really will happen what the theory says!

That is why economists are constantly looking for good non-experimental reasons to work with one particular idealisation rather than with another. A very good example is the way Adam Smith, one of the founders of economics, gives his reasons for the idealisations from which he derives the theorem of the tendency of rate of profit to equalise over an economy.